The 15cm Trick That Rewrites Penalty Theory

The penalty spot has never been fair

Economists Pierre-André Chiappori, Steven Levitt and Sohyoung Groseclose worked this out back in 2002, studying 459 penalties across the French and Italian top flights. Their conclusion was that both takers and keepers already play something close to a mixed strategy, similar to how a poker player varies their bluffs. A right-footed kicker in their data placed the ball left 43% of the time, centrally 24% of the time, and right 33% of the time. The goalkeeper, in response, dived left 60% of the time, stayed central 9% of the time, and went right 31% of the time.

The payoff matrix behind those numbers tells you why. When the keeper dives correctly to the taker's left, the shot still goes in 63.2% of the time. When the keeper dives correctly to the taker's right, it goes in only 44%. A left-side penalty from a right-footer is simply harder to stop even when the keeper guesses right, because that side carries more natural power and accuracy for a right-footed strike.

Why the edges of the goal matter more than the middle

Penalty scoring probability isn't flat across the width of the goal, it climbs the closer the shot gets to a post, because the keeper has less distance to cover and less time to get there. Combine that with the left/right asymmetry above and you get a clear picture: the most dangerous real estate in the goal is the far-left corner from the keeper's perspective, the taker's natural side.

That is the space a goalkeeper standing dead centre is defending worst.

What moving 15cm actually does

Shift the goalkeeper roughly 15 centimetres towards their own right, which is the taker's left, before the kick, and something specific happens to the geometry. You are effectively removing a sliver of space from the most dangerous part of the goal, the taker's favoured left side, and adding an equivalent sliver to the less dangerous right side. Because scoring probability rises steeply near the posts, the centimetres taken away from the left cost the taker far more than the centimetres added on the right benefit them.

The updated payoff matrix reflects this. Once the keeper commits to standing off-centre and the striker adjusts their aim accordingly, the numbers look like this:

Compare the top-left figure to the old central setup: a correctly-guessed dive to the left now concedes a goal 57.7% of the time instead of 63.2%. The taker's biomechanical edge, the extra power and accuracy on their strong side, has been largely cancelled out. Shooting left and shooting right now carry almost identical risk for the keeper, which is precisely the point: the shift doesn't just steal a few centimetres, it dismantles the asymmetry the taker was relying on.

Why the gain is real but small

Here is the catch. Once strikers know a keeper is standing 15cm off-centre, they adjust their placement to compensate, which is exactly what game theory predicts in a mixed-strategy contest. The keeper's advantage doesn't vanish, because the underlying geometry, edges being harder to reach than the middle, hasn't changed. But it does shrink once both sides are playing optimally again. The net gain in save rate once everyone has adapted comes out at a small fraction of a percentage point.

That sounds trivial, and across a single shootout it mostly is. But penalty saves are rare events decided by centimetres and half-seconds, and elite goalkeeping careers are increasingly built on exactly these margins. A keeper who reliably converts one extra save every few hundred penalties, over a career of shootouts and spot-kicks, has found something no amount of reflex training could give them.

The takeaway for goalkeepers and coaches

The practical lesson isn't that goalkeepers should stand visibly to one side, which would tip off the taker and invite an obvious counter. It's subtler: small, almost imperceptible positional bias before the run-up, combined with the existing mixed-strategy logic of diving left more often than right, squeezes extra value out of a contest that both sides have already mathematically optimised as far as dive direction and shot placement go. Fifteen centimetres, invisible to the taker, is enough to turn a 63.2% goal into a 57.7% one.