Estimated reading time: 12–14 minutes.

Monaco sit at the top of Ligue 1 with 14 points from six matches: four wins, two draws and no defeats. The FotMob league table, checked on 14 October 2026, places them one point ahead of Lille and Rennes, with ten goals scored and five conceded. That is a strong opening to the campaign, but a league position cannot explain how a team moves the ball forward, which routes become threatening or where its construction tends to stall.

Those questions are the starting point for this Monaco build-up analysis. Using 202 possessions that began in the first third across six matches, I grouped similar sequences, examined their passing patterns and compared the threat they accumulated in a designated target zone.

The aim is to show how football data analysis can turn a collection of possessions into recognisable behaviours, giving an analyst a useful team profile before beginning a more detailed video review.

Monaco provide the example, but the approach can be applied to any team with suitable event data. It can describe how a side progresses in possession and, with the perspective reversed, the types of progression it allows opponents. Even this relatively brief exploration produces specific questions about left-sided progression, circulation trapped on the defensive right and the importance of connecting different parts of the pitch.

Subscribe now

Why analyse possessions that start in the first third?

A possession recovered close to the opposition penalty area begins with a very different problem from one starting near a team’s own goal. Combining them without distinguishing their starting positions can blur the behaviour we want to study: some attacks are already close to dangerous territory before the first pass, while others must travel through most of the pitch. For this analysis, restricting the sample to possessions beginning in the first third makes the subject more precise: how does Monaco construct its progression from deep positions?

That restriction applies to the beginning of the possession, not to every subsequent pass. Once a qualifying possession advances into midfield or the final third, its successful passes remain part of the analysis. The analysis covers 928 successful passes across 202 possessions that began in Monaco’s defensive third. Every subsequent successful pass remains in the sample, allowing us to follow those sequences as they develop through midfield and into attacking areas.

The resulting sample includes short spells that barely leave the defensive third and longer constructions that reach advanced areas. That variation is useful because it allows us to compare different ways a build-up develops, rather than selecting only successful attacks. It also means that a profile’s name describes the sequence observed in the data; it should not automatically be read as the team’s intention when the possession began.

How KMeans identifies football build-up patterns

KMeans groups observations that resemble one another across a chosen set of characteristics. Here, each observation is a possession, described by eleven variables: nine spatial shares covering three thirds and three channels, the number of triangulations and the duration of the sequence. Together, these describe where the passing activity is distributed, how long the possession lasts and the presence of a particular combination pattern.

Before grouping the possessions, the variables are standardised so that their different measurement scales do not dominate the comparison. KMeans then assigns each possession to the nearest group centre and repeatedly adjusts those centres to improve the grouping. In football terms, the result is a set of families: possessions concentrated on the defensive right, sequences circulating through central-left midfield, or longer constructions extending across several thirds.

Seven clusters were retained because they provided a useful balance between separation and football interpretation. Inertia describes how tightly observations sit around their group centres, while silhouette assesses how well they fit their own group compared with alternatives. These diagnostics help assess the grouping, but the practical justification also comes from the heatmap: the seven clusters express recognisable differences without turning this introductory analysis into a catalogue of increasingly narrow subtypes.

The heatmap displays those characteristics on a standardised scale. Red indicates a value above the sample average and blue a value below it; neither colour means inherently good or bad football. Crucially, xT was not one of the eleven clustering variables. It is used afterwards to examine how the identified profiles relate to threat, while remembering that a profile defined partly by advanced territory will naturally have opportunities to register higher threat there.

Seven profiles, very different progression outcomes

The first useful distinction is between frequency and danger. A behaviour can account for a large share of a team’s possessions while contributing little threat in the target zone, and a relatively uncommon pattern can carry a disproportionate share of that threat. Monaco’s sample makes this contrast particularly clear: the three profiles concentrated in the first third contain 130 of the 202 possessions, or 64.4%, but contribute only about 1.3% of the recorded target-zone xT.

The seven profiles differ substantially in frequency, duration and threat. Central First-Third Quick Build-Up is the largest group, with 47 possessions averaging just 2.5 seconds, but only one registers target-zone xT. Right-Side First-Third Retention contains 43 possessions, while Left-Side First-Third Short Build-Up contains 40. Both average approximately 5.6 seconds, and neither produces target-zone xT in this sample. Together, these profiles describe frequent sequences whose passing activity remains predominantly close to the defensive starting area.

The midfield profiles show greater progression, although their attacking returns remain uneven. Central-Left Midfield Circulation contains 32 possessions averaging 31.8 seconds, with positive target-zone xT in four of them and an average contribution of 0.0153 per possession. Right-Lane Progression into Midfield contains 17 possessions averaging 14.8 seconds, with positive contributions in three and an average of 0.0098. The central-left group therefore accumulates more threat per possession on average, but its longer circulation does not consistently translate into valuable receptions inside the target zone.

The strongest attacking returns belong to two smaller groups. Central-Left Final-Third Progression contains twelve possessions averaging 21.8 seconds, ten of which register target-zone xT, with an average of 0.0907 per possession. Sustained Multi-Third Progression contains eleven possessions averaging 41.6 seconds, with positive contributions in five and an average of 0.0451. These figures distinguish a comparatively consistent route into threatening central-left areas from a longer, more sustained form of progression whose attacking output varies more between possessions.

The central first-third profile is the most frequent, with 47 possessions, but also the shortest. Its 83 successful passes amount to fewer than two per possession, and almost all its average spatial share remains in the first third. The label “quick build-up” therefore describes short duration; it does not mean these were consistently fast, successful attacks travelling from one end of the pitch to the other.

At the other end of the distribution, the two most threatening profiles contain just 23 possessions combined. Central-Left Final-Third Progression and Sustained Multi-Third Progression represent 11.4% of the sample yet account for approximately 69.8% of its accumulated target-zone xT. Their recorded entry counts also add up to 14 of the dataset’s 23 entries, providing another view of how concentrated the attacking output is within these groups.

What the xT boxplot says about Monaco’s most dangerous build-up

The metric in the boxplot sums the xT values at the destinations of passes inside the target zone, separately for each possession. It captures the accumulated value of those receiving locations rather than the net xT gained between the start and end of a pass. Several receptions can contribute within one possession, so the total should not be interpreted as a probability of scoring or as a count of chances created.

This distinction matters when reading the distributions. A zero means the possession registered no accumulated value under this target-zone measure; it does not mean every action in the sequence was useless. Equally, a high total describes valuable passing destinations within the defined zone, without telling us on its own whether the move ended in a shot. Within those boundaries, the boxplot gives a clear comparison of which build-up profiles repeatedly connect deep starts with threatening territory.

Central-Left Final-Third Progression stands out most clearly. Ten of its twelve possessions record positive target-zone xT, its median is approximately 0.0525, and its average is 0.0907. The profile contributes roughly 48% of the entire sample’s accumulated target-zone xT despite containing fewer than 6% of the possessions. Its importance is therefore visible both in the distribution and in its overall contribution.

Sustained Multi-Third Progression provides a second route to danger, but with a different degree of consistency. Five of its eleven possessions produce positive target-zone xT, giving it a median of zero even though the upper part of the distribution reaches substantial values. Its average of 0.0451 is around half that of the central-left final-third profile, while its total contribution is about 22% of the sample. Both profiles matter, but the first reaches positive target-zone output much more regularly within its small group.

The Central-Left Midfield Circulation profile illustrates why individual high values need to be read alongside the whole distribution. It contains a possession with approximately 0.2575 target-zone xT, comparable to the highest observation in the central-left final-third group, yet only four of its 32 possessions register a positive value. Those occasional dangerous sequences are worth reviewing, but they do not make every long spell of midfield circulation equally threatening.

Longer possession does not automatically mean greater threat

The comparison between the two central-left profiles is especially revealing. Midfield Circulation averages 31.8 seconds and 8.7 successful passes per possession, whereas Final-Third Progression averages 21.8 seconds and 7.5 passes. Despite being shorter and involving fewer successful passes, the final-third profile produces almost six times as much target-zone xT per possession on average.

This is a useful reminder of what a possession count or passing total can conceal. Monaco’s 279 successful passes in the midfield-circulation group are the largest total of any profile, but that volume does not translate into the strongest target-zone output. The distinction lies in where the sequence develops and the receiving locations it reaches, rather than simply how long Monaco keep passing.

The longer Sustained Multi-Third Progression profile tells a complementary story. Its possessions average 41.6 seconds and 13.2 successful passes, with eight recorded triangulations across eleven sequences. That is a different way of moving forward from the shorter central-left final-third pattern: more sustained combination, a broader distribution across thirds and greater variation in whether the possession eventually accumulates target-zone threat.

Reading the pass flow: where Monaco progress and where they stall

The pass-flow maps add the spatial detail needed to interpret those differences. Their blue shading represents the number of successful passes originating in each cell, using the same colour scale across profiles. The black arrows represent average passing direction and average pass length within each cell, allowing us to compare the main tendencies of the different groups without plotting every individual event.

Because these are aggregated maps, an arrow does not represent one specific move and the full set of arrows should not be followed as a single possession. A pale cell can still receive passes, while a dark one tells us that many passes started there. This is particularly relevant near the target zone: a team can pass into an advanced area without producing many successful passes out of it, so receiving threat and pass-origin density answer different questions.

The Right-Side First-Third Retention profile provides the clearest example of restricted progression. It contains 43 possessions and 132 successful passes, yet records no target-zone entries and no target-zone xT. Its mean spatial share is approximately 98.6% in the first third, and the flow concentrates heavily around the defensive right. The map and the possession-level numbers describe the same limitation: these sequences rarely develop beyond their initial area.

This supports a specific tactical observation about the sample: when Monaco remain confined to the right flank in their own third, their progression struggles to develop. It does not establish whether the opponent forced that restriction, whether Monaco deliberately recycled possession or whether another event interrupted the sequence. Those are questions for video, but the data have already identified a coherent group of possessions in which to investigate them.

It would also be a mistake to turn that observation into a general claim that Monaco cannot progress on the right. Right-Lane Progression into Midfield contains 17 possessions, with three recorded target-zone entries and positive target-zone xT in three sequences. The sustained profile also has a strong final-third-right characteristic in the heatmap. The relevant distinction is between being confined to the defensive right and using the right side as part of a progression that reaches other areas.

Monaco’s leftward tendency needs a territorial distinction

Taken together, the flow maps suggest a tendency for important progression routes to develop towards the left, particularly through central-left midfield and into central-left final-third areas. The strongest target-zone xT profile fits that reading: its heatmap values stand out in the central and left channels of the final third, while its flow extends well beyond the defensive starting area. The spatial description and the threat distribution reinforce one another.

However, the Left-Side First-Third Short Build-Up group contains 40 possessions and produces no target-zone xT. Simply placing the ball on the left does not make a sequence dangerous, just as using the right does not automatically prevent progress. What differentiates the more threatening leftward route is its connection with advanced territory: moving through and beyond midfield into receiving locations that carry greater value.

The contrast between these groups is useful for opposition analysis because it makes a broad observation more actionable. “Monaco tend to progress left” is a starting point, while “which connections allow Monaco to move from a deep start into central-left final-third receptions?” is a question that can guide a focused review. It directs attention towards the development of the possession instead of treating a flank preference as a complete tactical explanation.

Connecting the flanks: a hypothesis worth taking to video

The broader circulation visible in the more developed profiles raises another useful hypothesis: allowing Monaco to connect both flanks may help them escape restricted areas and sustain threatening progress. The Sustained Multi-Third Progression map is especially relevant because its activity extends across several zones, and its longer sequences include more triangulations than the other groups. This suggests combinations and connections worth inspecting rather than isolated forward passes alone.

There is a practical difference between observing activity on both flanks across a group and demonstrating a switch from one flank to the other inside an individual possession. The flow establishes the former; a sequential review is needed for the latter. The next question is therefore whether Monaco’s dangerous possessions use lateral connections to create a new forward route, and at what stage of the build-up those connections become decisive.

That investigation can cover the defensive, middle and final thirds without assuming the effect is identical in all three. Video can show whether a switch changes the opponent’s shape, opens a receiving lane or merely recycles the ball without creating an advantage. The cluster analysis has already narrowed the search to the profiles most likely to make that comparison useful.

Using possession clusters to guide video analysis

Even at this introductory level, the analysis provides a grounded initial profile: frequent short sequences near the defensive starting area, restricted progression in the defensive-right retention group, a threatening central-left final-third route and a longer combination-based alternative. That is enough to organise a meaningful video session. Instead of watching six matches with a general instruction to study Monaco’s build-up, the analyst can select possessions that represent each behaviour and compare how they unfold.

For the defensive-right group, the review can examine available passing options, the opponent’s positioning and the moments when progression breaks down. For the central-left final-third group, it can identify which connections make advanced receptions possible and whether the same mechanism recurs. Comparing those clips with the quieter midfield-circulation possessions would also help explain why some sequences reach valuable territory while others accumulate passes without the same return.

This makes the analysis more focused without pretending that a cluster label explains everything. Event data locate and quantify the pattern; video reveals the surrounding movements, pressure and decisions. A useful team profile emerges from that combination, with the initial statistical picture giving the analyst specific observations to test rather than a long list of unstructured impressions.

Analysing progression for and against any team

The same framework can be applied from the defensive perspective by selecting opponents’ possessions against the team being studied. An analyst could then identify which forms of build-up the team regularly contains and which ones are associated with valuable receptions or entries into the target zone. This moves the discussion beyond how a side attacks and towards the kinds of progression its defensive behaviour permits.

The comparison requires a consistent starting-point rule, spatial definition and passing direction. If profiles are to be compared directly between teams, the groups should also come from a common reference or be carefully matched by their characteristics; a cluster number has no fixed tactical meaning across independently fitted models. With those conditions in place, the method can support both a team’s own development work and preparation for a specific opponent.

That wider application is the purpose of this case study. Monaco’s patterns are interesting in their own right, but the repeatable process is the more valuable result: select a coherent set of possessions, group their characteristics, interpret the profiles and examine the threat associated with each one. The findings then lead back to the match, where the analyst can investigate the football behind the numbers.

Monaco vs Toulouse: taking the analysis into the Football Hacking Web App

The video below takes the discussion into the Football Hacking Web App using Monaco’s 2–2 home draw with Toulouse. I use the match to explore insights from the available tools and show how a broader data-led team profile can provide a starting point for examining an individual game. The purpose is to connect the patterns discussed here with a practical analysis workflow, while keeping the focus on team behaviour.

What this Monaco build-up analysis reveals

Across 202 possessions, clustering separates behaviours that a single passing total would merge together. The most frequent profiles spend much of their activity near the starting area, while a small group of central-left final-third and sustained multi-third possessions carries most of the accumulated target-zone threat. The boxplot distinguishes consistent danger from occasional high-value sequences, and the flow maps show where those differences take shape on the pitch.

There is already enough here to form a useful initial view of Monaco and make the next stage of analysis more deliberate. Investigate the restriction on the defensive right, trace the routes into central-left advanced areas and examine how connections between flanks support the more developed possessions. That is what makes clustering valuable in football analysis: it gives recurring behaviours a measurable form and helps us return to the video knowing what deserves a closer look.

When people evaluate a football player, they often begin with goals, assists, shots, and minutes played. Those numbers are useful, but they leave important questions unanswered. Where does a player fit into the team’s passing structure? Does he regularly connect teammates, influence the circulation of the ball, or become a preferred target? Where on the pitch does he receive and play successful passes, and how does the territorial threat associated with those actions change?

The new Player Analysis feature in the Football Hacking Web App is designed to help explore those questions. It combines a match-by-match view of a player’s role in the passing network with season-long pitch maps of successful open-play passes. To show how the feature works, this article uses the available example for Giorgian de Arrascaeta of Flamengo.

The goal is not to reduce a player to a single score or label. It is to give analysts a clearer way to inspect recurring patterns, notice how a role varies from match to match, and decide what deserves a closer look in the underlying games.

What the Arrascaeta player analysis shows

The example covers 21 appearances, with all 21 matches included in the role chart. Across the data shown, Arrascaeta has 740 received passes and 604 made passes. These totals provide context for the maps, while the charts show how those actions relate to his position within Flamengo’s passing structure and to their locations on the pitch.

The feature has three main visual components:

• A player roles in the pass network chart, with one point for each eligible match.

• A received passes pitch map, showing where the successful passes to Arrascaeta originated.

• A made passes pitch map, showing where Arrascaeta played successful passes from.

Together, these views move the analysis beyond a season total. They help distinguish a player’s recurring involvement from the different ways he may contribute in individual matches.

Reading the pass-network role chart

The role chart places Receiving Influence · PageRank on the horizontal axis and Flow Connector · Load Centrality on the vertical axis. Each point represents one match. The chart is divided into four regions, each describing a different relationship between the player and the passing network.

The vertical and horizontal dashed lines divide the chart according to the match-level reference values. A point’s position therefore helps answer a relative question: in that match, was Arrascaeta more involved as a connector, more influential as a receiving option, both, or neither compared with the players in the network?

In this example, many points appear in the lower-right “Preferred Target” region. This pattern suggests that Arrascaeta often ranks relatively high in receiving influence while sitting below the chart’s reference level for flow connector centrality. Put simply, teammates frequently connect with him, but his role in a particular match is not always defined by being one of the network’s main bridges for passing flow.

Several points appear in the upper-right “Structural Hub” region. In those matches, the chart indicates a stronger combination of receiving influence and flow centrality. Arrascaeta is not only an influential option to find; he also occupies a more central position in how passes move through the network. That can be a useful distinction when comparing his involvement across performances: a player may remain an important receiver while becoming more structurally connected in some games than in others.

There are also points in the upper-left “Flow Connector” region. These suggest matches in which Arrascaeta’s connector score is relatively high even though his receiving influence is below the reference line. This could correspond to a different kind of network contribution from his preferred-target matches. The chart identifies where that difference appears; match footage and event context are still needed to explain the tactical reason behind it.

Finally, some points sit in the lower-left “Peripheral Role” region. In those matches, both measures are comparatively low within the plotted network. This does not mean that Arrascaeta played poorly or had no influence on the result. It means that these two network measures place him in a less central role relative to the other eligible players in that match. A player can contribute through actions that the chart does not capture, and a less central network position may also reflect the team’s tactical plan, game state, or the way possession was distributed.

The key is the variation across the 21 matches. Rather than assigning Arrascaeta one fixed role for the whole season, the chart shows that his relationship with Flamengo’s passing network changes. The recurring concentration in the Preferred Target area, together with points in the Structural Hub and Flow Connector areas, provides a starting point for investigating when and how his network role shifts.

What does PageRank mean in a passing network?

In a passing network, players are connected through passes. PageRank helps describe a player’s receiving influence by considering not only how often he is found, but also the network context of those connections. It offers a way to identify players who occupy influential receiving positions within the structure.

That distinction matters because raw pass counts alone can be misleading. Two players may receive a similar number of passes, but their connections may sit in different parts of the network or involve different teammates. PageRank adds a structural perspective to the count.

For this feature, the role measures use the full team network for each match and are normalized across eligible players from both teams. That means the chart should be read as a match-by-match comparison within the network, not as an absolute rating that can be compared without qualification across every context.

What does Load Centrality reveal?

The chart’s Load Centrality measure helps describe a player’s position as a connector in the flow of the network. A player with a higher value may lie on more of the routes that link other players, which can make him an important intermediary in the team’s passing structure.

For Arrascaeta, the chart shows that high receiving influence does not always coincide with high connector centrality. Many points sit to the right of the vertical reference line and below the horizontal one, while a smaller group occupies the upper-right area. This separation is analytically useful: being a frequent or influential target and acting as a central bridge are related possibilities, but they are not the same role.

The distinction can guide more focused review. An analyst might compare matches from the Preferred Target region with those in the Structural Hub region, then inspect whether the difference relates to teammates’ positions, the opponent’s defensive shape, possession volume, or the phase of the game. The chart does not answer every tactical question by itself; it helps identify which comparisons are worth making.

How recency adds context

The points are shaded by recency. Lighter points represent older matches, while darker points represent more recent ones. This allows the chart to show not only where the player’s role has varied, but also when those variations occurred.

The recency scale is helpful when reviewing change over time. For example, an analyst can see whether recent points are concentrated in a different region from earlier ones, or whether a role pattern appears throughout the selected sample. It should still be treated as context rather than proof of a trend: a few recent matches may differ for reasons that require match-level review.

