Player reports for central defenders across various platforms in the football industry include block-related metrics. Most data analysts or scouts include blocks per 90 minutes or a ratio like StatsBomb’s blocks/shot. Almost every pizza chart or radar plot includes a raw or possession-adjusted version of shot-blocking. However, none of these analyses include a metric that evaluates the value of a block.
We use expected goals (xG) to evaluate quality and volume of chances, chance creation, and goalkeeper’s shot-stopping. There are some VAEP (Valuing Actions by Estimating Probabilities) models that include the value of individual blocks in an overall defensive rating, but these models don’t allow for the interpretation of shot-blocking ability specifically.
Because we are interested in blocked shots, it isn’t complicated to assign an expected value to individual blocks. We just need to calculate the xG value of the respective shot that was blocked.
StatsBomb’s open data includes event data for multiple historic competitions, including the four most recent major tournaments. I calculated xG prevented by each block and summarised my findings across each tournament. The table below reveals the 15 best players in terms of shot-blocking across the last two World Cups and EURO’s.
Current Bayern Munich head coach Vincent Kompany leads all players in xG blocked per 90. In second place we can find his former Belgium teammate Toby Alderweireld. He led all players in terms of total xG blocked. In addition, he recorded the highest xG blocked within one tournament with 1.03 xG at the 2018 World Cup, helping Belgium achieve a third place finish—together with Vincent Kompany.
Jules Koundé ranks third in xG blocked p90, but also recorded the third highest valued individual block. His 94th minute block on the goal-line to secure a place in the 2018 World Cup final prevented a shot of 0.51 xG.
The current methodology has its limitations. With an average of 7.3 total blocks across the 15 most valuable shot blocker, the sample size is very small. However, StatsBomb also provides free event data for the 2015/2016 season of top 5 European leagues. Although less recent (i.e., relevant), 34 to 38 possible games per season will provide a more stable shot-blocking estimate, less vulnerable for inflation by single block. This improves the measurement of consistent positioning, anticipation, reactions, and athleticism.
I will share the findings of that dataset later this week. Feel free to subscribe to get notified on the updated model.