Corner kicks are changing.
For years, they occupied an awkward position in football analytics: frequent enough to attract attention, but individually inefficient enough to be treated as a relatively minor part of attacking performance. That interpretation is becoming increasingly difficult to defend.
Sportradar’s Turning the Corner into Goals provides compelling evidence that elite European football is moving toward a new set-piece environment. Based on ten seasons of data across the Premier League, Bundesliga, LaLiga, Ligue 1 and Serie A, Sportradar reports that corner conversion reached its highest level of the decade during the 2025/26 season. The company also identifies another important tactical development: the rapid expansion of the inswinging corner.
That observation immediately raises a second question.
Why?
Why should a ball curving toward goal produce a systematically different attacking environment from one curving away from it?
Football Hacking approached the problem from a spatial perspective.
Rather than attempting to reproduce or audit Sportradar’s research, this analysis takes its findings as the starting point and asks a different question:
What does the spatial distribution of the resulting shots tell us about the tactical difference between inswinging and outswinging corners?
Using 16,724 corners from the 2025/26 Premier League, Bundesliga, LaLiga, Ligue 1 and Serie A contained in the Football Hacking dataset, we classified deliveries where foot information was available and examined the location of the immediate shot generated from the corner.
The resulting picture is fascinating.
Inswingers do not merely bend in a different direction.
They appear to create a different geometry of finishing.
And Football Hacking believes that the explanation may lie partly in something football data often struggles to measure directly: defensive body orientation, visual information and the biomechanics of reacting to a ball that is rotating toward goal.
First, a thank you to Sportradar
This study originated from research shared with Football Hacking by Sportradar.
Alex Bake, Senior Communications Manager at Sportradar, kindly provided Turning the Corner into Goals, a detailed analysis of how corner-kick efficiency has evolved across European football.
We would like to thank Sportradar and Alex Bake for making the research available to Football Hacking.
Their work provided the motivation for this analysis.
Importantly, the two datasets and analyses should not be confused.
Whenever this article refers to historical trends, decade-long conversion rates, league-wide corner efficiency, team rankings or the evolution in the use of inswinging deliveries, those figures come from Sportradar’s research.
The spatial analysis of shot locations, footedness and immediate corner-generated finishing locations presented below comes from Football Hacking’s own 2025/26 event dataset.
The objective is therefore not replication.
It is extension.
Sportradar identified a major tactical trend.
Football Hacking wanted to investigate the geometry underneath it.
The corner-kick revolution identified by Sportradar
The historical context matters.
According to Sportradar’s analysis of the Big Five European leagues, teams historically converted approximately one corner in every 28.
During 2025/26, that improved to approximately one in every 23.
Sportradar reports an overall Big Five corner success rate of approximately 4.4% in 2025/26, compared with 3.8% in each of the three previous seasons. The 2025/26 figure represents the highest corner-conversion rate in the decade covered by the research.
This is not simply statistical trivia.
A move from one goal every 28 corners to one every 23 fundamentally changes the expected value of a set piece that occurs repeatedly throughout a season.
But perhaps the most interesting development identified by Sportradar is not simply that corners have become more productive.
It is how teams are delivering them.
The rise of the inswinging corner
For most of the decade covered by Sportradar, inswingers and outswingers existed in relatively similar proportions.
Sportradar’s season-by-season analysis shows inswinging deliveries generally hovering close to 50% of corners for much of the period.
Then something changed.
The inswinger share climbed to approximately 60% in 2024/25 and 65.5% in 2025/26.
The shift was particularly pronounced in England and Germany.
According to Sportradar, approximately 80% of Premier League corners in 2025/26 were inswinging, while the Bundesliga was around 70%. Before 2023/24, both competitions had been much closer to an even split between inswinging and outswinging deliveries.
Sportradar also reports that inswingers generated a higher success rate than outswingers across the Big Five leagues in 2025/26.
This creates an intriguing tactical relationship:
teams are using inswingers more often at the same time that inswingers are producing better outcomes.