One practical approach is to select a recent cluster or outlying point and compare it with a similar match from earlier in the sample. That turns the chart into a navigation tool for analysis, helping the user move from a season overview to specific games that may explain the pattern.

The received-passes map: where Arrascaeta is found

The first pitch map is titled “Received passes · Origin at teammate.” It plots the origin of successful open-play passes played by teammates to Arrascaeta. In other words, each point represents where a teammate was positioned when making a pass that was received by the selected player.

In the example, the points span a broad area of the pitch, with visible concentrations across central and advanced zones as well as passes originating from deeper positions. This suggests that the map captures a range of situations in which Arrascaeta becomes an option, rather than a single fixed location from which teammates find him.

The color scale represents ΔxT, or the change in expected territorial threat associated with the pass. Blue marks a gain, red marks a reduction, and the shared scale is centered around zero. Most points appear close to neutral grey, with a smaller number of more clearly blue or red actions. That is a useful reminder that not every successful pass produces a large change in territorial threat, and that a pass can still matter for circulation, retention, or the next action.

The made-passes map: where Arrascaeta plays from

The second map, “Made passes · Origin at selected player,” plots the pitch locations from which Arrascaeta played successful open-play passes. As with the received-pass map, each point is a pass origin, and color indicates the associated change in xT.

The distribution again covers a wide range of locations, with many points in central and advanced areas. This lets the analyst compare where Arrascaeta tends to receive the ball with where he tends to play from. The two maps answer different questions: the first looks at the locations of his teammates when they pass to him; the second looks at his own locations when he makes a pass.

The same red-to-blue scale is used for both maps, making the visual comparison easier. In the screenshot, many actions on both maps sit close to the neutral center of the scale, while a smaller number show clearer gains or reductions in territorial threat. This is not a rating of whether a pass was “good” or “bad.” A negative ΔxT means lower modeled territorial threat after the action; it does not, by itself, establish that the decision was poor. A backward or lateral pass may help Flamengo retain possession or create a better route for a later action.

Why the two pitch maps use the same scale

The maps share a symmetric ΔxT scale, which places gains and reductions on opposite sides of zero. Using the same scale makes it easier to compare colors between the received-pass and made-pass maps without interpreting each map on its own color range.

The maps also follow the same coordinate convention as the match-analysis page, with the attack direction shown on the pitch. That consistency matters because it helps users build familiarity across different areas of the app. A user can move between match analysis and player analysis without having to relearn how the pitch is oriented or how the color scale is interpreted.

The data shown is limited to successful open-play passes with inferred receivers. Set pieces are excluded from the maps. Receiver identification is inferred from the event data, so the maps should be read with that methodological detail in mind. This filtering keeps the visual focus on open-play passing patterns, while also making clear what kinds of actions are represented.

How to use the feature in a real analysis

A useful analysis usually begins with a question, not a screenshot. For example: does Arrascaeta occupy a similar receiving role in every match, or does his network position change? Which games place him in the Structural Hub region? Do matches where he appears as a Preferred Target show a different distribution of received-pass origins? When his own pass origins shift, does the network chart also show a change in connector centrality?

The feature supports a workflow built around those questions. Start with the role chart to identify recurring regions and matches that stand out. Then use the pitch maps to examine the locations behind the player’s involvement. Finally, return to the match itself to evaluate the tactical context. The charts help narrow the review; they do not replace video, scouting judgment, or a wider assessment of performance.

For a recruitment or squad-planning workflow, this kind of player analysis can also help frame comparisons. Rather than asking only whether two players have similar goal or assist totals, an analyst can investigate whether their receiving influence, network position, and pass-origin distributions look alike. Any comparison still needs consistent samples, competition context, playing time, and tactical role. The feature makes those questions easier to explore; it does not declare that players are interchangeable.

Player analysis built for exploration

The Arrascaeta example shows why a player’s contribution is difficult to describe with a single label. Across the 21-match role chart, his points appear in multiple regions, with a recurring presence as a Preferred Target and additional matches in which the chart places him as a Structural Hub or Flow Connector. The pitch maps add another layer by showing where teammates’ successful passes to him originate and where he plays successful passes from.

That combination makes the new Football Hacking Player Analysis feature useful for analysts, coaches, scouts, and football fans who want to explore passing structure in more detail. It connects a match-by-match network view with season-level spatial maps, while keeping the underlying measures visible enough for the user to investigate the evidence rather than rely on a single summary label.

Watch the Player Analysis feature in action

Explore Giorgian de Arrascaeta’s passing-network roles and season-long received-pass and made-pass maps in the Football Hacking Web App.

A team’s passing style is easy to describe in broad terms. We might call a side patient, direct, possession-heavy or dependent on a playmaker. Those descriptions can be useful, but they can also flatten very different behaviours into the same label. Two teams may both keep the ball, for example, while distributing the work of circulation in completely different ways. One may rely on a small number of players to connect its passing network. Another may involve more players in moving the ball forward.

I wanted to explore that difference using two measures: Load Gini and final-third pass share. The first helps describe how evenly passing responsibility is distributed across a team’s network. The second measures what share of a team’s passes reach the final third. I used KMeans clustering to group 150 teams from eight European leagues according to their values on those two measures, then connected the resulting profiles to the outcomes in 438 matches.

The project is an early-season investigation, not a completed study. Some clubs have played only a handful of league matches, and some profile matchups have very few observations. Still, the first picture is useful. It gives us a way to compare teams beyond league standings, identify interesting differences between competitions, and ask more precise questions about how passing profiles relate to match results.

The point is not to declare that one cluster represents the “best” kind of football. It is to build a framework for examining what teams do with the ball, then test the patterns against more matches and more information as the season develops.

Subscribe now

What Load Gini measures in a passing network

Load centrality describes a player’s importance to the flow of a network. In a passing network, that can help identify players who act as important links in the team’s circulation. A player with a high Load value may sit on many of the routes through which the ball moves from one part of the team to another.

The Gini coefficient measures how unevenly something is distributed across a group. When applied to player Load values, it gives us a summary of how concentrated network responsibility is within a team.

A higher Load Gini means the distribution is more concentrated: some players carry a larger share of the network’s Load. A lower Load Gini means that responsibility is more evenly spread among the players included in the calculation.

This does not make one profile automatically better than the other. A concentrated structure may be intentional. A team may rely on a central midfielder, a deep playmaker or a centre-back to organise circulation. That player could be the main route through which possession moves. In another team, several players may share those responsibilities, producing a more balanced distribution.

Neither pattern tells us by itself whether the team progresses effectively, creates chances or wins matches. Load Gini describes the distribution of network responsibility. It does not explain why that distribution exists or what happens after the ball moves through those players.

A concentrated network might reflect a clear structure built around an influential organiser. It might also indicate that a team has few alternative routes through midfield. A more evenly distributed network could give a side several ways to circulate possession, but it could also reflect a lack of a clearly central connector. The number becomes more useful when we compare it with other evidence.

What final-third pass share tells us

The final-third pass share is the proportion of a team’s passes that go into the final third of the pitch. If the share rises, more of the team’s passing activity is reaching that advanced area. If it falls, a smaller proportion of its passes are entering the final third.

This metric describes where passes go. It does not measure the quality of the pass, whether it breaks a defensive line, or whether the move leads to a shot. A pass into the final third may be a valuable step in an attack, but it may also be safely recycled across the edge of a compact defence. The measure shows the frequency of reaching the zone, not the danger created there.

That distinction matters when interpreting the results. A high final-third pass share does not necessarily mean “more attacking” in every sense. It may indicate that a team reaches advanced areas frequently, but we would need additional measures to determine whether it creates high-quality chances or turns that access into goals.

The two metrics work together because they answer different questions. Load Gini asks how concentrated is the passing responsibility? Final-third pass share asks how much of the passing reaches an advanced area? One describes the network’s internal distribution; the other describes a broad spatial outcome of passing.

How KMeans created six passing profiles

KMeans grouped the 150 team observations into six clusters using Load Gini and final-third pass share. Instead of looking at each club as an isolated point, the model grouped teams with similar combinations of the two measures.

The resulting profiles are described as:

• Average Load Concentration / Average Final-Third Pass Share

• Lower Load Concentration / Average Final-Third Pass Share

• Higher Load Concentration / Higher Final-Third Pass Share

• Higher Load Concentration / Lower Final-Third Pass Share

• Average Load Concentration / Lower Final-Third Pass Share

• Lower Load Concentration / Higher Final-Third Pass Share

These names are relative descriptions of the clusters in this dataset. “Higher” and “lower” do not mean that a team has crossed a universal threshold, and the labels do not rank the clusters from good to bad. They make it easier to discuss the groups, but they should not be mistaken for complete tactical descriptions.

The largest group contains 46 teams with average Load concentration and average final-third pass share. Another 30 teams fall into the higher-concentration, higher-share profile. At the other end, the lower-concentration, higher-share profile contains just two teams: Strasbourg and Académico de Viseu.

The uneven size of the clusters is important. The larger groups give us more opportunities to compare teams and match outcomes. The two-team cluster is interesting because it represents a rare combination in this sample, but any conclusions about it must stay tentative. A profile represented by two clubs cannot support the same kind of broad comparison as one represented by 30 or 46.

The cluster model also does not tell us why a team falls into a particular group. It identifies similarity across the chosen variables. It does not know whether a player is a specialist passer, whether the team changed formation, whether an opponent pressed aggressively, or whether a match state altered the passing pattern. Those explanations require further analysis.

The broad shape of the team map

The overall passing-profile chart shows that teams are spread across several combinations of the two measures. Many sit around the middle of both. Others stand out through a more unusual pairing, such as high Load concentration with a high final-third share or low Load concentration with a high share.

One of the most useful points from the chart is that the two dimensions do not collapse into one simple idea of “progressive passing.” A team can have high Load concentration and a relatively high final-third share, suggesting that the measured network responsibility is concentrated while passing reaches advanced territory often. Another team can have similarly high concentration but a lower final-third share. Those sides may have comparable Load Gini values but different outcomes in the spatial measure.

The same is true at the other end. Two teams may share a lower Load Gini but differ in final-third pass share. One may distribute network responsibility broadly while reaching advanced areas at an average rate. Another may also distribute responsibility relatively broadly but have a higher share of passes enter the final third.

That is why the profiles are more informative as combinations than as single labels. Looking only at Load Gini would miss whether the team’s passing reaches advanced territory. Looking only at final-third pass share would miss how evenly the network responsibility is distributed.

The model gives us a starting map. The next question is what kinds of football sit behind those points. Does a team with concentrated Load build through one organiser? Does a team with a high final-third share progress through central combinations, wide switches or direct passes? Does a more evenly distributed network give the side several routes forward? The two-variable model cannot answer those questions on its own, but it helps identify teams and matchups worth examining more closely.

The Eredivisie’s early signal

The clearest league-level difference in this sample comes from the Eredivisie. Its teams have an average final-third pass share of approximately 28.1%, the highest of the eight leagues analysed. The other league averages fall between about 24.2% and 25.6%.

That does not prove that Dutch football is inherently more advanced or more attacking. It is an early-season result from this particular sample. But it is a clear enough difference to make the Eredivisie worth following as more matches are added.

Nine of its 18 teams fall into the high-Load-concentration, high-final-third-share cluster. That group includes PSV Eindhoven, ADO Den Haag, Telstar, Excelsior, Feyenoord and FC Utrecht. The point is not that these teams necessarily play the same way. They share a similar relationship between the two indicators in the current dataset.

The Dutch chart also shows teams in other profiles. Ajax appears in the high-concentration, lower-final-third-share group, while AZ Alkmaar sits in the lower-concentration, average-share cluster. Those placements are a reminder that the league average does not describe every club. Even in the competition with the highest mean final-third share, teams occupy different areas of the profile map.

The Eredivisie result gives us a useful next question: will this higher average remain visible once each team has played more matches? If it does, we can then investigate what is behind it. Are more passes reaching the final third because teams progress through possession, because they play forward earlier, or because of the way opponents defend? Those explanations require more than the pass-share measure, but the league-level signal gives us a clear place to begin.

La Liga leans toward the middle

La Liga has a different shape. Ten of its 20 teams fall into the average-concentration, average-final-third-share cluster, making it the most common profile in the Spanish sample.

That group includes clubs such as Barcelona, Real Madrid, Atlético Madrid, Real Sociedad, Real Betis, Espanyol and Athletic Club. Their presence in the same cluster does not mean they have identical tactics or that their passing networks are interchangeable. It means that, across these two measures, their current values are close enough for KMeans to group them together.

Other La Liga teams appear in profiles with lower final-third shares or lower Load concentration. Celta Vigo, Valencia and Deportivo Alavés are among those placed in the average-concentration, lower-share group, while Getafe and Osasuna appear in the lower-concentration, average-share cluster.

This spread helps show why a cluster is not a league ranking. A competition may contain a large central group and still have meaningful variation outside it. The chart lets us see both the common pattern and the exceptions without forcing every team into one broad league-wide description.

Different pictures in England, Portugal and Germany

The Premier League is spread across several clusters rather than being dominated by one profile. Four teams fall into the high-concentration, high-share group, while four are in the high-concentration, lower-share group. Five appear in the lower-concentration, average-share cluster, and another five in the average-concentration, lower-share group.

Among the names in the high-concentration, high-share group are Arsenal, Nottingham Forest, Coventry and Sunderland. Meanwhile, Manchester City, Brighton, Fulham and Aston Villa appear in the high-concentration, lower-share cluster. This may sound counterintuitive if we treat the labels as definitive statements about a club’s attacking style. But the labels are relative to these data and these two metrics. They do not describe every aspect of a team’s football, and they do not replace a match-by-match tactical analysis.

The Primeira Liga also shows a split across multiple profiles. Five teams are placed in the high-concentration, lower-share cluster, including FC Porto, Braga, Famalicão and Santa Clara. Three are in the average-concentration, lower-share group. At the same time, Moreirense appears in the high-concentration, high-share cluster with a final-third pass share of approximately 34.6%. Rio Ave is in that group as well, at around 33.6%.

That contrast is useful because it shows how clubs in the same league can share one side of a profile and differ on another. Porto and Moreirense both fall into higher Load concentration, but their final-third pass shares place them in different groups. The two metrics capture different parts of the passing picture.

In the Bundesliga, five teams appear in the lower-concentration, average-share group, while four fall into the high-concentration, high-share profile. Bayer Leverkusen and RB Leipzig, by contrast, sit in the high-concentration, lower-share cluster in the current sample. Again, that is a description of their measured position on the chart, not a complete account of how they attack or how successful they are.

Strasbourg and the small two-team cluster

The most striking outlier in final-third pass share is Strasbourg, at approximately 44.2%. That is the highest value among the 150 team observations. KMeans groups Strasbourg with Académico de Viseu in the lower-concentration, higher-share profile.

The combination makes the cluster worth mentioning, but its size makes it difficult to generalise. With only two teams, the group is better treated as a rare profile in this dataset than as a stable category with a well-established pattern. One unusual season start or one change in the number and type of matches available could move its shape considerably.

Strasbourg’s value also shows why looking at the chart matters. A team can reach the final third with a high share of its passes while having a lower Load concentration than many teams in the sample. That is a different combination from a side that pairs high final-third share with high concentration. Both may reach advanced territory frequently, but the distribution of network responsibility is not the same.

The next step would be to investigate how Strasbourg and Académico Viseu reach those areas. The current model tells us where their metrics place them; it does not identify whether the passes are central or wide, whether they come from sustained possession, or whether they follow direct transitions.

Connecting passing profiles to match results

The matchup charts compare the performance of teams with one passing profile against opponents assigned to another. The results cover 438 matches, represented as 876 team observations because each match contributes one observation for each team.

Subscribe now

The largest pairing in the chart is a meeting between two teams in the average-concentration, average-share profile. It contains 90 team observations, equivalent to 45 matches. Teams in that profile recorded a 45.6% win rate against opponents with the same profile and averaged 1.76 goals scored per match.

The high-concentration, lower-share profile has several matchup groups with substantial counts. Against the average-concentration, average-share profile, it has 46 team observations and a 43.5% win rate. Against opponents in the high-concentration, high-share profile, it has 29 observations, a 51.7% win rate, an average of about 1.83 goals scored per match, and a mean goal difference of approximately +0.90.

That result stands out in the chart, but it should be read as a pattern in the current sample, not as evidence that the profile causes teams to win. The sample is still modest, and the profile pairings do not control for team quality, home advantage, injuries, match state, finishing or the strength of the opposition.

The high-concentration, high-share profile also produced some notable figures. Against the average-concentration, average-share profile, it averaged 1.60 goals per match across 45 team observations. Against teams in its own profile, it averaged about 1.88 goals across 42 observations. Against the high-concentration, lower-share profile, however, it averaged less than one goal per match across 29 observations, with a negative average goal difference.

These contrasting results show why the matchup analysis can be more revealing than a simple overall profile ranking. The same profile may appear more productive against one kind of opponent and less productive against another. That possibility is worth investigating, but the present study cannot tell us whether the difference comes from a tactical interaction or from other factors in the matches.

The win-rate chart counts a draw as a match that did not produce a win for the team profile being shown. That is important when reading the percentages. A 45.6% win rate does not mean that the other 54.4% of matches were all defeats; some were draws.

What the matchup chart cannot prove

A profile-versus-profile chart can suggest that some combinations deserve a closer look. It cannot establish that one passing style caused a particular result.

A team with a high final-third pass share might score more because it advances the ball effectively. It might also score more because it has better finishers, faced weaker opponents during the sample, or converted a small number of chances unusually well. The same caution applies to goal difference. A strong figure may reflect a real matchup pattern, or it may change quickly as new results are added.

The smallest matchup groups make this especially clear. The lower-concentration, higher-share profile includes only two teams and appears in just a handful of team observations. Some cells in the chart have as few as one to three observations. A 100% win rate in such a cell is not evidence of a dependable advantage. It means that the limited number of observations all happened to be wins.

Even some of the larger cells remain early signals. A group with 29 or 45 team observations gives us more to work with than a group with two or three, but it is still not enough to make a broad causal claim. It is enough to identify an interesting relationship and ask whether it holds up across future matches.

The best use of these results is as a guide for the next round of analysis. Which specific fixtures sit behind the positive goal difference for the high-concentration, lower-share profile against the high-concentration, high-share group? Did those teams progress through the same areas? Did the losing sides attempt more passes into the final third but create fewer chances? Did the profile pairing behave differently depending on which team played at home?

Those questions turn an initial chart into a research path. They also help prevent the cluster labels from doing too much explanatory work. A label can organise the teams. It cannot replace looking at what happened in the matches.

Why the early-season sample matters

The analysis includes 150 team profiles across eight leagues, but it was created near the beginning of the season. Many teams have played only a small number of matches. A team’s early values can be influenced by an unusual fixture list, a difficult opponent, a red card, an early lead that changes its approach, or a short run of especially effective or ineffective passing.

That makes the current cluster assignment a snapshot. It may remain stable as more matches arrive, or it may change. A team that begins with a high final-third pass share could settle closer to the average once its sample grows. A side that currently appears in a lower-share profile could move upward after a run of matches against opponents that allow more possession in advanced areas.

The league comparisons need the same caution. The Eredivisie’s higher average is a meaningful feature of this sample, but it is not yet a fixed statement about the whole season or a permanent description of Dutch football. The difference should be monitored as the dataset expands.

There is also a practical distinction between team observations and match observations. The 150 team profiles describe the clubs on the profile chart. The matchup figure uses 438 matches and represents each match from both teams’ perspectives, producing 876 team observations. That is why the largest matchup cell can contain 90 observations even though the underlying number of fixtures is 45.

As the season progresses, this structure will let us examine whether the initial patterns persist. We can see whether teams remain in their clusters, whether certain profile pairs recur, and whether the early differences in goals and results become clearer or disappear.

What I want to test next

The current analysis is deliberately narrow: two passing indicators, six KMeans profiles and the results from the available matches. That makes the first comparison readable, but it also leaves several important parts of the game outside the model.

A natural next step is to add measures that describe what happens after the ball reaches the final third. Do teams in the high-share profiles take more shots? Do they create better chances? Are they progressing through central combinations, through the wide areas, or with longer passes? Does a concentrated network rely on one player to enter the final third, or does that player mainly circulate the ball before someone else advances it?

Another step is to follow profiles over time. Rather than assigning one label to a team and treating it as permanent, we can ask how often a team’s match-by-match behaviour resembles each cluster. That could help separate a stable passing tendency from a profile that changes with the opponent or the context of a particular game.

