Does Goalkeeper Height Predict Performance? What the 2023-24 UCL Data Says

The dataset is 1,751 shots on goal across both Champions League competitions in 2023-24. Each shot is tagged with whether it became a goal, the zone of the goal it was directed at, and an xGOT value — the probability that an average top-level keeper would concede that specific shot given its trajectory and placement. xGOT differs from xG: it conditions on the shot already being a shot on target, isolating shot-stopping skill from chance creation.

Goalkeepers' height was the candidate explanatory variable. The headline correlation between height and shot-stopping over-performance (i.e. conceding fewer goals than the xGOT model predicts) was statistically indistinguishable from zero in both the women's and men's competitions. The predictive value of xGOT itself — i.e. how well the model anticipated which shots became goals — was 0.61 ± 0.29 for the women's league and 0.60 ± 0.29 for the men's, almost identical.

The shot-direction analysis surfaced a clearer pattern. The "low centre" zone of the goal received the highest share of shots on target across the dataset — strikers prefer the body-line low strike, which the keeper has to react to with their feet rather than dive across. But the highest goal-conversion zones were "low left" and "low right" — the bottom corners. These are the shots that beat goalkeepers most often.

That distribution has tactical implications for both attackers and goalkeeper coaches. For attackers, the bottom corners remain the highest-yield placement target despite being the most-coached defensive zone. For goalkeeper coaches, ground-saving technique to the low-corner zones is doing more of the shot-stopping work than aerial coverage, which most height-led recruitment models implicitly prioritise.

The intuition that taller keepers cover more goal area is geometrically correct but operationally weak. The goal is 2.44m tall and 7.32m wide. A 195cm keeper with arms extended covers an aerial reach barely 5cm taller than a 190cm one, and the marginal aerial reach is the part of the goal where shots are least frequent. The high-value shot-stopping geography — bottom corners and centrally-low — is decided by reaction time, footwork, and reading angles, not by height.

The combined effect is that the recruitment shortlist filtered by height likely excludes shorter candidates whose shot-stopping fundamentals are excellent in the zones where most goals actually go in. The study's authors conclude explicitly: goalkeeper height should not be regarded as a determining physical characteristic for performance in either competition.

• 1,751 shots analysed across the 2023-24 UCL and UWCL seasons.
• No correlation between keeper height and xGOT over-performance in either league.
• Low centre received the most shots on target; low left + low right conceded the highest goal share.
• Reaction + footwork drive bottom-corner shot stopping, not aerial reach.
• Recruitment implication: height as a pre-filter excludes candidates whose stoppable-zone skills may be stronger.

Height does still help with crosses, set-piece coverage, and one-v-one aerial duels — situations the xGOT-on-shots framework doesn't capture. A goalkeeper's total contribution includes those phases, and the study isn't a comprehensive verdict on every facet of the position. What the study does say, and supports cleanly, is that the specific question of "does this keeper save more shots than the model expects?" doesn't depend on whether the keeper is 188cm or 198cm.

For analytics-led recruitment teams, the practical implication is to drop height from the shortlist filter stage and re-evaluate previously-rejected shorter candidates on the actual shot-stopping signal. For coaching staff, the implication is that ground-saving and reaction-time training are where the marginal goal saves live.