That is where our investigation begins.
Football Hacking methodology
The Football Hacking dataset contains 16,724 corner kicks from the 2025/26 Big Five league sample:
LeagueCorners in Football Hacking datasetPremier League3,786LaLiga3,674Serie A3,349Bundesliga2,985Ligue 12,930Overall16,724
Each corner contains spatial information including the location of the delivery, its destination coordinates, angle and—when present—the foot used by the corner taker.
For immediate resulting shots, the dataset also provides the finishing coordinates.
This allows us to move beyond the binary question of whether a corner produced a shot.
We can ask:
Where did that shot happen?
How Football Hacking classified inswingers and outswingers
There is an important methodological limitation that should be stated explicitly.
Our event data does not contain a direct field saying:
Inswinger
or
Outswinger.
We therefore use an operational classification based on corner side and kicking foot.
A delivery taken with the foot that curves the ball toward goal is classified as an inswinger; the opposite combination is classified as an outswinger.
This is conceptually similar to the methodology described by Sportradar, whose report states that corner type was identified using the location of the corner and the preferred foot of the taker.
Football Hacking, however, uses the foot information contained directly in the event qualifiers rather than assigning a preferred foot externally.
Consequently, we deliberately leave deliveries without usable foot information as Unknown rather than inferring a category.
The resulting classification is:
Corner typeCornersShare of total datasetInswinger6,58239.4%Outswinger3,59721.5%Unknown6,54539.1%
Among the 10,179 corners that can actually be classified, the split becomes:
Inswinger: 64.7%
Outswinger: 35.3%
Interestingly, that 64.7% classified share sits extremely close to Sportradar’s independently reported 65.5% inswinger share for the Big Five in 2025/26.
This should not be interpreted as a formal validation exercise—the underlying providers, definitions and classification procedures are not identical—but it provides useful context for the spatial sample.
What exactly counts as a shot in the Football Hacking analysis?
For this particular spatial study, we intentionally use a restrictive definition.
We identify the corner event and examine the immediate subsequent event when it is explicitly marked as originating from the corner and is a finishing event.
The finishing categories are:
Goal
Saved shot
Missed shot
Shot on post
We are therefore interested in the direct attacking geometry produced by the delivery, not extended second phases occurring many seconds later.
That distinction matters.
Sportradar’s published corner-success metrics and the Football Hacking immediate-shot sample should not be treated as interchangeable conversion-rate definitions.
Sportradar provides the macro picture.
Football Hacking is isolating a narrower event chain in order to investigate spatial mechanics.
Overall: inswingers and outswingers create very different shots
This is where the analysis becomes particularly interesting.
Among classified corners, Football Hacking recorded:
DeliveryCornersImmediate shotsImmediate goalsInswinger6,58249149Outswinger3,59753235
At first sight, this appears counterintuitive.
Outswingers produced an immediate shot more frequently in our restrictive event definition:
Inswinger: 7.46%
Outswinger: 14.79%
But stop there and we miss the most important information.
The shots are not equivalent.
Once an inswinger produces the immediate finishing action, the resulting shot tends to occur materially closer to goal.
And that spatial distinction is substantial.
The critical spatial result
Across the Big Five sample, the average finishing coordinates were:
Delivery / footShotsMean shot XMean shot YMean distance to goalInswinger — left foot23193.4549.848.97 mInswinger — right foot26093.1250.909.24 mOutswinger — left foot19590.0850.8611.69 mOutswinger — right foot33790.0049.5811.89 m
The lateral coordinates are remarkably similar.
All four groups concentrate around the central axis of the goal.
The major difference occurs longitudinally.
Inswinging corners move the finishing event toward the goal.
Left-footed inswingers generated shots from an average estimated distance of approximately 9.0 metres.
Right-footed inswingers: approximately 9.2 metres.
The corresponding outswingers produced shots from approximately 11.7 and 11.9 metres.
That is not a cosmetic difference.