The matchup results also invite a closer look at specific profile pairs. The positive goal difference for high-concentration, lower-share teams against high-concentration, higher-share opponents is one example. The next question is not whether the chart proves a tactical advantage; it does not. The more useful question is whether video and additional match data reveal a recurring explanation in those fixtures.

That is where the team labels become practical. They reduce a large collection of fixtures into groups that can be compared, helping us decide which matches to inspect and what patterns to look for. The numbers can point us toward a question. The deeper analysis still has to establish what happened on the pitch.

A first map to build on

KMeans has grouped 150 teams into six passing profiles based on Load Gini and final-third pass share. The early results show an Eredivisie sample with a higher average final-third pass share than the other leagues, a strong concentration of La Liga teams in the average-average profile, and considerable variation across the Premier League, Primeira Liga and Bundesliga.

The matchups add another layer. Some profile pairings are associated with higher goal totals or stronger goal differences, while others show less productive outcomes in this sample. Those are signals to follow, not proof that passing profiles determine results. The early-season dataset is small, the matchup counts are uneven, and several striking percentages come from very few observations.

That is exactly why the next stage matters. As the sample grows, we can see whether these clusters remain stable, whether the league differences persist and whether the matchup patterns hold up. For now, the analysis gives us a clear framework for asking a more precise question about European football: how does the distribution of passing responsibility relate to a team’s ability to reach the final third, and what happens when different passing profiles meet?

Download the Spreadsheets

Paid subscribers can download the spreadsheets behind this analysis below. The first file contains each team’s league, Load Gini, final-third pass share and KMeans profile. The second contains the analysed matches, including the teams, their assigned profiles, the score and the result.

Read more

Whatever ultimately happens with the accusations against Manchester City, and whatever decisions football authorities may eventually make about titles from that period, there is one thing no document can change: what actually happened on the pitch.

Manchester City played those seasons. They won the matches. Their players celebrated. Their supporters celebrated. Their opponents tried to beat them — and failed often enough for City to finish as champions.

That happened.

You can change a record book. You can put an asterisk next to a season. You can issue a statement, sign a document, remove a trophy from an official list or declare that history should now be interpreted differently.

But you cannot travel backwards in time and change what millions of people actually experienced.

And that distinction matters.

Football does not belong primarily to administrators, lawyers, committees or governing bodies. Football belongs to the people who actually make football exist: the players who play it, the coaches who build teams, and especially the supporters who invest their time, money and emotions in it.

A bureaucratic decision cannot retroactively erase the moment a supporter watched his team score the goal that effectively won a championship.

It cannot erase the celebration in the stadium.

It cannot erase the frustration of the opposing supporters.

It cannot erase the tactical battles, the mistakes, the brilliant performances or the tension of a title race.

Those things happened in the real world.

That is football history.

This is why, even if Manchester City were eventually stripped of titles, my personal view of those seasons would not change. On the pitch, Manchester City finished as champions. I watched it happen. Everyone else watching football at the time watched it happen too.

The official record might someday say something different.

Memory will not.

And perhaps football supporters should remember that distinction more often.

The emotions we experience through football should never depend entirely on what someone sitting behind a desk decides years later. Administrators do not own those memories. Institutions do not own them either.

The supporter does.

A bureaucrat was not sitting beside you when your team conceded in the 90th minute.

He did not spend the week thinking about the next match.

He did not buy your ticket.

Subscribe now

He did not feel the stadium explode after a decisive goal.

He did not go home furious after a defeat or spend the following morning replaying a victory in his head.

So why should a signature on a piece of paper suddenly become more important than everything that actually happened?

Football is alive precisely because people care about it far beyond regulations, balance sheets and administrative decisions.

Rules and financial regulations have their place, of course, and the allegations against Manchester City should be dealt with through the applicable process. But that process and the lived history of football are not exactly the same thing.

One belongs to institutions.

The other belongs to us.

And if one day somebody officially writes that Manchester City were not champions of a season in which millions of us watched them become champions, they are free to write it.

The paper may change.

What happened on the pitch will not.

Data disclaimer: No match images or video footage were used in this analysis. The conclusions are based exclusively on event-level match data and derived tactical information, including xT progression, passing networks, player centralities, defensive influence and event-derived positional tendencies. The movement analysis therefore represents functional patterns inferred from where players passed and interacted with the ball, not tracking-data running paths.

Barcelona won 3–1 away to Sevilla in La Liga 2026/27, with Raphinha scoring all three Barcelona goals after Sevilla had taken the lead. Yet the passing data tells a much less conventional story: Sevilla generated considerably more positive threat from its passing sequences, while Barcelona combined a much denser circulation network with extraordinary efficiency at the decisive end of attacks.

Sevilla 1–3 Barcelona analysis inside the Football Hacking Web App

Sevilla’s opening structure created real danger with relatively few passes

The opening phase immediately showed a contrast in philosophy. Barcelona established the more elaborate circulation structure, but Sevilla did not need long possession sequences to move the ball into threatening areas; Gabriel Suazo provided an aggressive left-sided outlet, while Juan Iglesias helped maintain width and progression on the opposite side rather than allowing Sevilla to become compressed into one corridor.

Sevilla’s network was relatively sparse across the full match, with an edge density of just 0.413 compared with Barcelona’s 0.525. That did not make it ineffective, however, because Sevilla’s passing sequences generated significantly more threat per action and were often designed to reach the next dangerous zone rather than simply preserve possession.

The home side’s early goal captured that distinction. Youssouf Fofana attacked the centre of the six-yard area and headed Sevilla ahead in the 18th minute, turning a relatively direct progression into a high-value finishing situation rather than requiring a long sequence around Barcelona’s defensive block.

Barcelona’s response came from movement around the structure, not from its main hub

Barcelona responded only three minutes later, and the scorer’s network profile is one of the most interesting features of the entire match. Raphinha equalised from the left side of the penalty area in the 21st minute, but he would finish the match with zero Load Centrality and a relatively modest PageRank of 0.035.

That apparent contradiction is actually tactically useful. Raphinha was not being used as the player through whom Barcelona’s possession repeatedly travelled; instead, his value came increasingly from arriving at the end of sequences, moving into finishing zones and exploiting the structure created by teammates rather than becoming the structure himself.

In the opening phase, Lamine Yamal maintained a high position on the right while Barcelona also kept an advanced reference on the opposite side. The resulting horizontal stretch created space between Sevilla’s defensive references, while Raphinha could occupy more central positions without being required to remain permanently wide.

Subscribe now

22.5’–HT: Barcelona rotate the midfield, but Sevilla still create more passing value

After the equaliser, Barcelona changed the height of several midfield roles. Fermín López began operating higher and more aggressively between midfield and the attacking line, while Pedri increasingly appeared underneath that level to support circulation and reconnect the first phase with the attacking structure.

This distinction becomes important when interpreting Barcelona’s possession volume. Their central and left-sided build-up could generate multiple routes toward the final third, but many of those sequences included backward or lateral actions after reaching more threatening areas, reducing the net xT value even when the team remained in control of the ball.

Sevilla were much more economical. Their full-match passing profile produced approximately 0.885 xT per 100 passes, with a positive net passing progression of +1.07 and +0.73 net xT generated in the final third despite recording only 197 analysed passes.

Barcelona, by contrast, recorded 549 analysed passes but only 0.240 xT per 100 passes, while their overall net passing progression finished at -0.88. Their final-third passing alone produced -0.84 net xT, meaning that many possessions moved from a more threatening location into a less threatening one even though Barcelona continued circulating the ball.

Negative passing xT does not mean Barcelona were harmless

That distinction needs to be understood correctly. Passing xT measures how possession changes the expected threat of the next state; it does not measure finishing quality, shot conversion or the value of the shot itself, so Barcelona could produce a negative net passing profile and still score three times from decisive attacking actions.

The game is therefore a strong example of why possession quality cannot be reduced to one metric. Barcelona’s circulation was less efficient at continuously increasing territorial threat, but when the structure finally delivered the ball into the right moment, their finishing and final-action execution were decisive.

Sevilla had the inverse problem. Their passing network produced threat at a much higher rate and lost very little of it through destructive backward progression, yet the advantage did not translate into enough goals after the opening finish.

Pau Cubarsí was Barcelona’s structural hub

The centrality data shows exactly where Barcelona’s possession was anchored. Pau Cubarsí recorded the highest Load Centrality on the team at 0.362 and also the highest PageRank at 0.150, while his Closeness Centrality of 4.13 placed him at the heart of Barcelona’s ability to keep the network connected.

This is a very different role from Raphinha’s. Cubarsí was responsible for carrying a large share of important network routes and repeatedly became an influential receiving destination, whereas Raphinha largely existed downstream of those connections, closer to the point where structure became finishing.

João Cancelo was another major connector, posting a Load Centrality of 0.271, while maintaining an important left-sided role throughout the match. His influence also changed spatially: he could support deeper circulation but remained capable of advancing outside the midfield line, especially when Barcelona needed width around Sevilla’s compact central defence.

Barcelona’s network was more connected — but not necessarily more progressive

Barcelona’s pass network contained 126 connections and a reciprocity rate of 0.841, compared with Sevilla’s 99 connections and 0.646 reciprocity. In practical terms, Barcelona had more established relationships and more two-way passing routes between players, which is exactly what we would expect from a team able to maintain long periods of structured possession.

The problem is that network richness and progression are different things. Barcelona’s density helped them retain and recycle the ball, but Sevilla’s less complex structure produced much greater xT efficiency because its connections were more frequently aimed at escaping a line or arriving directly in an advanced zone.

This is also reflected in the corridor data. Barcelona’s central starts were divided between central final-third entries and the right half-space, while left-sided possessions could finish through several different channels; the structure was flexible, but that flexibility frequently required additional circulation before the ball became dangerous.

Sevilla’s most important network relationship also produced attacking value

Sevilla’s centralities were distributed differently. Gabriel Suazo posted the highest Load Centrality at 0.357, making him the main carrier of routes through the network, while other midfield connections provided the receiving influence that helped Sevilla move from the first line toward attacking areas.

The relationship between Félix Correia and Youssouf Fofana was particularly productive. Their passing connection generated the largest positive progression among Sevilla’s leading links, and the same broad relationship appeared in the sequence that produced Sevilla’s opening goal; this was a network with fewer interactions but a much clearer tendency to convert specific combinations into danger.

That distinction helps explain why Sevilla’s overall Network Health was higher despite losing 3–1. Their structure, efficiency and defensive influence combined for a general score of roughly 0.765, compared with Barcelona’s 0.607, reinforcing that the scoreboard and the quality of the underlying passing system can tell very different stories.

45’–67.5’: Lamine Yamal stretches the right and Raphinha attacks the box

Barcelona’s second-half movement made their attacking roles clearer. Lamine Yamal continued to operate as the high right-sided reference, while Raphinha increasingly attacked central finishing spaces rather than dropping into circulation, separating the role of chance creator from the role of finisher.

The second goal arrived in the 56th minute and was the clearest expression of that relationship. Yamal created the opportunity from the right side, and Raphinha attacked the central zone to score with a first-touch header, giving Barcelona a 2–1 lead despite the underlying passing xT still favouring Sevilla.

Barcelona then changed their left-side structure around the hour mark. Gerard Martín replaced Andreas Christensen, giving the team a deeper left-sided support position while allowing the player ahead of him to maintain greater width and height instead of constantly dropping to stabilise possession.

Sevilla’s defensive work repeatedly forced Barcelona to reset

Sevilla’s strong passing numbers should not obscure their defensive contribution. Their defensive influence index reached 0.769, with the central defensive third producing the strongest concentration of interventions, and several of those defensive zones forced Barcelona into recirculation instead of immediate progression.

That helps explain Barcelona’s negative final-third xT. Sevilla were not necessarily preventing Barcelona from entering advanced areas, but they were often preventing the next action from becoming more dangerous; Barcelona would reach a threatening zone, encounter resistance, and then move the ball backward or laterally to restart the attack.

Barcelona’s own pressure operated much higher. Their most influential defensive actions were concentrated in central areas of Sevilla’s attacking build-up, and those interventions frequently caused Sevilla to lose rhythm before establishing longer possession sequences.

The consequence was a fascinating tactical trade-off. Sevilla produced more threat when they escaped Barcelona’s pressure, while Barcelona prevented them from escaping often enough to dominate possession and control the number of attacking sequences.

Raphinha’s third goal completes the separation between network role and finishing role

Barcelona’s third goal arrived in the 73rd minute, and once again Lamine Yamal was involved in creating the decisive situation. Raphinha received centrally and finished one-on-one with his left foot, completing his hat-trick from another action in which his contribution came at the terminal point of the attack rather than through extensive involvement in the pass network.

That is perhaps the strongest individual tactical takeaway from Sevilla 1–3 Barcelona. A player can have almost no Load Centrality and still decide the match because centrality measures structural importance inside circulation, not finishing importance; Raphinha did not need to organise Barcelona’s possession because his job was to exploit what that possession eventually created.

After the third goal, Barcelona could reduce some of the attacking risk in their positional structure. Raphinha was later replaced by Karim Adeyemi, while Barcelona continued to maintain a wide outlet and sufficient central occupation to prevent Sevilla from committing completely forward without exposing space behind their defensive line.

Why the final score and the passing model tell different stories

If this game were analysed only through the score, the conclusion would be straightforward: Barcelona won comfortably and Raphinha scored a hat-trick. If it were analysed only through passing xT, the interpretation could move too far in the opposite direction, because Sevilla generated dramatically more progression value per pass and a much healthier final-third profile.

Both views are incomplete on their own. Sevilla were more efficient at using passes to increase threat, but Barcelona were more connected, sustained possession for longer, suppressed many Sevilla sequences before they developed, and converted their decisive chances with exceptional efficiency.

That combination is why this match is so useful analytically. Barcelona did not need the stronger passing-threat profile to win because their network successfully delivered enough high-quality terminal actions, while Sevilla demonstrated that producing better progression numbers is not the same as converting those advantages into goals.

Final takeaway: structure created the platform, Raphinha decided the outcome

Barcelona’s 3–1 win over Sevilla was a match in which the passing network and the scoreboard pulled in different directions. Cubarsí was the structural hub, Cancelo helped connect and stretch the left side, and Barcelona maintained the denser and more reciprocal network; nevertheless, that circulation often reduced xT once possession reached advanced areas.

Sevilla were almost the opposite. They completed far fewer analysed passes but generated substantially more positive progression from them, used their wide-to-central relationships efficiently and produced the healthier overall passing structure, yet they were unable to turn that advantage into enough finishing output after taking the lead.

The decisive layer therefore came from player roles. Raphinha was not a central connector and did not need to be one: he occupied the final stage of Barcelona’s attacks, scored three times and demonstrated why network centrality and match-winning attacking influence are related concepts, but certainly not identical ones.

Download the data

Paid subscribers can download the complete Sevilla vs Barcelona match-event XLSX below the paywall. The spreadsheet contains the event-level data from the match and can be used for independent tactical analysis, visualisations, modelling and further football-data research.

Read more

Data disclaimer: No match images or video footage were used in this analysis. The conclusions are based exclusively on event-level match data and derived tactical information including pass-network structure, xT progression, player centralities, defensive influence and event-derived positional tendencies. These movement patterns come from event data, not tracking data, so they describe functional changes in where players passed, received and connected rather than continuous physical trajectories.

Atletico Madrid won 2–1 in La Liga 2026/27 after a scoreless first half, with all three goals arriving after the break. The result itself was important, but the deeper story lies in the contrast between two passing structures: Atletico generated considerably more positive threat from their possession, while Real Madrid’s network relied heavily on deeper structural players and became even more conservative after going down to ten men.

Atletico Madrid 2–1 Real Madrid analysis inside the Football Hacking Web App

The opening phase: Atletico stretch the right, Real Madrid load the left

Atletico’s opening structure already contained the asymmetry that would remain important throughout the match. Giuliano Simeone operated as an advanced reference in the right corridor, while Lee Kang-In repeatedly occupied the inside-right channel behind him; the relationship gave Atletico both external width and an interior passing option without asking the same player to perform both functions. Early event locations show Atletico repeatedly circulating and recovering the ball on that right side, rather than treating the flank as a simple crossing zone.

Real Madrid’s occupation leaned in the opposite direction. Marc Cucurella gave the team a wide left outlet from a deeper base, while the attacking structure repeatedly placed advanced players ahead of him in the same broad side of the pitch, creating a clear left-sided emphasis rather than a perfectly symmetrical build-up.

That mattered because Atletico did not need to dominate every zone equally. They could preserve their own right-sided combinations while forcing Real to circulate around a more predictable left-heavy structure, and the first half gradually became a contest over where possession could be converted into genuine progression rather than simply who completed more passes.

22.5’–HT: Atletico begin moving inside rather than simply outside

The second part of the first half produced a subtle positional change. Alejandro Grimaldo, initially much closer to the left touchline in the event-derived positioning, began appearing more often inside the left half-space, while Álex Baena also shifted toward more central areas rather than remaining permanently attached to one external corridor.

That inward movement created a different relationship with Atletico’s aggressive right side. Instead of building two independent flanks, Atletico could use one side to attract pressure and then access the opposite interior or corridor, an idea that is also visible in the progression routes: when Atletico began sequences centrally, 66.7% of those sequences reached the final third through the left corridor, while right-corridor starts also reached the opposite left corridor in 66.7% of the recorded sequences.

Real Madrid’s own interior structure dropped somewhat deeper as the first half developed. Jude Bellingham became more involved from central-left positions rather than only operating close to the last line, while Arda Güler increasingly participated from the right half-space at a lower height.

The result was a first half in which Real retained routes through midfield but did not establish the same relationship between circulation and threat. The difference would become much clearer after the interval, when the match suddenly shifted from positional competition to adaptation under numerical inequality.

Subscribe now

The decisive sequence: red card, penalty and a completely different match

The central turning point arrived shortly after half-time. Dean Huijsen committed the penalty foul, was subsequently shown a red card, and Alejandro Grimaldo converted the resulting penalty to put Atletico ahead; the sequence changed both the score and the structural conditions of the match within only a few minutes.

The red card is particularly important for interpreting everything that followed. Any second-half network analysis that ignores the numerical disadvantage risks treating Real Madrid’s deeper structure as an ordinary tactical preference, when in reality the team now had to protect larger spaces with fewer players while simultaneously chasing the game.

Atletico reacted by increasing occupation around Real’s reduced block rather than simply circulating horizontally. Koke, introduced at half-time, immediately became an important connector in the right-central area, while Marcos Llorente continued linking the right corridor and right half-space, giving Atletico two different ways to access the same side without occupying identical spaces.

45’–67.5’: width outside, control inside

The most productive phase of Atletico’s attacking structure came after the interval. Grimaldo again pushed much higher in the left corridor, while Giuliano Simeone maintained aggressive width on the right, forcing ten-man Real Madrid to defend the full horizontal dimension of the pitch rather than collapse exclusively around the penalty area.

That outside occupation created room for the interior circulation. Atletico’s underlying passing model finished with 551 analysed passes, +1.31 total net xT progression and 0.71 xT per 100 passes, while their attacking-third phase alone produced more than +1.02 net xT. The important point is not merely that Atletico passed more, but that those passes collectively moved possession toward more threatening zones rather than repeatedly destroying value.

The second goal illustrated the value of that width. Giuliano Simeone delivered the decisive cross from the right side and Jonathan David finished first time from the centre of the box, doubling Atletico’s lead only a few minutes after the penalty and converting the right-corridor occupation into a direct central chance.

For Real Madrid, the positional response was almost the opposite. Cucurella dropped into a substantially deeper left-sided position after the red card, while Bellingham also became more involved from deeper central-left zones, reflecting the need to reconnect a team whose first line had lost one of its major structural players.

Centrality explains why Atletico’s possession travelled better

Atletico’s centrality profile gives the clearest statistical explanation for their structure. Marcos Llorente recorded the team’s highest Load Centrality at 0.281 and highest PageRank at 0.131, meaning he was both heavily involved in important routes through the network and repeatedly selected as an influential destination.

Koke provided a complementary profile despite playing only the second half. His Load Centrality reached 0.248 and his PageRank 0.112, an unusually strong influence considering his reduced minutes, which fits the immediate change in Atletico’s circulation after the break: the ball had a stable interior connector without removing Llorente’s right-sided influence.

