The Carryover Effect Does Not Influence Football Results

Abstract

In a round robin tournament, it is often believed that each team has an effect on its opponent, which carries over to the next game of that opponent. Indeed, if team A plays against team B, and subsequently against team C, A’s performance against C may have been affected by B, and we say that team C receives a carryover effect from B. For instance, if team B is a very strong team, then team A could be exhausted and discouraged after this game, which could benefit its next opponent, team C. Clearly, any schedule will lead to carryover effects. In practice, the perceived influence of carryover effects has been used as an argument when producing a schedule. In this work, we develop an approach to measure whether carryover effects have an influence on the outcome of football matches. The authors apply this method on the highest division in Belgium, using data from over 30 seasons, amounting over 10,000 matches. In our data set, we find no evidence to support the claim that carryover effects affect the results, which has major implications for the sporting community with respect to generating fixtures.