In a penalty area packed with defenders, moving the average finishing action two to three metres closer to goal changes the entire defensive problem.
The six-yard box reveals the difference even more clearly
The strongest spatial evidence appears when we isolate the six-yard area.
Approximately:
40.3% of shots following left-footed inswingers occurred inside the six-yard box.
40.8% following right-footed inswingers occurred there.
Compare that with outswingers:
9.2% for left-footed outswingers
8.9% for right-footed outswingers
That is an enormous structural difference.
Roughly four out of every ten immediate shots generated by an inswinger in our sample occurred inside the six-yard box.
For outswingers, it was fewer than one in ten.
This may be the most important Football Hacking finding in the entire analysis.
Because it suggests that asking whether inswingers “create more shots” is potentially the wrong question.
A better question is:
What kind of shot does each delivery create?
The spatial answer is clear.
Inswingers disproportionately create finishing situations immediately around the goal.
Outswingers generate their immediate shots from deeper positions.
Left-footed inswingers
The left-footed inswinger produced 231 immediate shots in our classified sample.
Its average shot position was:
X = 93.45
Y = 49.84
That Y-coordinate is almost perfectly central.
The average estimated shooting distance was approximately 8.97 metres, and 40.3% of these shots occurred within the six-yard box.
More interesting still is the finishing efficiency of the immediate shots.
Left-footed inswingers generated:
32 goals from 231 shots — 13.85% shot conversion.
That is the highest shot-to-goal rate among the four foot/swing combinations in this sample.
The heat map should therefore be read not simply as a density map.
It is a representation of how an inward-curving ball can compress the decisive action toward the goalkeeper and goal line.
Right-footed inswingers
The right-footed version produces an extraordinarily similar spatial signature.
From 260 immediate shots:
Mean X = 93.12
Mean Y = 50.90
Mean estimated distance = 9.24 m
Six-yard-box share = 40.8%
This symmetry matters.
If the inswinger effect were overwhelmingly associated with one foot, we would have to consider whether particular takers, teams or tactical structures were driving the overall pattern.
Instead, the spatial behaviour appears on both sides of the pitch.
The ball changes direction.
The foot changes.
The side changes.
But the geometry survives.
The inward-curving delivery repeatedly moves the finishing event toward the goalmouth.
Left-footed outswingers
Left-footed outswingers generated 195 immediate shots.
Their average position:
X = 90.08
Y = 50.86
Again, the lateral position remains central.
But longitudinally the shot moves backwards.
Average estimated distance rises to 11.69 metres, while only 9.2% of shots occur inside the six-yard box.
The contrast with the left-footed inswinger is striking:
40.3% vs 9.2% inside the six-yard area.
Same foot.
Different side.
Different ball rotation.
Completely different penetration into the goalmouth.
Right-footed outswingers
The largest immediate-shot group in our classified sample consisted of 337 right-footed outswinger shots.
Their average location:
X = 90.00
Y = 49.58
Average estimated distance:
11.89 metres.
Six-yard-box share:
8.9%.
Once again, the spatial profile closely resembles the opposite-foot outswinger.
This strengthens the broader interpretation.
The most important explanatory variable does not appear to be simply left foot versus right foot.
It is the relationship between:
foot + corner side + direction of curvature.
The Football Hacking hypothesis: why the inswinger is so difficult to defend
Numbers describe the outcome.
They do not automatically explain the mechanism.
This is where Football Hacking proposes a tactical hypothesis.
We believe the greater danger of the inswinger may be partly explained by the interaction between three factors:
body orientation, visual field and support-foot rotation.
Consider the defender’s problem.
Defending an outswinger
An outswinging corner initially travels away from goal.
The defender can often adopt an orientation that gives access to two critical pieces of visual information:
the trajectory of the ball;
the movement of the attacker.
The ball is moving into space rather than aggressively toward the goalmouth.
This does not make the situation easy.
But it can make the visual problem more manageable.