This balance is important because centrality is not simply a ranking of who touched the ball most. Load identifies players who carry a larger share of meaningful network routes, while PageRank highlights players who become influential receiving targets; Atletico’s main hubs were therefore situated in areas capable of connecting build-up, half-spaces and advanced corridors.

Real Madrid’s network was concentrated much deeper

Real Madrid’s highest Load Centrality belonged to Marc Cucurella at 0.396, while Dean Huijsen reached 0.341 before his dismissal. Cucurella’s PageRank was also high at 0.120, showing how heavily Real relied on the left-sided defensive structure to maintain connectivity rather than establishing the same degree of central influence higher up the pitch.

The next structural reference was Ibrahima Konaté, whose PageRank of 0.129 was actually Real Madrid’s highest and whose Load Centrality reached 0.297. That gives the network a very different character from Atletico’s: Real’s major hubs were concentrated in deeper defensive positions, so much of the team’s circulation had to travel through the back line before it could reach attacking players.

Real were also more dependent on a narrow set of Load hubs. Their top three players accounted for roughly 77% of total Load Centrality, compared with about 59% for Atletico, indicating a network in which progression responsibility was much more concentrated.

Similar network density, very different network value

One of the most interesting findings is that the difference cannot be explained simply by network density. Atletico’s passing graph had an edge density of 0.604, while Real Madrid’s was 0.600; structurally, both teams maintained a broadly connected network.

The bigger separation came from reciprocity and progression. Atletico’s reciprocity reached 0.841 compared with Real’s 0.730, indicating more consistent two-way relationships between connected players, but the much more important distinction was what happened to possession after those connections were established.

Atletico generated 0.71 xT per 100 passes, while Real Madrid finished at only 0.44. Real’s overall net passing progression remained slightly positive, but their final-third passing finished at -0.080 net xT, whereas Atletico generated +1.029 in the same broad phase of possession.

That contrast is huge tactically. Atletico’s passes near goal tended to preserve or increase attacking value; Real Madrid could reach advanced zones, but the next action often moved the ball toward a less threatening state, a classic example of why final-third possession should never automatically be equated with final-third danger.

Real Madrid’s corridor progression became fragmented

Real Madrid still found some progression from the left corridor. Around 42.9% of their sequences beginning there also reached the final third through the same corridor, while another 28.6% moved from the left toward the central corridor, showing that the left side remained their clearest launching platform.

But the distribution elsewhere was fragmented. Right-corridor starts were split equally between central, left-half-space and right-half-space destinations, while left-half-space starts were divided across four different final-third zones; variability can be positive, but in this match it did not correspond to sustained xT creation.

Atletico’s progression had a clearer relationship between origin and target. Right-half-space possessions most frequently reached the right corridor, while central possessions strongly favoured an exit toward the left, allowing the team to alternate between same-side continuation and diagonal transfer rather than circulating without a clear destination.

Atletico’s defensive influence also mattered

The attacking numbers tell only part of the story. Atletico recorded a defensive influence score of 0.717, compared with Real Madrid’s 0.622, and their most influential defensive zones included the central defensive third and the left half-space.

When Atletico defended those central and left-sided areas, Real Madrid frequently lost rhythm or were forced into recirculation rather than progressing immediately. This is especially relevant because those were precisely the zones from which Real were trying to reconnect their left-heavy structure with central attacking players.

Real Madrid also created disruption with their own higher pressure, especially through central areas in Atletico’s half, but Atletico were generally better at surviving those moments and restarting possession. The defensive data describes several Real pressure zones that caused recirculation, yet Atletico’s superior final-third progression suggests those disruptions rarely became sustained territorial control.

67.5’+: Atletico protect the centre and preserve outlets

The final phase produced another positional change from Atletico. Marcos Llorente dropped into a noticeably deeper position on the right side, while Ademola Lookman became a much higher outlet on the opposite left corridor, giving Atletico a way to protect the lead without removing all depth from their counterattacking structure.

Real Madrid responded by pushing new players into advanced zones. Yan Diomande occupied a high left-sided position, while Bernardo Silva increasingly connected from a more central-right area, giving Real a fresher route around Atletico’s compact block rather than simply repeating the first-half structure.

Real eventually reduced the deficit from a set piece. Bernardo Silva delivered the free-kick cross and Antonio Rüdiger headed home from the centre of the box late in the match, but the goal arrived too late to overturn Atletico’s structural advantage.

Final takeaway: Atletico won the battle between circulation and threat

Atletico Madrid 2–1 Real Madrid was not simply a story about the red card, even though Huijsen’s dismissal was the event that radically changed the second half. Atletico had already built a more reciprocal network and a clearer relationship between central circulation, right-sided occupation and opposite-side progression; the numerical advantage then amplified structures that were already functioning.

The centralities are especially revealing. Atletico’s main influence ran through Llorente and Koke in midfield and the half-spaces, while Real Madrid’s strongest network influence was concentrated around Cucurella, Huijsen and Konaté deeper in the structure; one team’s hubs were positioned to connect possession with attacking zones, while the other relied much more heavily on players responsible for getting the ball out of the first line.

The xT numbers complete the picture. Atletico generated significantly more threat per 100 passes, produced strongly positive value in the final third and maintained useful width throughout the second half, while Real’s final-third circulation actually lost net xT despite their late push.

For tactical analysis, that is the main lesson from this Madrid derby: connectivity alone does not define control. The decisive question is where the network’s most influential players operate, how the structure changes across match phases, and whether possession increases or decreases the threat of the next action.

Download the data

Paid subscribers can download the complete Atletico Madrid vs Real Madrid match-event XLSX below the paywall for their own tactical analysis, visualisations and football-data projects.

Read more

Corinthians vs Fluminense brings together two efficiency-first teams with the same dominant bottleneck in Progression Space, but very different ways of dealing with resistance. Corinthians arrive with the stronger defensive influence and cleaner relationship between route frequency and value, while Fluminense can reach dangerous zones but are less consistent at preserving that value once pressure arrives.

Below the paywall, paid subscribers get the complete Football Hacking structural matchup graphic, a deeper PDF report designed for coaching-staff-level analysis, and the XLSX Poisson probability matrix generated from the Football Hacking models. Together, they combine tactical structure, progression and resistance analysis with probabilistic tools for subscribers who also want to build their own betting assessment.

Become a paid Football Hacking subscriber to unlock the full analysis.

Subscribe now

Explore the Football Hacking Web App:

Read more

Data disclaimer: No match images or video footage were used in this analysis. The conclusions are based exclusively on event-level match data and derived tactical information such as passing structure, xT progression, player centralities, defensive influence and functional movement patterns; the movement analysis therefore describes event-derived positional tendencies, not tracking-data running paths.

Liverpool beat Tottenham 3–1 in the 2026/27 League Cup, but the most interesting feature of the match was not simply the difference on the scoreboard. Tottenham built the denser passing structure and circulated the ball through more connections, yet Liverpool produced the cleaner progression, the stronger final-third value and the more influential defensive network.

Liverpool 3–1 Tottenham analysis inside the Football Hacking Web App

Liverpool’s first advantage came from structure, not possession volume

The opening phase already showed an important contrast. Liverpool’s central structure was built around Trey Nyoni, whose median passing position sat in the central corridor, while Jeremie Frimpong provided a much higher and wider reference on the right corridor; instead of filling the same vertical line, they separated the functions of central circulation and width.

That gave Liverpool an interesting asymmetry. Nyoni could participate in circulation without being permanently fixed next to the centre-backs, while Frimpong’s high starting point stretched Tottenham horizontally and created a direct route into advanced areas whenever the ball escaped the first pressure.

Tottenham’s early structure was more possession-oriented. Marcos Senesi operated as a deep left-sided distributor, while Mohammed Kudus occupied a much more advanced position in the right corridor, creating a clear long diagonal between the deepest circulation point and one of the highest attacking references.

That structure helped Tottenham keep connections alive, but Liverpool found the first decisive action. Alexis Mac Allister scored in the 21st minute from a central position inside the box, giving Liverpool a 1–0 lead before the first half had reached its midpoint.

Subscribe now

22.5’–HT: Mac Allister moves higher and Liverpool change the midfield geometry

After the opening goal, Liverpool’s event-derived positioning changed. Alexis Mac Allister moved from a deeper left-central passing position into a higher left half-space, while Trey Nyoni appeared noticeably deeper and more right-sided in the circulation structure.

That exchange matters because it changed who was responsible for receiving behind Tottenham’s first pressure and who was responsible for connecting the attack. Mac Allister became more available ahead of the ball, while Nyoni increasingly supported the structure underneath him rather than duplicating the Argentine’s zone.

Tottenham also adjusted their occupation. Rodrigo Bentancur became much more central and much more involved in circulation, while Mohammed Kudus remained the higher right-sided reference, giving Spurs a stable central-to-wide route without forcing Kudus to drop continuously toward the ball.

Yet there was already a warning in Tottenham’s possession. Their circulation could move around Liverpool, but it was not consistently moving Liverpool itself; the eventual whole-match data would show that Tottenham built a connected network without converting that connectivity into enough positive passing threat.

The centrality map explains who actually controlled Liverpool’s circulation

Liverpool’s passing network had one extremely important structural hub: Ronald Araujo. He recorded the team’s highest Load Centrality at 0.362 and the highest PageRank at 0.135, which means he was simultaneously the player through whom a large share of important passing routes travelled and the most influential destination inside the network.

That combination is particularly revealing. A high Load value alone can identify responsibility for connecting the structure, while a high PageRank indicates that teammates repeatedly selected the player as an important destination; when both are high, the player is not simply circulating possession but acting as a genuine structural hub.

Alexis Mac Allister offered a different profile. His Load Centrality of 0.248 was the second-highest Liverpool value, while his PageRank of 0.105 and team-leading Closeness Centrality of 2.802 show how consistently he could connect with the rest of the network from midfield.

This fits the match itself. Mac Allister was not only present in circulation but also able to move from the deeper left-central zone into more advanced spaces, and that dual role eventually produced both a goal and the final pass for Liverpool’s second.

Tottenham had an even stronger hub — but in a less productive structure

Tottenham’s network was even more concentrated around Marcos Senesi. His Load Centrality reached 0.514 and his PageRank 0.173, both higher than any Liverpool player, making him overwhelmingly important both as a route through the network and as a receiving destination.

Behind him, Martin Dúbravka also carried unusually high network influence, with a 0.195 Load value and 0.155 PageRank. The important point is therefore not that Tottenham lacked passing structure; it is almost the opposite — too much of that structure remained concentrated around the deeper circulation platform rather than being transferred efficiently into attacking threat.

The concentration numbers reinforce that reading. Tottenham’s three most important Load Centrality players accounted for roughly 67.8% of the total, compared with approximately 55.1% for Liverpool, so Spurs were more dependent on a small number of players to carry the network from one area to another.

45’–67.5’: Liverpool push both corridors higher

Liverpool’s movement after half-time became considerably more aggressive. Milos Kerkez shifted into a very high position in the left corridor, while Jeremie Frimpong maintained the opposite width on the right, leaving Liverpool with genuine occupation of both outside lanes rather than simply attacking through one overloaded flank.

That altered the space available inside. Tottenham’s block now had to respect two high external references, which increased the value of receiving between them and opened room for Liverpool’s midfielders and attackers to arrive rather than remain permanently stationed in the last line.

The second goal followed in this phase. Cody Gakpo scored in the 53rd minute from the left side of the penalty area, with Mac Allister involved in the assisted action, giving Liverpool a 2–0 advantage shortly after the restart.

Tottenham reacted with personnel changes and a slightly different central structure. Mateus Fernandes entered and began operating in central-to-left areas, while Lucas Bergvall remained a major circulation point around midfield, but the deeper network still struggled to turn its volume into sustained positive progression.

Tottenham passed more, but Liverpool’s passes carried more threat

This is where the xT profile becomes especially revealing. Liverpool’s passing model produced an xT-per-100-passes figure of 0.737 and a positive net progression of +0.677, while the attacking-third phase generated a positive +0.296 delta xT.

Tottenham’s structure tells almost the inverse story. Their xT per 100 passes was only 0.471, overall net progression finished at -1.211, and their final-third passing produced a remarkable -1.456 net delta xT despite the much larger number of actions recorded there.

That does not mean every Tottenham pass in advanced areas was poor. It means that, collectively, the team frequently moved from more threatening situations into less threatening ones, so possession volume in the final third became very different from actual progression.

Liverpool, by contrast, recorded only 59 passes in the filtered final-third sample but finished with +0.296 net xT from them. Tottenham recorded 156 while ending deeply negative, making this match a useful example of why counting entries or passes near the box can misrepresent attacking efficiency.

Liverpool could bypass the midfield bottleneck

Both teams showed a midfield bottleneck in the progression model, but Liverpool handled it more effectively. Their defensive and attacking phase scores were both close to 1.0, while the midfield phase fell to only 0.207, suggesting that the team often progressed by finding ways to jump or escape the congested middle rather than relying on slow, continuous advancement through it.

The corridor routes help explain how. From right-sided starting positions, Liverpool repeatedly reached different final-third corridors rather than following one predictable lane, while right-half-space starts also produced entries into the opposite corridor and other interior zones.

One of the clearest patterns involved Milos Kerkez, whose right-corridor-origin sequences frequently ended in the same wide channel, and Lewis Koumas, who appeared as an important advanced destination in those progressions. Liverpool were therefore able to use width as both a direct route and a platform from which to change the final destination.

Tottenham’s progression was more central — and easier to slow

Tottenham’s main progression pattern began through the central corridor, with 55.6% of those sequences also reaching the final third centrally. Omar Marmoush and Rodrigo Bentancur were the main connection associated with that route, showing how Tottenham repeatedly tried to advance through central relationships rather than relying exclusively on wide circulation.

However, Liverpool were particularly effective at disrupting rhythm in precisely the areas where Tottenham needed clean continuation. When Liverpool defended the central zone of their defensive third, Tottenham’s response was repeatedly classified as neutral rather than progressive, while pressure in Liverpool’s left midfield area often forced Spurs into recirculation.

That relationship is visible in the defensive influence numbers. Liverpool finished with a defensive influence index of 0.787, including very high territorial pressure and suppression values, compared with Tottenham’s 0.621 defensive influence score.

67.5’+: Tottenham finally stretch Liverpool, but the game becomes more transitional

Tottenham’s late movement became wider after the score moved to 2–1. Andy Robertson advanced much further along the left corridor, while Sávio eventually provided an aggressive reference on the opposite right side, producing one of Spurs’ clearest periods of two-sided attacking occupation.

The goal that reopened the match came from a different source, however. Conor Gallagher scored with a header from close range after a corner in the 69th minute, cutting Liverpool’s lead to one and changing the risk profile of the final phase.

Liverpool’s own structure changed with their midfield substitutions. Dominik Szoboszlai occupied a more advanced right half-space, while Alexis Mac Allister continued appearing higher toward the left half-space, creating two different interior receiving lines rather than asking one midfielder to dominate both sides of the pitch.

This mattered because Liverpool no longer needed to control every possession. The priority became preserving enough structural stability to survive Tottenham’s wider attacks while retaining players capable of attacking the space that appeared when Spurs committed more bodies forward.

Szoboszlai’s late movement turns into the final goal

The late winner is a particularly clean example of movement becoming action. Szoboszlai had increasingly occupied an advanced central-to-right position during the closing phase, and in stoppage time he arrived centrally outside the box to score Liverpool’s third with his right foot.

That goal also completes the broader story of Liverpool’s midfield. Mac Allister supplied the early scoring action and later helped create the second goal, while Szoboszlai’s later introduction gave Liverpool a fresh advanced receiver who could attack the space in front of Tottenham’s retreating defensive structure.

A denser pass network did not mean a better attacking network

Tottenham actually finished with the denser passing graph: 0.546 edge density and 0.763 reciprocity, compared with Liverpool’s 0.496 and 0.706. In simple terms, Tottenham had more interconnected passing relationships and more two-way links between players, but Liverpool’s network produced superior efficiency and a higher overall health score.

This is one of the most important analytical lessons from Liverpool 3–1 Tottenham. A network can look rich because the ball circulates through many connections, but if those connections are concentrated too deeply, repeatedly return toward safe zones or lose xT as they enter the final third, network density becomes a description of circulation rather than evidence of attacking superiority.

Liverpool’s network was less dense but more useful. Araujo gave the team a stable structural hub, Mac Allister connected circulation with higher-value zones, the full-width occupation increased the space between Tottenham’s defensive references, and the second-half substitutions preserved a threat behind the ball even after Tottenham reduced the deficit.

Final takeaway: Liverpool won the value battle, not the passing-volume battle

Liverpool’s 3–1 win was built on a more productive relationship between structure, progression and movement. Tottenham could circulate through Senesi and their deeper network, but Liverpool were better at converting their passing structure into positive xT and at disrupting Spurs whenever their possession approached the areas where it needed to become genuinely dangerous.

The four temporal phases also show why average positions are not enough. Liverpool shifted from a balanced opening structure, to a more advanced Mac Allister before half-time, to aggressive two-sided width after the break, and finally to a late configuration in which Szoboszlai attacked the right half-space and central finishing zones; the team’s structure evolved with the score rather than remaining fixed.

For Tottenham, the late expansion of the left and right corridors created more territorial presence but came after the underlying passing network had already shown its main limitation. More possession routes and more final-third actions did not guarantee more threat, and Liverpool’s combination of defensive suppression and more efficient progression ultimately separated the teams.

Paid subscribers: download the match-event XLSX

Paid subscribers can download the complete Liverpool vs Tottenham match-event spreadsheet below the paywall. The XLSX contains the event-level match data for further analysis, modelling, visualisation and independent tactical research.

Read more

This tactical analysis combines Football Hacking event data with a detailed tactical review of three Como matches: Napoli vs Como, Genoa vs Como and Como vs Parma. Rather than treating them as three isolated match reports, the objective is to identify the recurring behaviours that survive different opponents, pressing structures and game states — the principles that help explain how Como play. Paid subscribers will also find, at the end of the article, an XLSX spreadsheet containing the event data from all three matches, allowing the sequences discussed here to be explored independently. The sample covers three different tactical problems and three Como victories: 2–1 against Napoli, 4–1 against Genoa and 2–1 against Parma.

Como’s main idea: manipulate the opponent before attacking the space

The most useful way to understand Como is not to begin with a formation. Their positional structure matters, but the more interesting principle is manipulation. Como repeatedly try to move the opponent before attacking the zone they actually want. Width is therefore not always the destination of possession; very often it is the bait. A pass toward the flank can attract the defensive block laterally, increase the distance between opponents and create a new central or opposite-side route. The same logic appears at different heights of the pitch: during first build-up, during progression and around the final third.

This explains why Como should not simply be described as a team that “looks for the middle”. Against Napoli, for example, they were perfectly willing to circulate or force possession toward the outside before reopening the centre. The important detail is the sequence: first alter the opponent’s position, then attack the space created by that alteration. The corridor-progression data reinforces that interpretation. Como sequences do not consistently begin and end in the same lane; possessions starting centrally, wide or in a half-space can reach the final third through another corridor altogether. Against Napoli, for instance, possessions beginning on the left reached the final third centrally or through the opposite half-space, while right-sided sequences also migrated toward central and left-sided zones.

Subscribe now

One of the clearest video examples came in Como’s first goal against Napoli. Luis Milla initiated the sequence, Lucas da Cunha then played a 42-metre chipped long pass across the pitch toward Assane Diao, and Diao immediately connected inside for Martin Baturina, who had arrived centrally to finish. The event data records the movement from Milla to da Cunha, the long diagonal from da Cunha, Diao’s assist and Baturina’s finish from a central position.
That goal contains several recurring features of Como’s positional attack: the ball changes corridor, a wide player becomes the rupture option, and another attacking player compensates by attacking the central finishing zone. Space is not merely occupied; one movement creates the conditions for another movement elsewhere.

NAPOLI VS COMO | FOOTBALL HACKING WEB APP ANALYSIS

Third-man combinations are a principle — not a dogma

Another recurring feature is Como’s use of the third-man principle. During build-up and progression, the obvious receiver is frequently not the real destination of the play. A player receives in order to connect with another player positioned beyond the immediate pressure, allowing Como to progress without asking the final receiver to receive directly from the original ball carrier. This is especially valuable when opponents close obvious vertical lanes: the third man changes the angle of progression before the defensive structure can fully adjust.

This behaviour is closely connected to another detail visible on video: Como players repeatedly try to become available on the blind side of the nearest defender. The player providing progression support is therefore not simply searching for an empty zone. He is trying to occupy a position where the opponent has difficulty controlling both the ball and the receiver simultaneously. If that blind-side player is not initially available, Como often use movement around the ball to create him.

