A five-player wall hid early ball flight and delayed skilled goalkeepers’ movement by 70ms in a virtual-reality study. Use a wall only when its blocked shots outweigh the cost of the goalkeeper moving later. ¹

The study tested 10 skilled goalkeepers and 15 novices against identical trajectories with and without a central wall. The kicks began 23 metres from goal and had flight times of one second or 1.2 seconds. ¹

The wall obscured roughly 150 to 250 milliseconds of early flight. Skilled goalkeepers began moving 70 milliseconds later, and their median hand-position error increased by 1.5 centimetres. Their success rate fell from 85.8 percent without the wall to 79.1 percent with it.

Interception error rose more when the goalkeeper had to move sideways towards either corner. The wall also reduced early movement in the direction of the ball’s curve, but the study found no corresponding final-error difference between skilled goalkeepers facing congruent and incongruent trajectories.

The simulator kept the goalkeeper and wall central and required a step-and-reach action. It omitted the taker’s body, diving saves, off-centre wall positions and elite-match pressure. Every kick followed the same trajectory regardless of wall condition, so differences in goalkeeper performance reflect occlusion alone.

Five defenders and no defenders were the only wall conditions in the main experiment. The 70-millisecond delay cannot be assigned to removing one player, moving the outside edge or changing the goalkeeper’s starting point. The study did not test four-player walls. Teams still have to compare four-player walls with five-player and no-wall setups in training.

The authors recommend evaluating whether to use a wall against expert takers who repeatedly clear it at pace. Their suggested comparison uses the opponent’s historical rate of shots hitting the wall and field tests measuring how much the goalkeeper improves when the entire flight is visible.

Free kicks beyond 25 metres need a separate wall test

A 2023-24 Opta Analyst snapshot recorded 264 Premier League direct free-kick attempts worth 16 expected goals, or 0.06 xG per shot. Nine had been scored with 23 league matches still to play. ²

Seven of those nine goals came from 20 to 25 metres. That range overlaps the PLOS experiment’s 23-metre kicks and gives teams a sensible starting band for wall comparisons. Analysts still need to record where the wall stood, when the goalkeeper saw the ball and when the goalkeeper first moved.

The one-second trajectories produced 2.8 centimetres of additional error for skilled goalkeepers with the wall. The 1.2-second trajectories produced 0.3 centimetres. Faster shots therefore deserve separate testing because later first sight leaves less time for a lateral save.

The FIFA Training Centre’s long-range analysis identifies curl, knuckleball and power as distinct solutions beyond 25 metres. Curl attacks the wall-protected side, while knuckleballs and power shots seek late movement or reduced reaction time. ³

Fast knuckleballs were outside the controlled experiment. Teams should separate 20-to-25-metre kicks from longer attempts and divide the longer group by technique. Pooling a slower curling effort with a driven knuckleball would hide the difference in the goalkeeper’s available reaction time.

Test each wall with the same taker and shot location

No cited source compares wall sizes under match conditions. The same taker should shoot repeatedly from the same spot against the usual wall, one fewer defender and no wall, with the goalkeeper beginning from the agreed position for each setup.

A wrong-way step means the goalkeeper’s first lateral movement opposes the shot’s eventual side. A late step should be judged against that goalkeeper’s timing from matched attempts with clearer sight, avoiding a universal millisecond threshold drawn from ten participants in virtual reality.

IFAB Law 13 requires defenders to remain at least 9.15 metres from the ball. Attackers must stay one metre from a wall containing three or more defenders until the kick is taken, and a defender may legally lie behind the wall. ⁴

Training comparisons should preserve those distances and the intended prone defender. Moving the wall closer or allowing attackers to obscure contact would change the goalkeeper’s view and make the repetitions poor copies of the match situation.

SoccerNet Tracking contains 100 main-camera clips from 12 matches and tracks players, goalkeepers and the ball. Researchers would still need to label wall membership, wall edges, contact visibility, goalkeeper position and shot route before using those clips to test wall setups. ⁵

Staff should log the contact-to-first-visible-ball frame count for every matched attempt, compare the result across wall setups, and reject or reposition any configuration that delays visibility without improving the block rate.