The defender may be able to maintain a partially open body shape, track the incoming delivery and preserve peripheral information about the opponent.
The movement of the ball is, in a sense, helping maintain separation between the delivery and the goal.
The inswinger creates a different visual problem
Now reverse the rotation.
The ball bends toward the goal.
The defender must process:
the ball;
the attacker;
the goal;
the goalkeeper;
potential blockers;
the expected contact point.
But the most dangerous movement is now occurring behind or across parts of the defender’s natural visual field.
This can force a choice.
Watch the ball and lose information about the runner.
Or:
track the runner and temporarily lose precise information about the ball.
Elite defenders obviously attempt to manage both.
But human vision has constraints.
And football happens too quickly for those constraints to be irrelevant.
The body-orientation problem
There is another component: the defender’s feet.
When defending an inward-curving delivery, the defender may have to adjust his body while simultaneously responding to a trajectory that is moving toward his own goal.
A defender initially positioned to observe the corner taker may need to rotate his hips and shoulders as the ball travels.
If an attacker changes direction during that movement, the defender may need another adjustment.
Every additional rotation has a cost.
The defender’s support foot must plant.
The hips must rotate.
The centre of mass must move.
The head must relocate to recover visual information.
Then the defender must jump, accelerate or challenge.
The attacker has an important informational advantage:
he knows the intended routine.
The defender does not.
The attacker may know approximately where the delivery is supposed to arrive.
The defender has to infer it in real time.
Why rotation of the support base matters
Think of defending a corner not merely as running and jumping, but as a sequence of biomechanical commitments.
Once a defender plants a foot to rotate toward one direction, immediately reversing that movement is costly.
A well-designed inswinging routine can exploit exactly that moment.
The attacking player can move while the defender is reorganising his support base.
The difference may be fractions of a second.
But fractions of a second inside the six-yard box are enormous.
This offers a plausible tactical interpretation of our spatial finding.
Inswingers may not need to generate more immediate shots to be more threatening. They may generate more invasive shots.
The Football Hacking data shows exactly that spatial signature.
Premier League: the most aggressive adopter
Sportradar identifies the Premier League as one of the strongest adopters of the inswinging corner, with roughly 80% of deliveries classified as inswingers during 2025/26. The league also recorded the highest overall corner efficiency among the Big Five in Sportradar’s research.
The Football Hacking spatial data is consistent with an extremely aggressive goalmouth profile.
Premier League immediate finishing locations
TypeShotsMean XMean distanceSix-yard shareInswinger — left6392.779.72 m50.8%Inswinger — right7392.669.62 m49.3%Outswinger — left1291.0711.38 m16.7%Outswinger — right5889.2212.41 m10.3%
The headline number is extraordinary:
approximately half of the immediate shots produced by classified inswingers occurred inside the six-yard box.
This is exactly the type of spatial behaviour we would expect from a competition that has increasingly professionalised set-piece coaching.
Sportradar’s broader analysis also demonstrates the competitive significance of corners in England. Its research identifies Arsenal as Europe’s leading overall corner-efficiency team and Tottenham Hotspur as the most efficient attacking corner team.
The tactical arms race is already visible.
Bundesliga: perhaps the strongest spatial inswinger signature
Germany produces an even more dramatic spatial pattern.
Bundesliga
TypeShotsMean XMean distanceSix-yard shareInswinger — left5394.268.03 m45.3%Inswinger — right4495.227.66 m68.2%Outswinger — left3391.0310.56 m12.1%Outswinger — right6689.1112.51 m4.6%
The right-footed inswinger is particularly striking.
Its average finishing coordinate is X = 95.22.
Almost seven out of ten immediate shots in that category occurred inside the six-yard area.
Average estimated shooting distance was only 7.66 metres.
This is not merely a difference in conversion.
It is a different attacking territory.
Sportradar independently identifies the Bundesliga as another major adopter of inswinging corners and reports that the competition had the second-highest corner conversion rate in 2025/26.