However, Como are not doctrinaire about combination play. As the attack approaches the penalty area, there are moments when the team deliberately becomes more aggressive. The third-man mechanism can give way to what we can call a “space of force”: once a sufficiently dangerous space has been created, Como do not necessarily need another elegant combination. They may attack it immediately with a direct pass, run, cross, cut-back or shot.

That distinction matters. The third man is a method of creating advantage; it is not an obligation to continue combining when the advantage already exists. Como can patiently manufacture the opening and then become vertical when the probability of turning that opening into a shot increases. In that sense, their positional play is pragmatic rather than decorative.

Read more

Two upcoming matches. One question: what do the numbers actually support?

The short video below shows Football Hacking’s prediction engine in action — combining goal dispersion, Poisson probabilities and fair odds with Monte Carlo robustness testing to measure not only what is likely to happen, but how much confidence we can place in the prediction.

▶️ Watch the predictions below.

Want to test it yourself? Get 7 days free access to the Football Hacking Web App and explore the prediction tools, tactical metrics, passing networks, xT analysis and more.

Run the numbers. Test the prediction. Make your own decision.

Data disclaimer: No match images or video footage were used in this analysis. The conclusions are based exclusively on event-level match data, pass-network structure, Voronoi spatial control, xT profiles, Network Health, player-role centralities and Functional Movement Maps.

Manchester City survived the red card, changed their spatial behaviour and still found the winning goal

Manchester United’s 0–1 defeat to Manchester City in Round 4 of the Premier League 2026/27 is an unusually interesting tactical case because the game effectively contains two different matches.

The first lasted a little over 22 minutes.

City had eleven players, broader territorial control and the stronger xT profile across practically every attacking category.

Then Phil Foden was sent off in the 23rd minute. The dismissal was immediately followed by a City formation change, making the red card the clearest structural breakpoint in the entire game.

From that point onward, the question was no longer whether Manchester City could dominate Manchester United in conventional terms.

It became something more interesting:

Could City protect enough of their passing structure, maintain enough threat after recoveries and survive United’s increasing territorial control with ten men?

The answer was yes.

And the way they did it explains the 1–0 result far better than the scoreline alone.

Subscribe now

0’–23’: Manchester City controlled more of the pitch before the red card

Before Foden’s dismissal, the xT radar presents the clearest version of City’s original plan.

Manchester City hold the advantage in Offensive Transition xT, Recovery Zone xT and all three main progression corridors.

The differences are not enormous in every category, but they are remarkably consistent.

City were not dependent on one flank.

They were producing useful destinations through the right corridor, central corridor and left corridor, while also creating greater threat immediately after recovering possession.

The pass network reinforces that picture.

City occupy large areas of the pitch through their Voronoi control and maintain connections from the first line into midfield and then toward their advanced players.

United are present in advanced areas, but their network is less territorially dominant.

The Network Health data tells the same story in another way.

City post strong early readings of 0.73 between 5’–10’ and 0.74 between 15’–20’. United remain competitive, but City possess the healthier passing structure across much of the opening period.

This was a Manchester City side capable of controlling both circulation and territory.

Then the match changed completely.

Foden’s red card removed an important right-sided reference

The dismissal matters not only because City lost a player.

It matters because of which spatial function disappeared.

Before the red card, Phil Foden was operating repeatedly through City’s right corridor and right half-space, moving between wide and interior positions and offering a progressive reference ahead of the ball.

His presence helped connect City’s right side with more central attacking spaces.

Alongside him, Rayan Cherki occupied more interior and right-central zones, giving City another receiving option between lines.

That created a useful relationship: one player could stretch or move outside, while the other could occupy the space created inside.

Once Foden disappeared, City could no longer reproduce that same geometry with eleven-man spacing.

The movement map after 22.5 minutes changes immediately.

And that is where this derby becomes much more tactically interesting.

23’–45’: Manchester United gain territory — but City protect transition threat

The xT radar from roughly 25’–45’ looks very different from the opening phase.

Manchester United now create a substantial advantage through the left corridor.

Their central threat also becomes more competitive.

This is exactly what we would expect from the numerical advantage: United can push another reference forward, stretch City’s ten-man block and occupy more territory around the final third.

But City retain two important advantages.

Their Recovery Zone xT remains clearly stronger, and their Offensive Transition xT also stays above United’s.

That distinction is crucial.

United had more players.

United gained more positional territory.

But every additional player committed ahead of the ball also created a possible transition environment for City.

The red card changed possession control.

It did not eliminate City’s capacity to hurt United.

United stretch the pitch after the dismissal

The Functional Movement Maps make United’s reaction especially clear.

After the red card, Patrick Dorgu pushes high into the left corridor, with his reception pattern showing movement away from his median passing position.

On the opposite side, Bryan Mbeumo provides depth and width through the right corridor.

United are now trying to make the numerical superiority physically visible.

Instead of concentrating an additional player in the same central zone, they stretch City’s defensive distances from both sides.

That is the correct spatial logic against ten men:

make the opponent defend width first, then attack the gaps that appear inside.

The pass-network map from 25’–45’ reflects this territorial change. United control considerably more of the areas around City’s half, while their central network becomes denser.

Yet territorial control and attacking control are not identical.

City were already beginning to prove that.

City become less expansive and more selective

The post-red-card movement map shows City abandoning part of their earlier positional ambition.

Rayan Cherki, now operating toward the right-sided attacking corridor, shows a large reception displacement rather than simply remaining in a fixed advanced position.

Further back, Enzo Fernández becomes increasingly important through the central and left-central midfield zones.

This is a meaningful change.

With eleven players, City could occupy many spaces simultaneously.

With ten, they increasingly needed individual players to connect multiple spaces.

That creates a different type of movement.

Instead of the structure placing a player everywhere the ball might go, fewer players have to move farther to become available.

City become more functional and selective.

They stop trying to own every zone.

They try to own the next useful one.

The first-half Network Health numbers reveal something unexpected

Manchester United’s structural response after the dismissal is strong.

Between 25’ and 35’, United reach 0.67 and 0.64 Network Health, their best sustained spell of the first half.

City fall relative to their eleven-man levels, but they do not collapse.

That is important.

A sending-off often creates immediate structural damage: fewer passing options, greater distances between players and isolated receivers.

City avoid the worst version of that problem.

Even with ten players, they remain around 0.49, 0.61, 0.52 and then 0.65 through different five-minute phases before halftime.

The shape is more volatile.

But it remains functional.

That volatility becomes one of the major themes of the match.

United become the more stable territorial team. City become the team capable of surviving structural instability without losing all attacking potential.

45’–70’: City stop trying to reproduce eleven-man football

The beginning of the second half shows the tactical adjustment even more clearly.

The xT profiles between 45’ and 70’ are remarkably balanced across the three main corridors.

United no longer hold the obvious left-sided advantage they produced immediately after the sending-off.

Central and wide progression values become similar.

Only one category remains consistently in City’s favour:

Recovery Zone xT.

That is the key.

City’s greatest weapon after the red card was not prolonged positional domination.

It was what happened after possession changed hands.

Their compactness reduced the distances required to recover defensive positions, but it also gave them opportunities to attack once United’s structure was temporarily open.

City’s movement becomes dramatically more economical

The second-half opening movement map is perhaps the most revealing graphic of the entire match.

City show far fewer major reception displacements than United.

That is not accidental.

With ten men, constant positional rotation carries a cost. Every movement away from the structure creates another space that someone else must cover.

City therefore become noticeably more static.

Erling Haaland remains primarily around the central attacking channel, providing a reference for direct progression without constantly leaving the zone.

Behind him, Enzo Fernández stays concentrated around the central midfield corridor, helping preserve access between the deeper structure and the forward line.

This is not the fluid City visible before the dismissal.

It is a more economical version.

Movement is no longer being used primarily to expand the pitch.

It is being rationed.

United move more because they now have to break a compact ten-man block

Manchester United display almost the opposite behaviour.

Patrick Dorgu continues making significant movements from the left corridor, seeking higher reception points.

On the other side, Diogo Dalot operates through the right corridor, with United trying to maintain width around City’s compact central structure.

The numerical advantage gives United more passing options.

But City can now defend with a clearer priority:

protect the centre, accept some external circulation and wait for moments when United’s spacing becomes vulnerable after possession is lost.

This helps explain a paradox in the match.

United increasingly control where the game is played.

City still create the decisive action.

Subscribe now

62’: Manchester City score with ten men

The winning goal arrives in the 62nd minute.

The sequence begins with Josko Gvardiol attacking from the left side of City’s advanced structure and attempting a cross into the penalty area.

The ball is not cleanly completed, United make contact, and the second phase remains alive.

Erling Haaland, positioned in the right side of the six-yard area, reacts first and scores with his first touch.

This goal fits the tactical story almost perfectly.

City did not need a long sequence of territorial domination.

They needed the structure to get the ball into an advanced area, survive the imperfect first action and have a central reference ready for the second one.

Haaland’s role is especially important.

In ten-man football, the striker can easily become disconnected.

City avoid that.

They keep him close enough to the central attacking route that one unstable moment inside the box is enough.

The goal occurs during one of City’s weakest Network Health phases

This is where the Network Health chart becomes particularly valuable.

Around 55’–60’, City’s Network Health is only 0.32.

Between 60’–65’, when the goal arrives, it is approximately 0.35.

United are significantly stronger structurally during those windows, around 0.62 and 0.57.

Yet City score.

Again, this is why structure and outcome cannot be treated as the same thing.

Network Health tells us how coherent and functional the passing network is in a given phase.

It does not say which team will score next.

Manchester United had the healthier passing structure.

Manchester City produced the higher-value decisive action.

That distinction is central to understanding this derby.

The player-role map explains how City survived with ten

The full-match centrality analysis provides another important clue.

Despite playing most of the match with ten men, City finish with two structural hubs, while United have only one.

For City, Antoine Semenyo, operating largely from the left-sided corridor and left half-space, records the highest Flow Connector Index of the entire match.

He was not simply an external destination.

He repeatedly connected different parts of City’s possession structure.

Alongside him, Enzo Fernández, functioning through central and left-central midfield areas, combines high receiving influence with strong flow responsibility.

That pairing becomes vital after the red card.

City lose a player but retain two players capable of linking possessions rather than merely receiving them.

That is structural survival.

United had many preferred targets, but fewer true circulation hubs

Manchester United’s player-role profile is different.

Lisandro Martínez, operating from the left-central defensive and buildup zones, is the only United player classified as a structural hub.

Further forward, Kobbie Mainoo has the highest receiving influence in the entire match, mainly through the central midfield corridor, but his Flow Connector score is much lower.

That distinction matters.

United had multiple players the network wanted to find.

But receiving importance does not necessarily mean circulation responsibility.

A preferred target can help advance a possession.

A structural hub helps determine how the entire possession continues.

City had fewer players.

But some of those players carried a disproportionately important share of the connective burden.

70’–100’: Manchester United take even greater control of the field

Once City are ahead, the final phase becomes increasingly territorial.

United’s xT radar now shows a significant advantage through the right corridor, as well as greater Offensive Transition xT.

City continue to lead in Recovery Zone xT.

This creates a fascinating late-game split.

United are producing more attacking value through sustained territorial pressure and right-sided progression.

City remain most dangerous when the game changes state.

The pass network also becomes much denser for United.

Their spatial control extends across large areas of City’s half, while City’s network becomes sparse.

On paper, this is exactly the environment United wanted after Foden’s red card.

In practice, they still cannot convert that field control into an equaliser.

United’s late movement becomes wider and more aggressive

The final Functional Movement Map shows United making a final attempt to stretch City horizontally.

Patrick Dorgu remains very high in the left corridor, almost functioning as a permanent wide attacking reference.

On the opposite side, Diogo Dalot produces one of United’s clearest wide reception displacements through the right corridor.

That width forces City to defend the complete horizontal dimension of the pitch.

Meanwhile, United’s central players remain compact enough to attack second balls and recycle possession.

Structurally, this is one of United’s strongest phases.

The Network Health chart confirms it.

United peak structurally late — City nearly disappear from the network

From 70 minutes onward, United’s Network Health is extremely consistent:

0.61, 0.63, 0.61, then a match-high 0.73 between 85’–90’, followed by 0.65 and 0.63.

City’s numbers tell a very different story.

They fall to 0.16 between 75’–80’ and later just 0.13 between 90’–95’.

These are extraordinary contrasts.

United are circulating with their strongest structural coherence of the match.

City’s passing network is barely functioning during some of the same periods.

Yet City remain ahead.

This tells us something important about defensive game-state management.

At 1–0 with ten men, City’s objective is no longer to maximise passing-network quality.

Their priority becomes preserving the spaces that matter most.

The ball can belong to United.

The territory can belong to United.

Even Network Health can belong to United.

The central defensive zones cannot.

City’s late movement confirms the change in priorities

The final movement map reinforces that interpretation.

City’s player positions become much more static.

Matheus Nunes, on the right-side external corridor, provides one of the few higher outlets.

On the opposite side, Antoine Semenyo remains deeper around the left corridor, offering an escape route without permanently abandoning City’s compact block.

There are far fewer major reception displacements than earlier in the match.

City are no longer moving to create new attacking structures.

They are moving only when the game requires an outlet.

That is a fundamental tactical transformation from the opening 20 minutes.

Manchester City’s match can be divided into three tactical identities

The data suggests three clearly different versions of Manchester City.

Before the red card, they were expansive.

They controlled multiple corridors, had strong Recovery Zone xT, broad spatial occupation and a healthy passing network.

After the red card but before scoring, they became selective.

They reduced movement, relied more heavily on a small number of connectors and maintained transition threat despite conceding territory.

After taking the lead, they became protective.

The passing structure deteriorated sharply in some windows, but the team preserved compactness and external escape routes.

That progression is what makes this Manchester United 0–1 Manchester City tactical analysis much more interesting than a conventional derby win.

City did not win by maintaining one superior tactical model.

They won by abandoning one model and surviving with another.

Manchester United controlled more — but not the part of the game that decided the score

United’s frustration becomes clearer when all the graphics are viewed together.

After Foden’s dismissal, they had the numerical superiority.

Their pass network became healthier.

Their spatial control expanded.

Their left corridor became more dangerous before halftime.

Their right corridor became more productive late.

Their movement maps show greater positional variation and more aggressive reception patterns.

By almost every structural measure, United successfully forced the game toward City’s goal.

But City continued to dominate one metric throughout almost the entire match:

Recovery Zone xT.

That persistence matters.

Even with ten players, City remained better positioned to turn changes of possession into valuable attacking situations.

United controlled more of the field.

City controlled more of the dangerous moments immediately after the field changed ownership.

And eventually, one of those unstable moments became the only goal.

Manchester United 0–1 Manchester City: the tactical takeaway

The most important lesson from this Manchester derby tactical analysis is that numerical superiority does not automatically translate into functional superiority.

Manchester United successfully changed the geography of the match after the sending-off.

They stretched the pitch.

They increased territorial control.

They created a healthier passing network.

They forced City into extremely poor Network Health readings during the closing stages.

But Manchester City adapted the purpose of their structure.

Before the red card, structure was used to control the game.

After the red card, structure was used to survive transitions.

After the goal, structure was used to defend critical spaces.

That is a completely different challenge.

The defining feature of Manchester City’s 1–0 win was not that they remained dominant with ten men. It was that they understood they no longer needed to be dominant in the same way.

That distinction explains almost everything about the match.

Download the Manchester United vs Manchester City match data

For readers who want to work directly with the underlying information, paid subscribers can download the complete Manchester United vs Manchester City match-event XLSX below the paywall, ready for independent tactical analysis, custom visualisations, modelling and football data research.

Read more

Flamengo vs Corinthians brings together Brazil’s two largest fan bases, but the real story here is structural. The Football Hacking matchup looks at how both teams progress, react to pressure, preserve attacking value and behave across the five-pass window after resistance — without reducing the game to possession, shots or isolated metrics. The full tactical picture is available below the paywall.

Paid subscribers get the complete Flamengo vs Corinthians structural matchup graphic, a deeper PDF report designed for coaching-staff-level analysis, and the XLSX Poisson probability matrix generated from the Football Hacking models for those who want to explore score probabilities and build their own betting analysis. The full package combines tactical interpretation and probabilistic modeling in one match preview.

Subscribe now

Read more

Data disclaimer: No match images or video footage were used in this analysis. The conclusions are based exclusively on event-level match data, pass-network structure, Voronoi spatial control, xT profiles, Network Health, player-role centralities, progression patterns and Functional Movement Maps.

At the end of this Liverpool 2–1 Atletico Madrid tactical analysis, paid subscribers will be able to download the complete match-event XLSX and explore the underlying event data independently.

Subscribe now

How Liverpool turned an early Atletico advantage into a 2–1 Champions League win

Liverpool’s 2–1 victory over Atletico Madrid in Round 1 of the UEFA Champions League 2026/27 was not a simple story of one team establishing control and keeping it.

Atletico scored first and, during important sections of the first half, actually produced the stronger structural readings. Their ability to create threat after regaining possession remained one of the most persistent characteristics of the match.

Liverpool’s response was different.

They gradually improved the value of their right-sided progression, connected those wide routes to central receiving positions and, by the start of the second half, found a more effective balance between territorial occupation and vertical threat.

The most interesting part is that the two Liverpool goals came from different structural situations. The equaliser emerged after the right side became more productive. The winner came through the centre just after halftime.

And once Liverpool went ahead, the match changed again.

0’–25’: Atletico lead because their early threat is real

Atletico’s opening was not merely opportunistic.

The first xT radar gives them advantages in several important areas, most clearly in Recovery Zone xT. They also show stronger early values in offensive transition and through the right corridor.

The Network Health data supports that picture.

Liverpool begin at 0.49 and gradually improve through the opening phase, while Atletico move from 0.38 to 0.63 and then 0.66 around the period in which they score.

At 16 minutes, Julián Alvarez finds the ball in the left/central attacking channel and sends a through ball diagonally inside. Marcos Llorente, arriving centrally, finishes from outside the penalty area to make it 1–0.

The geography of that move is important.

Atletico did not simply progress vertically through one corridor. The ball moved from an external starting position toward the centre, where the final threat appeared.

That tendency appears elsewhere in their progression data as well.

When Atletico started through the left corridor, their attacks did not necessarily remain wide: 42.9% reached the final third through the left half-space, while another 42.9% reached it through the central corridor.

Their wide possession was often a starting point, not the destination.

The early movement maps show Atletico constantly changing the receiving picture

The Functional Movement Maps make this early advantage easier to understand.

On Atletico’s right corridor, Marcos Llorente shows a substantial difference between his median passing position and his main reception region. He was not simply stationed on the flank; his functional position shifted according to where the next possession developed.

Further along the same side, Marc Pubill also offered a changing high reference.

This matters because it created a moving target for Liverpool’s defensive structure.

Liverpool’s own width was visible, but organised differently.

On the left corridor, Milos Kerkez provided a high positional reference, while Liverpool retained another wide outlet on the opposite side. The structure was broad, but the early reception patterns were not yet producing the same level of threat after turnovers that Atletico were generating.

That difference is captured particularly well by Recovery Zone xT.

Atletico were creating more value from the moments immediately after winning the ball.

25’–46’: Liverpool change the most productive side of the game

The second part of the first half is where this Liverpool vs Atletico Madrid tactical analysis begins to shift.

Liverpool’s right corridor xT increases considerably.

Atletico retain their advantage in Recovery Zone xT, but Liverpool now have a more obvious route for advancing possession into valuable territory.

The Functional Movement Map shows the same behavioural adjustment.

The team becomes less stretched vertically and more connected between the middle and attacking zones.

Instead of relying simply on fixed wide references, Liverpool create shorter relationships around the central block and then use the outside corridor as an entry point.

That becomes directly relevant to the equaliser.

Liverpool’s equaliser connects the right side to the centre

At 39 minutes, Ronald Araujo receives in the advanced right-side channel and plays a short ball inward. Dominik Szoboszlai finishes from the central area of the box to make it 1–1.

This is a useful example of why corridor analysis should not be reduced to “Liverpool attacked on the right.”

The important part is what happened after the right corridor was accessed.

The external route created the entry.

The final action moved inside.

And that fits the broader change in Liverpool’s attacking structure during this window: more useful occupation outside, but greater connectivity around the centre.

