Game Theory in Soccer Set Pieces
Article: https://statsbomb.com/2017/02/changing-how-the-world-thinks-about-set-pieces/
With the rise of data in sports, many data analysts have been able to find areas that their teams can exploit.
A notable example of this was with “set pieces” (corner kicks, free kicks, or throw ins) in soccer. In the above article, the author, a data analyst, mentions that set pieces were something professional clubs places placed a low emphasis on, with many of them doing set piece training for only “5 to 15 minutes a week” even though set pieces typically account for .3 goals scored per game.
As such, he helped FC Midtjylland, a Danish team that had never won the league before, implement a set piece system prior to the 2014-2015 season. At the end of the season, FC Midtjylland had finished first. During that season, they scored .75-.8 goals from set pieces per game, a large improvement compared to the average of .3.
Example set piece strategy
Since we know by how much on average training set pieces improves expected goals, we can model soccer clubs choosing whether or not to train set pieces as a game.
Players: two soccer clubs playing each other
Strategies: train (T) or not train (NT) set pieces
Outcomes: representing the expected set piece goals (SPG) per game, with .3 representing the average number of SPG if set pieces are not trained, and .7 representing the average number of SPG if set pieces are trained.
We can see that the only Nash Equilibrium is the pure strategy to train set pieces. Training set pieces is also an example of a game in which there’s one strictly dominant strategy for both players (training set pieces). As such, once FC Midtjylland won the league, other teams followed their strategy in training set pieces.
This emphasis on set pieces has gained prevalence throughout the sport. For example, one of the stories of the World Cup this summer was England’s effective use of set pieces en route to the semifinals.