Germany therefore deserves particular attention when discussing the tactical evolution of set pieces.
LaLiga: strong inswinger penetration, but a more balanced delivery ecosystem
LaLiga produces a slightly different profile.
LaLiga
TypeShotsMean XMean distanceSix-yard shareInswinger — left3893.388.92 m42.1%Inswinger — right4293.558.70 m33.3%Outswinger — left4190.5610.83 m2.4%Outswinger — right6691.0411.06 m7.6%
Again, the key distinction is not necessarily lateral.
It is depth.
The inswinging groups produce shots approximately around X=93–94.
The outswinging groups sit around X=90–91.
The left-footed outswinger is particularly revealing: only 2.4% of its immediate shots occurred inside the six-yard area.
Sportradar characterises LaLiga somewhat differently from England and Germany: its research highlights strong defensive corner performance among leading Spanish clubs while also identifying room for greater attacking corner optimisation.
Our spatial data suggests that the inswinger still creates substantially deeper finishing positions when it produces the immediate shot.
Serie A: inswingers produce the better finishing geometry
Italy follows the broader pattern.
Serie A
TypeShotsMean XMean distanceSix-yard shareInswinger — left4493.768.77 m27.3%Inswinger — right4793.488.88 m29.8%Outswinger — left5189.5912.24 m7.8%Outswinger — right5990.7311.04 m10.2%
The difference in average shooting distance is again considerable.
The left-foot comparison is particularly pronounced:
8.77 m for the inswinger versus 12.24 m for the outswinger.
Sportradar’s league-level assessment describes Serie A as sitting around the middle of the Big Five in overall corner conversion and notes that the league still appears to have room for greater set-piece optimisation.
From a Football Hacking perspective, the spatial incentive for inward-curving delivery is nevertheless visible.
Ligue 1: the exception that helps explain the rule
France provides perhaps the most useful counterpoint.
Ligue 1
TypeShotsMean XMean distanceSix-yard shareInswinger — left3393.159.37 m27.3%Inswinger — right5491.3910.72 m22.2%Outswinger — left5889.4212.53 m12.1%Outswinger — right8889.9012.26 m11.4%
The inswinger advantage in shot depth remains.
But it is weaker than in England or Germany.
That matters because tactical effects are rarely universal constants.
League environments differ.
Defensive structures differ.
Player profiles differ.
Delivery quality differs.
Set-piece coaching differs.
And Sportradar’s own league summary describes Ligue 1 as displaying greater variation between clubs and lower attacking corner conversion than the leading competitions.
The French sample therefore reinforces an important principle:
“Inswinger” is not a magic word.
Execution still matters.
An important paradox: more shots versus better shots
The Football Hacking dataset produces a result worth examining carefully.
Outswingers generated immediate shots more frequently in our restrictive event chain.
Yet inswingers generated those shots much closer to goal.
This distinction may reconcile two apparently contradictory observations.
A delivery can increase the probability of producing some kind of shot without necessarily producing the most dangerous shot.
An outswinger naturally travels away from the goalkeeper and goal line.
That trajectory can facilitate contacts around the penalty spot or deeper central areas.
Those situations may be easier to convert into an identifiable finishing event.
But the attacker is farther from goal.
The inswinger does something else.
It attacks the space between the defensive line and goalkeeper.
That space is more difficult to access.
But when the attacking team wins it, the reward is potentially much larger.
Shot volume is not shot quality
This principle extends far beyond corners.
Football analytics has spent years learning that counting shots without considering their locations can be misleading.
Twenty speculative attempts from outside the penalty area are not necessarily superior to eight shots around the six-yard box.
The same logic should be applied to set pieces.
Simply asking:
“How many corners resulted in shots?”
is insufficient.
We also need:
“Where were those shots taken?”
Our results indicate that the second question fundamentally changes the interpretation.
The left-footed inswinger deserves special attention
Across all five leagues, left-footed inswingers generated:
231 immediate shots and 32 goals.