Subscribe now

A fascinating Network Health contradiction

There is another detail worth highlighting.

Atletico reach 0.72 Network Health around 25’–30’ and an even higher 0.74 around 35’–40’.

Liverpool equalise during that second window.

So Atletico’s passing structure was functioning extremely well when they conceded.

That is precisely why structural analysis and goal analysis need to remain separate.

A healthy network tells us that a team is circulating effectively, maintaining relationships and distributing the ball through a coherent structure.

It does not guarantee that the next decisive action belongs to that team.

Liverpool were not necessarily dismantling Atletico’s entire structure.

They were becoming more efficient at reaching the areas where the structure could be punished.

First-half control was divided between two different strengths

By halftime, neither team had achieved complete control.

Atletico had shown the stronger ability to transform recoveries into threatening possession.

Liverpool had gradually created the better connection between corridor occupation and central finishing access.

This difference also appears in their progression patterns.

Liverpool are strikingly persistent on the left: when possession starts through the left corridor, 62.5% of the identified sequences also reach the final third through the left corridor.

When they begin through the left half-space, 50% continue into the final third through that same half-space.

That suggests Liverpool were not always using width to switch the point of attack.

Sometimes the objective was to keep progressing through the same lane until Atletico were forced deeper.

Atletico’s progression was more inclined to convert wide possession into an interior destination.

Two different spatial ideas were competing.

45’–70’: Liverpool take the lead by attacking through the centre

The beginning of the second half brings another tactical change.

Liverpool’s Offensive Transition xT improves, while Atletico remain clearly superior in Recovery Zone xT.

The corridor values themselves become more balanced.

That is important: Liverpool no longer need one particular flank to carry the entire attacking threat.

Their winning goal arrives almost immediately.

At 51 minutes, Alexander Isak provides the final central connection before Alexis Mac Allister scores from outside the box, also from a central position.

Liverpool’s two goals therefore tell two slightly different tactical stories.

The first connected the right side to the centre.

The second emerged directly through the middle.

By this point, Liverpool had become more difficult to defend because the final action no longer depended on the same entry route.

Liverpool become more positional after halftime

One of the most interesting details comes from the 45’–67.5’ Functional Movement Map.

Liverpool display fewer large reception displacements than in other phases of the game.

That should not automatically be interpreted as less attacking ambition.

It suggests that their passing positions and principal receiving positions were becoming more closely aligned.

In the central corridor and right half-space, Dominik Szoboszlai and Florian Wirtz operate within a much more recognisable positional framework.

Liverpool are now asking the structure to do more of the work.

The distances between the passing references are stable.

The ball moves through the structure.

The players do not constantly need to abandon it to become available.

That is a form of control.

Atletico react with more movement around the outside

Atletico’s behaviour after falling behind is different.

Their movement map remains more dynamic, especially in the external lanes.

Marc Pubill continues changing his reception point along the right corridor, while Dávid Hancko begins showing wider reception behaviour on the opposite side.

Atletico are trying to move Liverpool’s block by altering the location of the next receiver.

That helps explain why their Recovery Zone xT remains so strong.

When they regain possession, they have players capable of becoming available somewhere different from their previous passing position.

The problem is what happens next.

Liverpool are increasingly effective at slowing those possessions once they approach the central defensive spaces.

Liverpool’s defensive control was about killing rhythm, not stopping every progression

The full-match defensive patterns are revealing here.

When Liverpool defend the central zone of their defensive third, Atletico’s reaction tends to become neutral and the attack loses rhythm.

When Liverpool engage in the right half-space of midfield, Atletico are frequently forced to recirculate.

This is an important distinction.

Atletico can progress.

Atletico can recover the ball in useful areas.

Atletico can even produce high Recovery Zone xT.

But Liverpool are often able to prevent one good action from immediately becoming the next good action.

That interruption matters.

Control does not always mean winning the ball. Sometimes it means forcing the opponent to restart the attack.

The player-role map reveals Liverpool’s central structural advantage

The full-match centrality analysis adds perhaps the clearest explanation for Liverpool’s ability to reorganise the match.

Florian Wirtz is the standout structural hub.

Operating predominantly around the central corridor and adjacent attacking half-space, he combines extremely high receiving influence with the highest Flow Connector value in the match.

In other words, he is important in two different ways.

The network wants to find him.

And once possession reaches him, a significant amount of Liverpool’s circulation continues through him.

That is more valuable structurally than simply recording a high number of passes.

Meanwhile, Dominik Szoboszlai, operating through central and right-sided interior areas, has the highest Receiving Influence Index in the match.

His profile is different.

He is an especially important destination.

Wirtz is both destination and bridge.

This distinction helps explain why Liverpool could change attacking routes without completely changing their passing identity.

Atletico had structural influence, but more peripheral roles

Atletico also possess influential players.

Dávid Hancko, working mainly through the left defensive corridor and left-side buildup, combines strong receiving influence with above-average connection responsibility.

But the overall player-role picture is more fragmented.

The chart identifies 11 Atletico players in peripheral roles compared with six for Liverpool.

That does not mean those players were uninvolved.

It means fewer of them combined high network influence with high responsibility for the flow of possession.

Liverpool therefore had more ways to redistribute the circulation when one route became difficult.

This becomes especially important after they take the lead.

70’–100’: Atletico become more dangerous without taking control of the result

The final xT radar changes sharply.

Atletico now have their strongest Offensive Transition xT of the match.

Their Recovery Zone xT remains substantially above Liverpool’s.

This is exactly the profile of a team pushing for an equaliser while trying to exploit every turnover.

The pass network also becomes more dispersed.

Atletico create wider receiving points and increasingly stretch the match across several lanes.

They are no longer simply trying to construct the cleanest possible possession.

They need to destabilise Liverpool.

Atletico stretch both sides; Liverpool compress the middle

The late Functional Movement Maps show that contrast beautifully.

For Atletico, Marc Pubill continues to move along the right corridor, repeatedly changing the width of his receiving position.

On the opposite side, Dávid Hancko creates a similarly wide reception pattern through the left corridor.

This gives Atletico width on both sides while central players occupy the spaces between them.

Liverpool respond differently.

Their late shape becomes more concentrated through the middle.

Dominik Szoboszlai now shows more pronounced reception movement around the central corridor and adjacent half-space, while the team’s external positions remain available as escape routes.

Liverpool are conceding some territorial initiative.

But they are protecting the centre of the structure.

The late Network Health numbers explain why Liverpool survived the pressure

Atletico’s increased threat does not translate into a structural collapse from Liverpool.

Quite the opposite.

Liverpool reach their highest Network Health reading of the match — 0.72 — around 80’–85’.

After a temporary drop, they recover again to 0.61 and 0.63 in the final windows.

Atletico also remain competitive, reaching 0.61 around 90’–95’.

So this was not a finish in which one team had structure and the other simply attacked desperately.

Both teams retained functioning networks.

The difference was that Liverpool could maintain their own while defending a lead.

That is a valuable form of late-game resilience.

Field control changed repeatedly during Liverpool 2–1 Atletico Madrid

This match is particularly interesting because control never belonged permanently to one team.

During the opening phase, Atletico had the better recovery-generated threat and scored first.

Liverpool gradually improved their right-sided progression and connected it to the centre for the equaliser.

At the beginning of the second half, Liverpool became more positionally stable, increased their transition threat and scored through a central route.

Then Atletico changed again.

They created more movement on the outside, increased their late transition threat and continued producing strong value after recoveries.

Liverpool responded not by chasing complete territorial domination, but by protecting central relationships and keeping their own passing network functional.

There were therefore at least four versions of this match.

And the scoreline alone shows none of them.

Liverpool 2–1 Atletico Madrid: the tactical takeaway

The most important lesson from this Liverpool vs Atletico Madrid Champions League tactical analysis is the difference between movement, structure and threat.

Atletico often had more functional movement.

Their reception points shifted significantly, particularly in the external corridors.

They also maintained an exceptional ability to generate value after winning possession.

Liverpool had something else.

Their network eventually became more stable around influential central players, their right corridor became productive before halftime, and their second-half attacking threat no longer depended on one lane.

That gave them tactical redundancy.

When the outside route worked, they used it.

When the central route opened, they attacked it.

And when Atletico increased the pressure late, Liverpool could reduce movement, compress the structure and defend the spaces that mattered most.

Atletico repeatedly changed the geometry of the game. Liverpool were better at deciding what happened once that geometry opened a decisive space.

That is the difference the 2–1 score does not explain on its own.

Download the Liverpool vs Atletico Madrid match data

For readers who want to go deeper, paid subscribers can download the complete Liverpool vs Atletico Madrid match-event XLSX below the paywall, containing the event-level data needed for independent analysis, custom visualisations, tactical research and your own football data projects.

Read more

Data disclaimer: No match images or video footage were used in this analysis. The conclusions are based exclusively on event-level data, pass-network structure, Voronoi spatial control, xT profiles, Network Health, player-role analysis, progression patterns and Functional Movement Maps.

Real Madrid won the decisive moments. Inter controlled more of the game than the 2–0 first-half score suggested.

Real Madrid’s 2–1 win over Inter in the UEFA Champions League 2026/27 is a strong example of why the scoreboard and tactical control should never be treated as synonyms.

Madrid scored twice in the opening 22 minutes. Yet Inter’s passing structure remained competitive — and often healthier — through much of the first half. Their Recovery Zone xT was consistently stronger, their central progression remained functional, and their movement patterns repeatedly changed the point from which players received the ball.

Real Madrid’s advantage came from something different: greater efficiency in decisive spaces, a more distributed network of influential players, and an increasingly strong ability to attack from the left while keeping multiple central references available.

By the second half, the tactical balance changed again.

Madrid became structurally stronger. Inter became more aggressive in their movement. And the final 20 minutes turned into a contest between Real Madrid’s attempt to stabilise territory and Inter’s ability to generate new reception points around a more dynamic shape.

Subscribe now

0’–22.5’: Inter’s structure was healthier — Real Madrid scored twice

The opening phase is the most counterintuitive part of this Real Madrid vs Inter tactical analysis.

Inter’s Network Health starts at 0.66, compared with Real Madrid’s 0.51. They remain competitive in the following windows and clearly outperform Madrid structurally during parts of the first 25 minutes.

The xT radar tells a similar story.

Inter have a particularly clear advantage in Recovery Zone xT, while their central-corridor threat is at least comparable to Madrid’s. Real Madrid’s most distinctive attacking advantage appears instead through the left corridor.

So Madrid were not simply overpowering Inter territorially.

They were finding better outcomes from specific attacking situations.

At 13 minutes, Brahim Díaz received centrally and supplied Kylian Mbappé, who finished from the centre of the penalty area. The move converted a central progression directly into a high-value finish.

The first goal therefore did not require Madrid to dominate every zone. They needed one clean progression through the middle.

Nine minutes later, the score became 2–0 for a completely different reason. An Inter error close to goal preceded Federico Valverde’s finish from the central area of the box.

That distinction is critical.

Real Madrid’s two-goal advantage was larger than their structural advantage.

The opening movement maps show two very different ways of occupying the pitch

Real Madrid’s first Functional Movement Map is comparatively positional.

On the left corridor and left half-space, Marc Cucurella occupies a high reference while Federico Valverde works inside him. Valverde’s reception pattern suggests movement away from his median passing position, but the overall structure remains recognisable: width outside, connection inside.

This relationship becomes important because Madrid’s full-match progression data shows a remarkably consistent route: when they started through the left corridor, every identified sequence reached the final third through the left half-space.

Madrid were using width to open an interior destination.

Inter’s movement was different.

On their left side and central-left channel, Carlos Augusto showed much greater displacement between passing and reception positions, while Alessandro Bastoni provided the deeper structural reference behind him.

This is an early indication of a pattern that becomes even more visible later: Inter were often more mobile around their base structure, even when Madrid were more effective in the decisive areas.

Inter’s central route remained alive despite the 2–0 deficit

The Inter progression patterns help explain why the match did not become comfortable for Real Madrid.

When Inter began through the central corridor, 80% of those sequences also entered the final third centrally.

And when they started through the left corridor, 75% eventually reached the final third through the centre.

That is a revealing combination.

Inter were not using the flanks only to progress along the touchline. Wide possession could become a mechanism for accessing the middle.

The pass networks show this repeatedly: several connections converge around central midfield before the next forward action.

For Madrid, that created a defensive problem very different from simply stopping crosses. They had to protect the half-spaces and central receiving zones after Inter had already moved the ball laterally.

Subscribe now

22.5’–45’: Inter refuse to disappear from the match

The period after Madrid’s second goal is where the scoreline becomes especially misleading.

Inter continue to show a much stronger Recovery Zone xT profile.

Their passing network also remains highly connected through the middle and defensive structure, while Madrid’s Voronoi control does not expand into anything resembling complete territorial domination.

The movement maps make the contrast even sharper.

In Inter’s upper-left and central-left areas, Andy Diouf begins receiving away from his median passing position, while the deeper structure around the same side changes its reception points significantly.

Those arrows matter because they indicate that Inter were not simply circulating in fixed positions.

Their players were passing from one zone and repeatedly appearing in another to receive again.

Madrid, meanwhile, remain more compact around the middle. Their upper-left attacking structure continues to be important, but the movement is generally attached to a more stable positional skeleton.

This is not necessarily better or worse.

It tells us that the teams were solving the match differently.

Madrid preserved references. Inter kept changing them.

Inter came close to reducing the deficit before halftime

The structural resistance was not purely cosmetic.

At 35 minutes, Curtis Jones struck the post from the central area of the penalty box.

That moment fits the broader picture of the first half.

Inter’s Recovery Zone xT had remained stronger, their network had not collapsed after conceding twice, and their central progression routes were still functioning.

The 2–0 score did not describe a team being tactically removed from the match.

It described a team being punished much more efficiently than it was punishing its opponent.

The player-role map explains why Real Madrid could survive without dominating every phase

The full-match player-role analysis contains one of the strongest clues to Madrid’s eventual control.

Real Madrid have four structural hubs compared with Inter’s two.

More importantly, Inter have a much larger group classified in peripheral roles.

That means Real Madrid’s influence is spread across more players capable of combining receiving importance and connection responsibility.

The clearest example is Federico Valverde, who sits at the very top of the flow-connector scale while also showing strong receiving influence.

Alongside him, Jude Bellingham appears as another important structural reference rather than simply an advanced target.

Madrid therefore had redundancy.

If one route was blocked, the passing structure did not necessarily lose its principal connector.

Inter had powerful hubs too. Alessandro Bastoni is arguably their clearest structural reference, with extremely high scores in both receiving influence and flow connection, while Yann Bisseck also occupies the structural-hub area.

But around those hubs, Inter’s network contains considerably more peripheral players.

That difference becomes increasingly important as the match progresses.

45’–70’: Real Madrid change the nature of their control

The start of the second half is where Madrid’s performance becomes more convincing structurally.

The xT profile changes substantially.

Real Madrid now lead clearly in Offensive Transition xT and produce their strongest threat through the left corridor, while Inter maintain their familiar advantage in Recovery Zone xT.

Madrid’s Network Health also rises sharply.

Around 55’–60’, they reach 0.70, one of their best values of the entire match, while Inter drop to approximately 0.28.

For perhaps the first time, scoreline control and structural control begin to align.

The Voronoi map also shows Madrid occupying broader areas around the attacking half rather than relying on isolated forward positions.

This is an important change from the first half.

Madrid are no longer simply winning the decisive moments.

They are becoming better at controlling the conditions from which those moments can emerge.

Madrid’s second-half movement becomes much more expansive

The Functional Movement Map makes that behavioural change especially visible.

On the left corridor, Marc Cucurella now has an enormous horizontal difference between his median passing position and his principal reception areas.

Inside and ahead of him, Vinícius Júnior operates much higher on the opposite side of that attacking structure, giving Madrid a second reference that can stretch the pitch.

The key is not simply that both occupy wide areas.

It is that the reception patterns increase the distance Inter must defend between where the ball is released and where the next player becomes available.

Madrid are beginning to move the opponent rather than merely moving the ball.

That is a different form of territorial control.

Inter continue to move — but more of that movement is happening behind Madrid’s pressure

Inter’s second-half opening movement map remains highly dynamic.

In the central-left corridor, Hakan Çalhanoğlu moves between deeper and more central reception zones, while Carlos Augusto continues to create large changes in his receiving position on the flank.

But notice where much of the movement happens.

Inter are frequently moving to escape pressure before progressing.

That fits another full-match pattern: when Inter are under resistance in the central attacking third, 42.9% of their subsequent passes return into the central midfield zone.

In practical terms, Madrid were increasingly forcing Inter’s possession away from immediate attacking value.

The ball was not necessarily being lost.

But the attack was being reset.

Real Madrid’s defensive control worked through the half-spaces

Madrid’s most effective defensive patterns appear around their own penalty area.

When they defended the right half-space in the defensive third, Inter tended to recirculate and lose attacking rhythm.

On the opposite left half-space, the reaction was more neutral, but again the sequence usually lost momentum.

This is significant because Inter’s preferred progression repeatedly aimed at central destinations.

Madrid did not have to prevent the first forward pass.

They needed to prevent the next central reception from becoming an immediate continuation.

That is the difference between stopping progression and merely delaying it.

For long stretches of the second half, Madrid were better at the latter.

70’–95’: the match changes again — Inter recover threat

The final xT radar contains perhaps the most dramatic contrast of the game.

Real Madrid maintain an advantage in Offensive Transition xT and remain slightly stronger through the right and central corridors.

But Inter’s Recovery Zone xT explodes upward, reaching by far the largest value on the late radar.

This is exactly what you would expect from a team chasing the match while still capable of turning regained possession into forward value.

And the movement maps confirm that Inter were not simply adding bodies forward.

They were changing where players became available.

On the left and left half-space, Andy Diouf shows a substantial displacement toward a new reception area, while Petar Sučić also appears between zones rather than remaining fixed to his median passing position.

The late Inter structure becomes more fluid precisely as the match becomes more urgent.

Inter’s goal was the consequence of that late functional threat

The goal arrives at 78 minutes.

Lautaro Martínez supplies the final pass from an advanced inside-right position into the centre, where Carlos Augusto finishes from the central area of the penalty box.

The sequence is significant because it compresses several of Inter’s match-long tendencies into one action.

A player receives high on one side.

The attack is redirected centrally.

A different player arrives into the finishing zone.

Inter had been trying to turn external or half-space possession into central access throughout the match.

At 78 minutes, it finally produced the goal.

Real Madrid become more positional after Inter score

The final Madrid movement map is striking for the opposite reason.

There are far fewer significant reception arrows.

Players are still distributed across the pitch, but their main passing and receiving zones are now much closer to one another.

That suggests a more fixed late-game structure.

On the central-right attacking side, Kylian Mbappé remains an advanced reference, while Jude Bellingham occupies a deeper interior position rather than repeatedly moving into radically different reception zones.

Madrid are no longer trying to generate the same volume of functional displacement.

They are protecting the geometry of the team.

That is a subtle but important behavioural change.

Early in the second half, movement was being used to expand the pitch.

Late in the match, position itself becomes the tool for controlling it.

Network Health shows why goals and control must be separated

This game offers another useful warning about reading structural metrics too literally.

Real Madrid’s first goal arrives during a relatively strong structural window.

But the second comes around a period when Inter’s network is extremely healthy.

Later, Madrid’s strongest Network Health window appears around 55’–60’ without producing a goal.

And Inter score at 78 minutes even though Madrid’s structural reading remains competitive.

That does not make Network Health less useful.

It makes its meaning clearer.

Network Health tells us how well the passing structure is functioning. It does not tell us who will score the next goal.

Football is still decided by what happens when structure meets the penalty area.

Why Real Madrid ultimately had the more resilient passing network

Inter’s best structural players were highly influential, but Madrid distributed responsibility more effectively across the network.

That can be seen in the player-role map:

Real Madrid have more structural hubs.

They have fewer peripheral roles.

And their most influential players occupy different areas rather than clustering the network around one zone.

That makes the passing structure harder to dismantle completely.

If the left route is closed, the central hub remains.

If the central route is pressured, another receiving reference exists elsewhere.

Inter had strong circulation and excellent periods of structural health, but more of their overall network remained dependent on a smaller core.

This difference is easy to miss if we only watch possession totals or count passes.

It becomes much clearer when receiving influence and flow responsibility are considered together.

What Real Madrid 2–1 Inter tells us about control

There was no single form of control in this Champions League match.

Inter had more structural control than the first-half score suggested.

