Substitutes, Part II: Clubs, Leagues and Managers

or, England hates subs

In the first full Expecting Goals study, I tried to quantify the substitute effect. How much does it inflate a player's statistics to play more minutes coming off the bench than starting?

In this study, I want to look at the question from the other side. Rather than asking about the effects that substitutions have on player statistics, we can consider the changes themselves. Substitutions offer an opportunity to investigate managerial impact on the game, as they reflect active decisions by managers to use their bench at particular times. Or, as we will see, not to.

What are the tendencies in the use of substitutes? Are there persistent tendencies among clubs and managers to use their subs in particular ways? I had hoped this could be a straightforward and relatively quick study, but it turned out to be far more confounding than I expected. Such is the world of soccer statistics.

Substitute Usage Doesn’t Tell Us That Much

This graph displays the 20 teams which had the heaviest substitute usage in league play since the 2010-11 season and the 20 teams which used the fewest substitutes, as a simple percentage of minutes played.

One factor here I had predicted, but the other surprised me. Obviously, all of the clubs with the highest percentage of substitute usage played in leagues which allowed managers to make five personnel changes, not just three.

Until the world shut down in March 2020, the rules of just about every top level professional football league allowed teams to make only three substitutions per match. When league matches kicked off again that summer in front of empty stadiums, teams were allowed five substitutions instead of three. To make matters more confusing for study, this pandemic-era rule was retained as the standard in most leagues for the 2020-21 season, but the Premier League said, no, we’re not gonna. We’re going to have a sandwich.

For two seasons, there was one rule in the Premier League and another on the continent. Finally in 2022 the Premier League adopted the five changes rule and standardization returned.

So when we are talking about substitutions, we are drawing from two different samples. In leagues with a five sub rule, about ten percent of minutes are played by substitutes, compared to between six and seven percent when teams are allowed three changes.

That I knew, more or less. But the league effect here is so much larger than just a question of rules. Of the 20 teams which used their substitutes most aggressively, 19 played in either La Liga or Serie A. And of the 20 teams which used the fewest substitutes, 12 played in the Premier League and none in La Liga or Serie A.

There also appears to be some meaningful club and managerial effects, especially in the teams which use substitutes the least. Lucien Favre and Sean Dyche….. hello. We will be returning to this. But the managerial tendencies can only be understood once various context issues have been accounted for.

League Effects

You can see this clearly in a simple graph of how many minutes substitutes have played between the big five leagues. Sub usage increases during the re-opening in 2020 (with the exception of Ligue 1, which cancelled its season), and then Premier League substitute minutes fall dramatically compared to the rest of Europe as the rules diverge. But what is also clear is that even without the context of these rules, La Liga and Serie A teams use their substitutes more aggressively. And teams in the Premier League use them the least, with somewhat more variation among teams in the Bundesliga and Ligue 1.

It also appears that adaptation to the new substitute rule takes time. In leagues which did not revert to the pre-pandemic rules, the average percentage of minutes played by substitutes in 2020-21 under the new rules was 9.6 percent, and this rose to 10 percent in 21-22 and then to 10.6 percent in 22-23. It is reasonable to expect based on this pattern that substitute usage in the Premier League may catch up to some of the continental leagues over the next few seasons.

But until then, England stands out. Another way of getting at the same issue is to look an unused substitutes. Premier League teams only manage to get a fifth substitute into the match about one-third of the time under the new rules. In one match in ten they don’t even get around to the third substitute, while such a choice would be vanishingly rare in Serie A or La Liga.

The Cultural Context of Substitutions

This effect is strange, persistent and telling. Managers in the Premier League make substitutions later and more sparingly than managers in Serie A or La Liga, and managers in the Bundesliga and Ligue 1 fall somewhere in between.

It is hard to attribute this to anything other than cultural factors. I do not mean that something inherent in the Italian soul longs to bring on a backup striker, but rather that accepted norms in Italian and Spanish football make managers more likely to use their substitutes. Players likewise, it seems, have become acculturated by training and experience to know they are more likely to be pulled from the match or called up from the bench. We can even see this acculturation process happening in the increase in the use of substitutes the longer the five subs rule remains in place.

This finding makes it difficult to argue that current patterns of substitute usage are already optimal. You can imagine an argument that whatever positives are found from players entering a match fresh, they are already balanced out by other factors and so managers already know what they’re doing. But this cannot be squared with the cultural effects. It is not plausible that there is an optimal way of using substitutes in La Liga which is 30 percent more aggressive than the optimal way of using substitutes in the Premier League. Better ways are surely possible if both of these norms are currently possible.

But of course identifying better ways in any truly convincing way is a difficult problem. What will be useful first is understanding what these different trends in substitute usage across Europe really are, and the team and manager tendencies behind them.