That means 13.85% of those immediate shots became goals.
For comparison:
right-footed inswinger: 6.54%
left-footed outswinger: 6.67%
right-footed outswinger: 6.53%
This result needs to be treated cautiously because sample sizes are much smaller than the full corner sample and team/taker effects are not controlled here.
But it is sufficiently large to justify further investigation.
Is this a tactical effect?
Is it driven by particular elite left-footed takers?
Are defensive systems less comfortable dealing with the corresponding trajectory?
Does the left-footed inswinger more frequently reach a specific goalmouth zone?
Those questions deserve a dedicated model rather than an immediate causal conclusion.
The central corridor is universal — depth is the differentiator
One of the most useful results is actually what does not change.
Roughly 90% of immediate shots in every major delivery group occurred within our central goal-facing corridor.
That means inswingers and outswingers are not primarily differentiated by one producing central shots and the other producing wide shots.
Both generally seek central finishing areas.
The difference is how deep into those central areas they penetrate.
This is crucial.
The tactical contrast is less:
central versus lateral
and more:
goal-line proximity versus penalty-spot depth.
The heat maps make this immediately visible.
Inswingers concentrate their strongest finishing density closer to the goalmouth.
Outswingers pull the distribution backwards.
The goalkeeper’s problem
The defender is not the only player affected by the direction of the ball.
The goalkeeper faces a fundamentally different decision tree.
With an outswinger, the ball’s natural trajectory is moving away from goal.
With an inswinger, the goalkeeper must evaluate a ball that is:
entering the goalmouth;
potentially reachable;
potentially screened;
surrounded by attackers and defenders;
capable of being redirected;
and potentially dangerous even without a clean attacking contact.
This creates uncertainty.
Should the goalkeeper attack the delivery?
Stay?
Move toward the flight and then recover?
How does he react when a runner crosses his visual path?
The more aggressively the ball enters the six-yard area, the less time exists for that decision.
Our spatial data cannot directly measure goalkeeper uncertainty.
But it can measure the environment in which that uncertainty occurs.
And the inswinger is producing far more immediate shots inside precisely that environment.
The attack knows; the defence reacts
There is another asymmetry that should not be underestimated.
The attacking team knows the routine.
The defending team can prepare for tendencies, but it does not know the exact execution.
The attacker may know:
intended delivery zone;
blocking pattern;
first movement;
second movement;
target player;
expected flick;
timing of the run.
The defender has to recognise those signals while simultaneously tracking the ball.
This is why body orientation matters so much.
The attacker can initiate movement based on predetermined information.
The defender must perceive, interpret and respond.
An inswinging ball compresses the available time and space for that process.
What Sportradar’s team results tell us about specialisation
Sportradar’s research provides another important piece of context.
Its attacking-efficiency rankings identify Tottenham Hotspur as Europe’s most efficient attacking corner team in the 2025/26 Big Five sample, while Freiburg, Arsenal, Bayern Munich and Inter Milan also ranked among the leading teams.
Sportradar also highlights specialist corner takers and argues that delivery quality remains important even as tactical structures become increasingly sophisticated.
This is an essential qualification.
A tactical concept does not execute itself.
“Take an inswinger” is not a strategy.
Delivery speed matters.
Height matters.
Trajectory matters.
Target zone matters.
Blocking matters.
Movement matters.
Timing matters.
The preferred solution is becoming clearer, but the execution remains highly specialised.
Corners are now affecting league tables
Sportradar’s most powerful argument may actually have little to do with delivery geometry.
It is about points.
Their research evaluates match-defining corner goals and estimates how league standings would have changed without them.
According to Sportradar, Arsenal gained 14 points through match-defining corner goals during their 2025/26 Premier League title-winning campaign. Their analysis concludes that without those goals, Manchester City would have won the league.
Tottenham’s corner goals also contributed important points during their fight against relegation.