Their passing network stayed healthy, their Recovery Zone xT was consistently strong, and they repeatedly found central progression routes.

Real Madrid had greater control of decisive spaces.

They converted a central combination for the opener and punished an error for the second.

After halftime, Madrid added something they had lacked earlier: sustained territorial and structural control. Their Network Health improved, their left-sided threat increased and their functional movement began stretching Inter across the pitch.

Then the match changed once more.

Inter’s late movement increased. Recovery-generated threat surged. They scored through a move that entered the central finishing zone after originating from the side.

Madrid responded by becoming more positional and less expansive.

That final change may be the most interesting one.

The team that had used movement to establish control eventually used structure to preserve it.

The tactical takeaway from Real Madrid 2–1 Inter

The final score describes a narrow Real Madrid victory.

The underlying data describes a match with at least three distinct tactical phases.

First, Inter circulated better than the score suggested, while Madrid finished better than they controlled.

Then Madrid expanded their influence after halftime, with stronger movement, better territorial occupation and more attacking value through the left corridor.

Finally, Inter increased their functional movement and recovery threat, forced the game back into contention, and made Madrid retreat toward a more stable positional structure.

That is why combining xT, pass networks, Voronoi field control, Network Health, player roles and Functional Movement Maps matters.

Each model answers a different question.

Together, they explain not only who controlled the game, but what kind of control existed — and when it changed.

Download the Real Madrid vs Inter match data

The analysis does not have to stop with the article. Paid subscribers can download the complete Real Madrid vs Inter match-event XLSX below the paywall, with event-level information for independent analysis, custom visualisations and deeper tactical research.

Read more

Data disclaimer: No match images or video footage were used in this analysis. The conclusions are based exclusively on event-level match data, pass-network structure, Voronoi spatial control, xT profiles, Network Health, player-role analysis and Functional Movement Maps.

How Arsenal overturned Chelsea through structure, movement and control of the pitch

Arsenal’s 2–1 win over Chelsea in Premier League 2026/27 Round 3 is a particularly useful match for separating three ideas that are often treated as if they were the same thing: having territorial occupation, having a healthy passing structure, and actually turning those advantages into dangerous movement.

Chelsea scored almost immediately. Arsenal then equalised before half-time and completed the comeback early in the second half. But the most interesting part is not simply the order of the goals.

The data suggests that Arsenal were structurally stronger for long periods before the scoreboard reflected it, while Chelsea remained capable of producing danger whenever they escaped that structure quickly.

0’–25’: Chelsea score first, but Arsenal already control more of the structure

Chelsea took the lead after only one minute. Morgan Rogers scored from a central area outside the box following a set-piece sequence.

That detail matters.

The opening goal did not emerge from sustained control in open play, and the Network Health chart immediately shows the contrast. In the opening five-minute window, Arsenal are at 0.71, while Chelsea are down at 0.19.

So the team that conceded was already displaying the healthier passing structure.

That pattern continues through most of the opening 25 minutes. Arsenal remain between roughly 0.58 and 0.72, producing their strongest first-half structural reading just before the equaliser.

The xT profile adds another layer. Chelsea have an advantage in Recovery Zone xT, suggesting they can create value after regaining possession, but Arsenal show more threat through the right corridor while remaining competitive through the central route.

The pass network explains why.

On Arsenal’s right corridor and right half-space, Bukayo Saka provides the wide reference while Kai Havertz occupies a more interior and advanced position. The movement map shows that neither is simply standing on a fixed coordinate: their main reception regions extend beyond their median passing positions, creating depth and a constantly changing receiving picture on that side.

This is an important distinction.

Arsenal were not simply occupying the right side. They were using it as a platform from which players could change their receiving positions.

Chelsea’s early attacking structure was different. Pedro Neto occupied a high position in the right corridor, while Cole Palmer worked closer to the central and right-half-space zones. The structure offered Chelsea an outlet, but the movement map suggests fewer large functional displacements around the central network than Arsenal were producing on the opposite side.

Subscribe now

The result was an interesting early contradiction: Chelsea had the lead; Arsenal had more structural control.

Arsenal’s right-side structure eventually produces the equaliser

The equaliser at 24 minutes is much more closely connected to what the positional data had been showing.

The move developed through Arsenal’s right side. Declan Rice received in the advanced right corridor and played the final pass toward the central attacking zone, where Kai Havertz scored from outside the box.

This was not an isolated visual pattern suddenly appearing with the goal.

Arsenal’s Functional Movement Map had already shown the right side as their most clearly defined attacking platform. The pass network had also positioned the main attacking references there, while the xT radar showed Arsenal’s advantage through the same corridor.

That is exactly the kind of alignment that makes a tactical pattern more convincing: position, passing relationships, movement and eventual chance creation all point in the same direction.

25’–45’: Arsenal expand control beyond the right flank

The second part of the first half is where Arsenal’s advantage becomes broader.

In the 25’–45’ xT profile, Arsenal now lead Chelsea not only through the right corridor, but also show stronger values through the central corridor and left corridor.

That matters because Arsenal are no longer dependent on one attacking route.

The pass network becomes more balanced around midfield, while the Voronoi picture shows Arsenal gaining territory across several zones rather than simply building around Saka’s side.

The movement map confirms the change.

On the left side and left half-space, Riccardo Calafiori begins to provide a higher receiving reference, while the central structure remains connected enough to transfer the ball across the pitch.

At the same time, Arsenal’s right-sided players continue to stretch the opposite side.

The result is a wider functional structure: Chelsea now have to defend not only an overload but the possibility that the attack will be moved away from that overload.

That matches one of the clearest full-match progression patterns in the data. When Arsenal begin through the right half-space, 60% of those sequences reach the final third through the left corridor.

That is an extremely useful tactical clue.

The right half-space was not always Arsenal’s final destination. It could be the starting point for shifting Chelsea before accessing the opposite side.

Chelsea were still dangerous when the game opened

Arsenal’s superior structure did not eliminate Chelsea’s transition threat.

Near the end of the first half, Pedro Neto attacked from the right side of Chelsea’s attack and hit the post from inside the right side of the box during a fast transition.

This is important when evaluating control.

Arsenal controlled more of the passing structure, but Chelsea could still break that control when the game became less organised.

That distinction runs through almost the entire match.

Arsenal were better at controlling the framework of the game. Chelsea were often most dangerous when they could temporarily escape that framework.

Second half: Arsenal turn positional control into attacking control

The 45’–70’ period is the decisive phase of this Arsenal vs Chelsea tactical analysis.

Arsenal’s xT profile becomes stronger across Offensive Transition xT, the right corridor and the central corridor. Chelsea retain an advantage in Recovery Zone xT, but Arsenal are generating threat through more routes.

The movement map also becomes more compact through Arsenal’s midfield.

Martin Ødegaard shifts into a more central/right-half-space function, while Declan Rice operates slightly deeper around the central lane, giving Arsenal both connection and forward movement.

This is where the difference between a static position and a functional role becomes useful.

Arsenal’s midfielders are not merely shown where they pass from. Their reception arrows indicate where they repeatedly become available again after releasing the ball.

The team is using positional occupation to create movement rather than using movement without a stable structure underneath it.

Subscribe now

The winning goal comes from the opposite side

Arsenal’s second goal at 54 minutes is perhaps the clearest expression of their developing control.

Christos Tzolis progresses from the left side into a central attacking route and delivers the decisive pass into the box, where Martin Ødegaard finishes from the central area. The sequence is classified as a fast break and creates a major scoring opportunity.

Compare that with the first goal.

The equaliser had emerged from Arsenal’s right-sided progression.

The winner arrived after penetration from the other side into the centre.

That is why Arsenal’s growing territorial balance matters more than simply saying they attacked down both wings. Their structure allowed one corridor to move Chelsea and another to become the productive route.

The attack was becoming harder to predict.

Chelsea’s strongest structural response comes after falling behind

One of the most revealing details in the Network Health chart appears immediately after Arsenal take the lead.

Between roughly 55 and 60 minutes, Chelsea rise to 0.85 — their highest Network Health reading of the entire match.

That is a huge change.

Chelsea did not collapse structurally after going 2–1 down. In fact, they responded by creating their healthiest passing network of the game.

But Arsenal remained relatively stable themselves, with strong readings around the same period.

This tells us something important about the meaning of control.

A good structural window does not automatically produce a goal. Chelsea’s network became healthier, but they were now trying to solve an Arsenal structure that already had the advantage on the scoreboard and several effective routes for moving the ball away from pressure.

Chelsea improved the circulation.

Arsenal still controlled where much of that circulation could go.

Arsenal’s player roles reveal a clear hierarchy

The full-match player-role map helps explain why Arsenal’s passing network remained resilient.

Two players stand out for very different reasons.

Declan Rice is the strongest receiving influence in Arsenal’s network and also one of its most important flow connectors. He is not merely receiving possession frequently; the structure repeatedly routes meaningful circulation through him.

Christos Tzolis, meanwhile, records the strongest flow-connector profile in the match while also carrying significant receiving influence.

This gives Arsenal two powerful structural references operating in different parts of the pitch.

Rice stabilises and redirects central circulation.

Tzolis provides a high-value connection between possession and progression on the left side.

That distribution matters because Arsenal are not dependent on a single player performing both jobs in one zone.

Chelsea’s network was influential, but more distributed

Chelsea’s structure has its own strong hubs.

Wesley Fofana stands out as a major flow connector, while Cole Palmer has one of the highest receiving-influence profiles in the team.

This produces a different network shape.

Chelsea have important players occupying different functional roles, but the map suggests less concentration of both receiving influence and flow control in the same dominant points than Arsenal achieve through their strongest hubs.

That difference can matter when a game becomes difficult.

A team with clear structural anchors can often preserve its passing identity even as individual movements around those anchors change.

Arsenal did exactly that.

70’–100’: Chelsea gain central threat, Arsenal protect width

The final phase changes again.

Chelsea’s xT profile improves through the central corridor and Recovery Zone, while Arsenal’s strongest remaining territorial advantage shifts toward the left corridor.

That is consistent with a Chelsea side increasingly trying to attack through central access rather than building exclusively around the wings.

Their progression patterns reinforce this: when Chelsea begin through the left half-space, every identified sequence in this sample reaches the final third through the central corridor.

Chelsea are therefore trying to turn wide or half-space possession inward.

Arsenal’s defensive behaviour gives us the other side of that equation.

When Arsenal defend the left half-space, Chelsea frequently recirculate rather than continuing forward, reducing the rhythm of the attack.

That is an important piece of the late-game control.

Arsenal do not need to prevent Chelsea from possessing the ball everywhere. They need to stop the first route from becoming the next dangerous route.

The late movement maps show two different approaches

Arsenal’s final Functional Movement Map becomes much wider.

On the right corridor, Bukayo Saka remains a deep attacking reference and his reception region extends considerably beyond his median passing position.

On the opposite side, Arsenal maintain enough occupation to prevent the entire team from collapsing toward the ball.

This width stretches Chelsea’s defensive responsibility even when Arsenal are no longer dominating every xT category.

Chelsea’s late movement becomes more central.

Cole Palmer drops into a deeper central receiving role, while Reece James occupies an important central connection point.

Chelsea are trying to increase circulation around the middle and find the route into Arsenal’s defensive block.

The strategy improves their central threat.

But it also means Arsenal can increasingly identify where the attack is being funnelled.

Arsenal did not simply “sit back” after taking the lead

The Network Health data makes this particularly clear.

Arsenal remain strong through the 60’–80’ period, reaching 0.64, 0.65 and 0.66 in several windows.

Only late in the match does their structural health fall noticeably, including a drop to 0.28 around 90’–95’.

Chelsea, by contrast, regain stronger readings late and reach around 0.65 in the 85’–90’ window.

So the closing stages do show a genuine shift.

Chelsea are imposing more pressure.

Arsenal’s structure is becoming less comfortable.

But this happens after Arsenal have already accumulated the tactical advantage that matters most: they can now defend the lead while retaining enough width to escape pressure.

The late phase is therefore better described as Chelsea gaining initiative rather than Arsenal losing all control.

What Arsenal 2–1 Chelsea tells us about field control

The key lesson from this Arsenal 2–1 Chelsea Premier League tactical analysis is that control changed form several times.

Early in the match, Arsenal had structural control without scoreboard control.

Chelsea scored, but Arsenal possessed the healthier passing network.

By the middle of the first half, Arsenal converted that into territorial control, expanding from their strong right side toward central and left-sided progression.

After halftime, that became functional attacking control. Players were not merely occupying useful zones; they were receiving in new zones after releasing the ball, creating movement across corridors.

And after going 2–1 ahead, Arsenal gradually shifted toward game control: maintaining enough structural stability and width to prevent Chelsea’s improved circulation from becoming sustained high-quality access.

Chelsea had real moments.

They scored first, threatened in transition, hit the post before halftime and produced their best structural spell immediately after Arsenal’s second goal.

But Arsenal’s response was more complete.

They did not rely on one corridor, one player or one type of attack.

Their right side initially provided the strongest platform. Their left side became increasingly important. Their central network then connected both.

That is what ultimately separates the two teams in the data.

Arsenal 2–1 Chelsea: the tactical takeaway

The scoreline suggests a narrow Premier League win.

The underlying structure tells a richer story.

Arsenal won because their passing network could change shape without losing its main reference points.

The right corridor created the early platform.

Movement between corridors made the attack less predictable.

Central connections gave the team continuity.

And when Chelsea increased the pressure late, Arsenal still had enough width and structure to protect the advantage.

Chelsea were capable of disrupting the game.

Arsenal were better at reorganising it.

Download the Arsenal vs Chelsea match data

The analysis can continue beyond the article. Paid subscribers can download the complete Arsenal vs Chelsea match-event XLSX below the paywall, with event-level information that can be used for independent analysis, custom visualisations and deeper tactical exploration.

Read more

Data disclaimer: No match images or video footage were used in this analysis. The conclusions are based exclusively on event data, pass-network structure, Voronoi spatial control, Median xT Destination, corridor xT, Offensive Transition xT, Recovery Zone xT, Network Health, Match Diagnostics and the new Football Hacking Functional Movement Maps.

How Monaco turned PSG’s territorial control into a 2–1 Ligue 1 win

This PSG 1–2 Monaco tactical analysis is especially interesting because the result and the structural data do not tell exactly the same story. PSG produced several strong Network Health windows and often maintained a recognisable positional structure. Monaco were less stable by that measure, but their movement maps reveal greater functional displacement between where players passed the ball and where they subsequently received it.

That difference becomes increasingly important as the match develops.

The attached event dataset was also used to verify the match sequence alongside the visual models.

At the end of this analysis, paid subscribers will be able to download the match-event XLSX, making it possible to explore the underlying data independently.

0’–22.5’: Monaco showed more threat while PSG held a clearer positional skeleton

The opening xT radar already challenges a simple possession-based interpretation of the match. Monaco had the larger profile in Offensive Transition xT, Recovery Zone xT and central/right-corridor threat, while PSG’s shape was more conservative.

The pass network and Voronoi map show PSG maintaining a recognisable structure across the pitch, but the Functional Movement Map adds something the static network cannot.

On PSG’s left side and left half-space, Warren Zaïre-Emery repeatedly shifted between his median passing position and different reception zones rather than remaining fixed. Further forward, Khvicha Kvaratskhelia occupied a much more advanced left-sided position, giving PSG a stable high reference instead of constantly changing his functional location.

That is an important distinction: one player was creating movement around the structure, while the other was helping define the structure itself.

On Monaco’s right corridor, the movement of Vanderson was much more expansive. His reception region extended away from his median pass-origin position, while Sadibou Sané operated higher on the same side. Monaco therefore had a clearer capacity to change the location of an attack after the first pass rather than simply circulating inside fixed positions.

Their Match Diagnostics support that interpretation: when Monaco began sequences through the right corridor, 37.5% eventually entered the final third through the central corridor.

The right side was not necessarily the destination. It was an entry point.

Subscribe now

22.5’–halftime: PSG increased connectivity, but Monaco’s xT profile remained stronger

PSG’s structure changed noticeably during the second part of the first half.

The movement map shows more concentration around midfield, with Vitinha increasingly connected to the central zone and Désiré Doué operating between the central lane and right half-space. Their reception directions suggest more attempts to move away from the original pass location and create an additional line ahead of the ball.

The pass network for 25’–46’ reflects the same idea. PSG’s connections become denser and more interconnected around the centre, yet Monaco’s Median xT Destination radar remains larger in several dimensions, particularly transition and recovery-related threat.

This is where the scoreline becomes deceptive.

PSG scored after 30 minutes, but the JSON shows that Marquinhos’ goal came from a corner, with Kvaratskhelia providing the final ball into the central six-yard area.

That matters because the pass-network and Voronoi visualisations shown here explicitly exclude set pieces.

So PSG’s 1–0 lead did not necessarily validate their open-play structure. In fact, the open-play indicators were already suggesting that Monaco possessed routes capable of producing danger.

PSG’s Network Health looked good — but structural health is not scoreboard control

The full-match Network Health chart contains perhaps the most useful warning in this game.

PSG recorded a 0.78 Network Health value around 30’–35’, and later climbed to 0.73 around 55’–60’.

Yet Monaco equalised at 60’.

This is a perfect example of what Network Health should — and should not — be used for. It measures the quality and organisation of the passing structure in isolated windows. It does not mean the structurally healthier team will score the next goal.

PSG were still functioning as a network.

Monaco were finding ways to attack outside the logic of PSG’s structural superiority.

45’–67.5’: the match becomes more balanced — then Monaco finds the central finish

At the start of the second half, the xT profiles become much closer.

PSG improve their Offensive Transition xT and left-corridor profile, while Monaco continue to maintain meaningful central and recovery-zone threat. The Voronoi map becomes increasingly fragmented around midfield, which is consistent with a match becoming less territorially clean.

PSG’s movement also changes.

In the central corridor, Vitinha begins receiving in a different zone from his median pass origin, creating a more visible horizontal movement between the centre and the right side. Behind him, Willian Pacho shows a much longer functional displacement from his defensive passing position toward a wider/deeper reception region.

PSG were moving more, but their movement was also stretching the distances between structural points.

Monaco then punished them.

At 60’, Jordan Teze delivered from an advanced central area into the centre of the six-yard box, where Flávio Nazinho headed the equaliser. The event data identifies Teze’s pass as an intentional goal assist and a big-chance creation. The finish came from the central small-box area.

This was an open-play goal arriving directly into one of football’s highest-value spaces.

And unlike PSG’s set-piece opener, it connected much more directly with Monaco’s underlying open-play threat.

67.5’+: PSG stretch the field, Monaco attack the spaces created by that urgency

The final Functional Movement Map is arguably the most revealing image of the match.

PSG’s late structure is radically more stretched.

After substitutions, Mika Godts occupies a very high position on the left corridor, while Ousmane Dembélé provides width and depth on the opposite right side. Both entered around the 71st expanded minute.

The objective is easy to see even without video: increase the horizontal occupation of the pitch and create more reception possibilities around Monaco’s block.

The arrows become longer.

The team is no longer merely maintaining positional references; it is actively trying to move between them.

But stretching the attacking structure has a trade-off. Space becomes available elsewhere.

Monaco’s late movement map is more compact around its central and defensive structure, but it retains functional movement through the flanks and half-spaces rather than simply collapsing around the penalty area.

That helps explain the late xT radar.

From 70’ onward, Monaco again display a clear advantage in Offensive Transition xT, Recovery Zone xT and right-corridor xT. PSG, despite chasing the score, are stronger primarily through the left corridor.

In other words, Monaco did not merely defend the result. Their positional profile still contained mechanisms to escape pressure.

Monaco’s second goal shows the value of moving across corridors

The winner at 74’ gives the tactical picture a very concrete ending.

Aleksandr Golovin played forward from the central corridor into the left side of PSG’s penalty area, where Stanis Idumbo finished from the deep left side of the box.

That sequence is especially interesting alongside Monaco’s Match Diagnostics.

Across the match, when Monaco began through the right half-space, 66.7% of those sequences reached the final third through the left half-space.

That tells us something fundamental about Monaco’s attack: their progression was not tied to the corridor where it started.

They could enter through one side, use central connections, and arrive on the opposite side.

The movement maps make that behaviour visible in a way a static pass network cannot.

Why the Functional Movement Maps add another layer

This match is a good example of why pass networks and movement maps answer different questions.

A pass network tells us who connects with whom and where those connections occur.