Similar effects appear elsewhere in Europe, although Sportradar notes that the impact was less pronounced in Ligue 1 and Serie A than in England, Germany and Spain.
This changes how corners should be conceptualised.
They are not isolated micro-events.
Accumulated over 38 matches, small improvements in corner efficiency can become league-position variables.
Corners and overall attacking environments
Sportradar also identifies a relationship between corner efficiency and overall attacking output.
Its analysis finds that leagues generating more goals from corners also tended to generate more goals overall, with Bundesliga and Premier League teams occupying the higher end of both measures during 2025/26.
The report correctly warns that this does not establish causality.
High corner efficiency may be one expression of a broader attacking environment rather than the cause of it.
That distinction is important.
Strong teams may simultaneously have:
better corner takers;
better aerial players;
better coaches;
more attacking possession;
more corners;
better general attacking structures.
Separating those effects requires substantially more modelling.
Why the 2026 World Cup matters
The trend extends beyond club football.
According to Sportradar, the expanded 2026 FIFA World Cup produced 40 goals from corners, compared with 17 at the 2022 tournament.
The corresponding corner success rate rose from approximately 3.0% in 2022 to 4.4% in 2026.
The increased number of matches obviously matters when comparing raw goal totals, which is why the conversion-rate increase is the more informative figure.
Set-piece optimisation is not merely a Premier League fashion.
It is spreading through elite football.
The beginning of a set-piece arms race
Sportradar describes the current environment as a developing tactical arms race.
That description feels appropriate.
Once one tactical innovation generates measurable value, competitors copy it.
Then defences adapt.
Then attacking teams counter-adapt.
We may therefore be observing a familiar evolutionary process.
First:
teams discover that inswingers generate superior outcomes.
Then:
inswingers become more common.
Then:
defences redesign their structures specifically for inswingers.
Then:
the relative value of another solution may increase.
Football tactics rarely reach permanent equilibrium.
Success creates imitation.
Imitation creates defensive adaptation.
Adaptation creates the next innovation.
The next phase may be defensive
Sportradar makes an especially interesting observation in its discussion of 2026/27: as attacking corner routines become more sophisticated and more widely copied, the next competitive advantage may emerge from defending them better.
Our spatial findings suggest where that defensive battle may occur.
Not simply “inside the box.”
Not even simply “centrally.”
The critical territory appears to be the corridor immediately around the six-yard area where inward-curving deliveries produce their deepest finishing actions.
Defending that space may require more than assigning taller players.
It may require optimising:
starting orientation;
visual access;
goalkeeper-defender coordination;
blocking resistance;
support-foot position;
recovery rotation;
runner handover;
first-contact zones.
That is a much richer tactical problem.
What coaches can take from this analysis
The practical implication is not that every corner should automatically be inswinging.
Football does not work like that.
The stronger conclusion is that delivery direction changes the spatial distribution of the resulting finishing action.
For attacking coaches, this suggests evaluating corners through a spatial chain:
delivery type → delivery zone → first contact → shot location → shot outcome.
For defensive coaches, the chain can be reversed:
expected delivery → defensive orientation → first-contact protection → second-ball structure → shot suppression.
The corner itself is only the beginning of the sequence.
What analysts should measure next
Traditional corner analysis frequently stops at:
corners taken;
goals scored;
conversion percentage.
That is no longer sufficient.
A more complete set-piece model should include:
Delivery
inswinger / outswinger;
foot;
angle;
speed;
height;
destination.
Structure
number of attackers;
number of defenders;
zonal/man-marking configuration;
blockers;
starting positions.
First contact
player;
zone;
body part;
direction.
Finishing
shot X/Y;
distance;
angle;
body part;
goalkeeper location;
goalmouth coordinates.
Outcome
shot;
shot on target;
goal;
retained possession;
defensive clearance;
transition conceded.
Only then do we begin to model the true value of a corner.
Limitations
A serious analysis should make its limitations explicit.
First, the Football Hacking swing classification is available only when the event data provides usable foot information.
Of the 16,724 corners in the sample, 6,545 therefore remain unclassified.
We do not guess.
Second, the Football Hacking spatial analysis deliberately examines immediate finishing events explicitly associated with the corner. It should not be interpreted as using exactly the same definition of corner success employed by Sportradar.
Third, the analysis is descriptive.
It does not control for:
team strength;
individual taker quality;
defensive quality;
game state;
marking scheme;
goalkeeper behaviour;
delivery velocity;
attacking player height;
score effects.
Fourth, our body-orientation interpretation is a Football Hacking tactical hypothesis, not something directly measured by the event dataset.
We can observe where the ball and shot occur.
We cannot observe a defender’s gaze direction or exact foot-loading mechanics from these event coordinates.
The hypothesis should therefore be treated as a proposed mechanism consistent with the spatial pattern—not as established causality.
The most important finding
If we reduce the Football Hacking analysis to one observation, it is this:
Inswinging and outswinging corners do not merely produce different trajectories. They produce different finishing geographies.
Across our classified Big Five sample, roughly 40% of immediate shots generated by inswingers occurred inside the six-yard box.
For outswingers, the figure was approximately 9%.
The average inswinger shot occurred roughly 9 metres from goal.
The average outswinger shot occurred roughly 12 metres away.
The lateral distributions were surprisingly similar.
The difference was penetration.
That is a tactically meaningful result.
From conversion rate to defensive information
Sportradar’s research established the broader context.
Corner efficiency has increased.
Inswingers have become more common.
Inswingers are producing stronger conversion rates.
Premier League and Bundesliga clubs have moved particularly aggressively toward them.
Football Hacking’s spatial analysis suggests a possible part of the explanation.
The inswinger pushes the decisive action closer to goal.
But perhaps more importantly, it does so while creating an extremely difficult perceptual problem for defenders.
A defender needs information.
Where is the ball?
Where is the attacker?
Where is the goalkeeper?
Where is the contact point?
Where is my support foot?
Where can I rotate?
The inward-curving corner may deliberately make those questions harder to answer simultaneously.
And the attacker has the one advantage the defender can never completely eliminate:
the attacker knows what is supposed to happen.
Conclusion: the corner is becoming a designed attacking system
The modern corner should no longer be understood as simply putting the ball into the penalty area and hoping somebody wins a header.
At elite level, it is becoming a designed attacking system.
Sportradar’s ten-season research demonstrates the macro evolution: corner efficiency has reached a decade high, inswinging deliveries have expanded rapidly, specialist coaching is becoming increasingly important and corner goals are now materially affecting championship, European qualification and relegation outcomes.
Football Hacking’s 2025/26 spatial analysis adds another layer.
The distinction between inswinger and outswinger is visible not merely in aggregate outcomes but in where the resulting finishing action occurs.
Both left- and right-footed inswingers pull immediate shots significantly closer to goal.
They dramatically increase the proportion of those shots occurring inside the six-yard area.
And they potentially create precisely the defensive environment in which body orientation, visual information and rotational movement become hardest to manage.
This does not mean the outswinger is obsolete.
Quite the opposite.
As defences increasingly optimise themselves for inward-curving deliveries, tactical variation may become more valuable again.
But the direction of travel is difficult to ignore.
Corner kicks are no longer a marginal source of attacking value.
They are becoming a specialised phase of possession where geometry, biomechanics, visual information, player movement and data analysis converge.
And if the tactical evolution identified by Sportradar continues into 2026/27, the next competitive advantage may not come from discovering that inswingers work.
Elite football already appears to know that.
The next advantage will come from understanding why they work, exactly where they create danger, and what defenders can do about it.
Football Hacking would like to thank Sportradar and Alex Bake for sharing Turning the Corner into Goals and providing the research that inspired this independent spatial analysis. All historical figures and Sportradar-specific findings referenced above are attributed to Sportradar. The spatial event analysis, classifications and tactical interpretations are Football Hacking’s own work.