The Functional Movement Map asks a different question: does a player tend to pass from the same area where he receives, or does his functional role require him to move into another zone before becoming available again?

Across the full match, PSG appear relatively more positional. Several players retain clearly defined zones, with movement occurring around a stable structure.

Monaco show more displacement between pass-origin and reception zones, especially through their wide and half-space players.

That does not automatically make one model better than the other.

But in PSG 1–2 Monaco, it helps explain how Monaco could look less dominant structurally while remaining difficult to control functionally.

Control of the field changed meaning after halftime

In the first half, PSG’s control was mainly positional.

They occupied predictable zones, maintained a connected structure and eventually took the lead.

In the second half, the game increasingly demanded functional control: the ability to follow changing reception zones, defend movements across corridors and manage transitions when the opponent escaped the initial pressure.

That is where Monaco became more effective.

The Match Diagnostics show that PSG were capable of disturbing Monaco when defending in the central midfield zone and right half-space: Monaco often responded by recirculating, producing a loss of rhythm.

But Monaco also forced similar behaviour from PSG when defending higher in the attacking-third central zone and left half-space.

Neither side simply controlled the ball and imposed itself continuously.

The battle was about where circulation could be converted into the next useful reception.

What PSG 1–2 Monaco tells us

The most interesting conclusion from this PSG vs Monaco Ligue 1 tactical analysis is that territorial occupation, structural health and functional movement are related — but they are not the same thing.

PSG often had the cleaner network.

Monaco often had the more disruptive movement.

PSG’s first goal came from a set piece. Monaco’s two goals came from open-play attacks entering the central six-yard area and the left side of the box.

And when PSG stretched the pitch late in search of a response, Monaco’s transition and recovery-zone xT remained strong enough to prevent the match from becoming one-way pressure.

The 2–1 result therefore looks less surprising when the pass network, xT, spatial control and functional movement maps are analysed together.

Download the PSG vs Monaco match data

Paid subscribers can continue below the paywall and download the complete PSG vs Monaco match-event XLSX used for this analysis. The spreadsheet provides the underlying event-level data so you can reproduce parts of the analysis, create your own visualisations or test entirely different tactical questions from the same Ligue 1 match.

Read more

São Paulo vs Atlético-MG presents a fascinating structural contrast. The data gives São Paulo the clearer overall edge, particularly because their possession sequences tend to preserve — and often increase — value after resistance. Atlético-MG can reach advanced areas, but their biggest weakness appears after pressure arrives: promising central progression too often turns into backward circulation and a reset into lower-value zones.

One of the most important tactical collisions is especially revealing. Atlético’s preferred final-third access runs through the central corridor — precisely where São Paulo show some of their strongest defensive suppression. Meanwhile, São Paulo have demonstrated the ability to escape pressure from deep zones and reach the opposite attacking side within five passes, turning resistance into genuine territorial and xT progression rather than mere survival.

The complete structural matchup is in the Football Hacking tactical intelligence graphic below, including progression routes, post-resistance behavior, support functions, structural risks, key adjustments and the main collision zones between the two teams.

For paid subscribers

Alongside the full tactical analysis, paid subscribers can also download the XLSX Poisson probability matrix for São Paulo vs Atlético-MG, generated from the Football Hacking models. You can use the matrix to explore the model probabilities yourself and build your own match and betting analysis rather than relying on a single predicted outcome.

Unlock the full Football Hacking newsletter and become a paid subscriber

Subscribe now

Explore the Football Hacking Web App:

The newsletter delivers the deeper tactical interpretation. The Web App gives you direct access to Football Hacking’s analytical tools, models and match data. If you want to go beyond scores and conventional football statistics and understand how teams progress, resist pressure, connect their passing structure and create — or destroy — value, that is exactly what Football Hacking is built for.

Read more

Football analytics should make video analysis more focused

Football Hacking was created around a simple idea: analysts, coaches and technical staff should not have to watch every minute of every match blindly, hoping that an important tactical pattern eventually reveals itself.

Video remains essential. No graph, model or football analytics platform can reproduce everything that happens on the pitch. The timing of a run, a player’s body orientation, the pressure applied by an opponent and the space created for a teammate often need to be seen.

But that does not mean the analyst should always begin with a blank page.

The more tactical information we can extract from football data before watching the match, the more intelligently we can use video. If a graph already indicates where a team may be building, which players connect different areas and where certain players tend to receive the ball, the analyst can concentrate on the deeper questions.

Why is that player moving there? What triggers the rotation? Which teammate occupies the space he leaves behind? Does the movement attract an opponent, create a passing lane or prepare the next progression?

This is the philosophy behind Football Hacking: use data to identify the questions worth investigating, then use video to understand the football behind the numbers.

Subscribe now

A new player movement visualization

With that objective in mind, I am currently developing and testing a new tactical movement graph for the Football Hacking web app.

The basic concept is straightforward.

Each player is positioned according to the average location from which he makes his passes. From that reference point, dotted arrows indicate the main areas in which he tends to receive the ball.

The arrows are not passing connections. They also do not claim to reproduce a player’s exact run from one point to another. Event data cannot provide the same level of physical movement information as optical or GPS tracking data.

Instead, the visualization compares two different positional signals:

• Where does the player usually release the ball?

• Where does he tend to appear when receiving it?

The difference between those locations may reveal functional movement: a defender stepping into midfield, a full-back moving inside, a midfielder dropping into the first line, a winger receiving in different corridors or an attacker changing height during possession.

It is a way of transforming passing and receiving locations into tactical hypotheses.

From a graph to a better video question

The image attached to this email is one of my current internal tests. The labels are still in Portuguese because this is an early prototype that I am validating before deciding whether it should become a new Football Hacking web app feature.

One possible interpretation immediately caught my attention.

In the section of the match represented by the graph, Arsenal appear to be building with three players in their first line.

Is Arsenal really forming a back three during that phase? Is one full-back staying deeper? Is a midfielder dropping between or alongside the central defenders? Is the apparent structure stable, or is it only the result of a few isolated possessions?

The graph cannot answer all of those questions by itself—and it should not pretend to.

What it can do is tell me exactly what to investigate when I open the match video.

Instead of watching a long sequence without a specific target, I can go directly to that period and verify whether Arsenal consistently use a three-player build-up structure. I can then examine how the opposition reacts, where the free player appears and what happens after the first line is broken.

That is the workflow I want Football Hacking to support:

Data identifies the possible pattern. Video confirms, rejects or explains it.

Even if the hypothesis turns out to be wrong, the visualization has still served a purpose. It has generated a precise tactical question that can be tested against reality.

Reducing screen time without reducing analytical depth

For analysts and coaching staffs, time is one of the most limited resources in football.

A professional may need to examine several matches from the next opponent, review his own team, prepare individual clips, evaluate set pieces and communicate the findings to coaches and players. An enthusiast may face a different schedule, but the challenge is similar: there is far more football available than anyone can examine deeply.

The objective is not necessarily to watch less football for its own sake. The objective is to avoid spending valuable time searching blindly for patterns that data could have highlighted first.

If an analyst needs to watch fewer matches—or watch the same match fewer times—to understand its basic structure, that time can be invested in details that demand human interpretation.

The analyst can focus on pressing triggers, cover shadows, body shape, synchronization, player intentions and the opponent’s responses. In other words, the elements that are difficult to capture with event data alone.

A useful football analytics tool should not compete with the analyst’s knowledge. It should make that knowledge more productive.

Subscribe now

Why this is still an experiment

I am not presenting this visualization as a finished metric or a confirmed web app feature.

I am currently testing different matches, teams, match periods and tactical contexts. I also need to compare what the graph suggests with what actually happened on video.

That validation is crucial.

A visually convincing graph can still produce a misleading interpretation if its assumptions are not tested carefully. Before adding this visualization to Football Hacking, I want to understand where it works, where it fails and how much confidence we can place in the patterns it highlights.

I am also evaluating how the graph responds to substitutions, red cards, extra time and changes in possession structure. Set-piece situations are excluded because corners, free kicks, throw-ins and goal kicks can distort the open-play positional picture.

The goal is not simply to create an attractive graphic. It is to create something that genuinely helps people understand a match more efficiently.

Building Football Hacking around better questions

This prototype represents the direction I want Football Hacking to continue exploring.

Football data should not merely describe how many passes a player completed or where he touched the ball. It should help us move closer to tactical interpretation.

Where does a player operate? Where does he reappear? How does his position change during possession? What structure might the team be creating around those movements? And, most importantly, what should we look for when we return to the video?

The answers will never come from one graph alone. But a strong visualization can reduce uncertainty, guide attention and allow analysts to begin their video work with better questions.

That is what I am testing now.

The attached image is only a prototype, and the Arsenal interpretation still needs to be checked against the match itself. I wanted to share the process because Football Hacking is not only about presenting finished results. It is also about developing new ways to connect football data, tactical analysis and reality on the pitch.

If the validation proves that this graph consistently produces useful and reliable insights, it may become one of the next features available inside the Football Hacking web app.

Data disclaimer: No images or video footage from the match were used in this analysis. The conclusions below are based exclusively on event data, pass-network structure, Voronoi field control, network-health windows, xT destination profiles, offensive-transition xT, recovery-zone xT and spatial progression patterns.

Arsenal’s structure changed the game without completely dominating it

Aston Villa 0–1 Arsenal in Premier League 2026/27 was not a simple story of one team controlling everything. Arsenal found the decisive structural advantage, while Villa remained competitive enough to make the final stages uncomfortable.

At the end of this analysis, the match event JSON will also be available for download, allowing subscribers to explore the underlying positional and event data themselves.

The overall Network Health chart immediately shows why the 0–1 scoreline makes sense. Aston Villa repeatedly produced healthy five-minute structural windows and actually finished the match with one of their strongest periods. Arsenal, meanwhile, did not maintain overwhelming structural superiority from beginning to end.

The difference was where Arsenal’s advantages appeared and what they did with them.

Subscribe now

First half: Arsenal built their clearest advantage on the right

During the opening 0’–25’ window, Arsenal’s Offensive Transition xT was clearly higher. Their structure also occupied important territory around the right side and right half-space.

This is where Bukayo Saka and Martin Ødegaard become particularly relevant. Saka occupied the advanced right corridor while Ødegaard operated more internally, creating a route from the right side into central attacking territory rather than relying exclusively on width.

That pattern is reinforced by the Match Diagnostics data: when Arsenal started progression through the right half-space, 80% of those sequences reached the final third through the central corridor. Starting from the right corridor itself, 42.9% progressed into the right half-space.

That is an important distinction. Arsenal were not simply moving the ball down the wing. They were using the wing as an entrance into more valuable central territory.

The JSON also confirms Saka’s early involvement on that side: within the opening minute he was already receiving and playing passes from the right zone.

Aston Villa were competitive — and dangerous

Villa’s structure should not be interpreted as passive.

Their Network Health remained competitive throughout much of the first half, and their xT profile became much closer to Arsenal’s between roughly 25’ and halftime. Their central-corridor xT was particularly relevant during this phase.

Emiliano Buendía and Ian Maatsen illustrate the threat better than a long list of players would. Buendía hit the post from a central area in the 11th minute, while Maatsen later generated another attempt from an advanced area in the 26th minute.

So this was not Arsenal simply pushing Villa backwards.

The contest was much more balanced spatially: Villa could reach dangerous territory, but Arsenal were better at turning their preferred progression route into a repeatable structure.

The second-half change: Arsenal increased control around recovery and progression

The most important tactical shift appears in the 45’–70’ window.

Arsenal’s Recovery Zone xT rises considerably above Villa’s, while their left-corridor xT also improves. In other words, Arsenal became less dependent on a single attacking route.

Their Voronoi map reflects the same transition. The team occupied a broader central platform while retaining access to both sides, making possession losses less destructive because players were already positioned around the next interaction.

This is one of those situations where “control” does not necessarily mean endless possession. It means reducing the cost of losing the ball.

And this period produced the decisive moment.

The winning goal was the final expression of Arsenal’s spatial control

Arsenal scored around the 60th-minute mark through Bukayo Saka, but the event sequence is more interesting than the name of the scorer.

Immediately before the finish, Riccardo Calafiori received in an extremely advanced central position and played the intentional assist into the centre of the six-yard area. Saka finished first time with his right foot. The data classifies the chance as a big chance and the pass as an intentional goal assist.

That matters because the goal did not emerge from an isolated long-range action.

It ended with Arsenal occupying the most valuable central space possible.

The pass network had been progressively bringing Arsenal from wider and half-space structures toward the centre; the goal converted that spatial principle into the scoreboard.

Subscribe now

Villa responded after going behind

Aston Villa did not collapse after the goal.

Their strongest attacking spell actually came soon afterward. Matty Cash forced a save from inside the box around 64’, while Ross Barkley produced another saved attempt only a few minutes later.

This coincides with Villa’s stronger Network Health periods around the final third of the match.

There was also a clear structural intervention around the 72nd minute: Tammy Abraham replaced Nicolas Jackson, while another midfield change was followed by a recorded formation change.

Villa consequently became more direct and vertically stretched.

That change is visible in the 70’–105’ network. Their players push into higher positions, while Arsenal’s structure becomes more compact around midfield and its own half.

Arsenal changed from attacking control to game control

This is perhaps the most interesting difference between the two halves.

Early Arsenal control was about progressing into advantageous spaces.

Late Arsenal control was about protecting those spaces.

The final xT radar still gives Arsenal an advantage in Recovery Zone xT, but the overall shapes become considerably closer. Villa’s Network Health even reaches 0.79 in the final 100’–105’ window, its highest value of the match.

Yet Arsenal’s late network becomes denser centrally rather than disintegrating.

That is significant. The team no longer needed to reproduce the same attacking structure that created the goal. It needed to keep Villa’s attacks from turning territorial pressure into repeated high-value penetration.

And that is ultimately why Aston Villa 0–1 Arsenal looks different through structural data than through the scoreline alone.

Villa remained alive. Arsenal remained organised.

What the data says about Aston Villa 0–1 Arsenal

The key takeaway is not that Arsenal completely dominated Aston Villa.

They didn’t.

It is that Arsenal produced the more useful form of control at the decisive moments.

Their right-sided progression repeatedly connected the right corridor and right half-space with central territory. Their recovery structure strengthened after halftime. And when the opportunity finally emerged, Calafiori and Saka transformed advanced central occupation into the only goal of the match.

Villa subsequently changed behaviour and pushed harder, but Arsenal successfully changed with them.

The result was therefore less about permanent superiority and more about structural adaptability.

Download the match data

The analysis does not have to end here. Paid subscribers can download the Aston Villa vs Arsenal match JSON below, containing the underlying event-level data used alongside the Football Hacking structural models. Use it to reproduce the analysis, build your own visualisations or test completely different questions against the same match.

Read more

Disclaimer: No match images or video footage were used in this analysis. The conclusions are based exclusively on event data, pass-network structure, Voronoi spatial control, Median xT Destination, corridor xT, Network Health and Football Hacking Match Diagnostics.

Chelsea’s direct efficiency overcame Brighton’s territorial circulation in a match that changed shape repeatedly

Chelsea’s 4–3 win over Brighton in Premier League 2026/27 Round 2 is a good example of why territorial control, passing control and attacking efficiency are not the same thing.

Brighton frequently established the more elaborate circulation structure, particularly through central areas and the left half-space. Chelsea, however, repeatedly found ways to turn much shorter periods of control into decisive attacking actions.

The match-event data reinforces that contrast. Chelsea were already 2–0 ahead by the 13th minute, with goals from Roméo Lavia and Pedro Neto, before João Pedro made it 3–0 at 31 minutes. Brighton responded through Malick Yalcouyé at 34’.

Subscribe now

0’–25’: Brighton had structure, Chelsea had damage

The opening network already explains much of the game.

Brighton occupied a broad central structure with multiple connections around midfield, while Chelsea were less interested in building a perfectly symmetrical possession network. The Blues’ attacks were more vertical and selective.

The key areas for Chelsea were the central channel and the right side of the attacking structure, where Cole Palmer and Malo Gusto helped stretch Brighton’s defensive occupation. Chelsea did not need permanent territorial domination to create danger.

That difference matters because the Network Health chart shows something more nuanced than the score. Brighton still recorded substantial structural health during several opening windows despite conceding early. Chelsea’s effectiveness came from converting favourable moments rather than simply owning every phase.

In other words: Brighton could have the ball structure without having control of the consequences.

25’–45’: Chelsea’s right side becomes much more threatening

The second window produces one of the clearest changes in the entire match.

Chelsea’s right-corridor xT expands sharply, while the left corridor also remains productive. The passing map simultaneously shows a wider attacking occupation.

This is where Jorrel Hato and João Pedro become especially relevant. At 31’, the event data records Hato’s successful central pass as part of a FastBreak, immediately preceding João Pedro’s goal from the central area of the box.

That sequence captures Chelsea’s first-half approach almost perfectly: the team did not require a long possession chain to manufacture high-value progression.

Brighton, meanwhile, continued to find routes through midfield and the left side. The Match Diagnostics identifies a particularly clear tendency: when Brighton started through the central corridor, 60% of their final-third progressions arrived through the left half-space, with the Ferdi Kadioglu–Lewis Dunk connection highlighted as the main pathway.

So Brighton were not without a plan. The problem was that Chelsea were converting transitions faster than Brighton could turn territorial organisation into defensive security.

45’–70’: the match becomes less stable

The beginning of the second half changes the geometry.

Chelsea no longer appear as consistently expansive as they did late in the first half. Brighton occupy larger areas of the pitch and their network becomes more capable of sustaining possession around midfield.

Yet this is also where the Network Health chart needs to be interpreted carefully.

Chelsea reach an exceptional 0.99 Network Health reading in the 60’–65’ window, but the next major scoring event is actually damaging to them: the match file identifies the 65’ event as a João Pedro own goal from a corner, reducing Chelsea’s advantage.

That is an important reminder of what Network Health represents. It measures structural conditions; it does not claim that the structurally healthier team must score next.

Football can punish a team during a perfectly reasonable structural phase.

At 3–2, the game had suddenly reopened.

70’–100’: Chelsea change the route to goal

The final xT radar is perhaps the most revealing of the four windows.

Chelsea’s Offensive Transition xT rises strongly, while the central corridor becomes the dominant source of destination value. This is different from the stronger right-corridor emphasis seen earlier.

The passing network also becomes more selective. Chelsea are no longer trying to occupy every attacking lane with the same intensity. They increasingly look for the moment when a wide or half-space action can be converted into a central attack.

The fourth goal is a perfect data-level example.

At 76’, João Pedro plays from the right-side zone before Cole Palmer scores from the central area outside the box. The event file records the pass as a key pass and Palmer’s finish as an OutOfBoxCentre goal.

That sequence matches what the radar is showing: progression from a lateral area, value concentrated centrally.

Brighton still refused to disappear. Pascal Groß eventually scored with a header from the centre of the box at 98’, bringing the final score to 4–3.

Spatial control explains why this was never a simple Chelsea domination

The Voronoi maps are probably the best antidote to reading 4–3 as a conventional home-team victory.

Chelsea repeatedly controlled strategically useful pieces of the pitch without controlling the entire pitch.

Brighton often retained meaningful territorial presence around the centre and their build-up zones. Chelsea were more dangerous when the structure opened and the game could be moved rapidly toward the final third.

The defensive diagnostics support that interpretation. When Chelsea defended the central zone of their defensive third, Brighton frequently reacted neutrally and lost rhythm. When Chelsea defended the left side of their defensive third, Brighton tended toward recirculation, again reducing attacking rhythm.

That means Chelsea’s control was partly behavioural rather than purely territorial: they could allow Brighton to retain circulation while influencing where that circulation eventually went.

What Chelsea 4–3 Brighton tells us

This was not a match where one team simply controlled possession and therefore controlled the game.

Brighton were capable of sustaining an interconnected passing structure and repeatedly entering advanced areas through recognisable routes. Chelsea were much more opportunistic.

Across the approximately 20-minute windows, the progression is clear:

early efficiency → stronger right-side threat → second-half instability → greater transition and central-corridor value late in the game.

That changing attacking profile is probably the most important tactical feature of Chelsea’s performance.

Chelsea did not win because their network looked superior in every phase. They won because their structure could change its route to danger as the match changed around it.

Paid subscribers can continue below the paywall and download the complete Chelsea vs Brighton match-events xlsx used in this analysis. The file contains the underlying event-level data, giving you the possibility to reproduce the analysis, build your own metrics or explore entirely different questions from the same match.

Read more