You Don’t Have to Succeed, Just Don’t Fail: The Effect of a Draw Outcome on Soccer Coaches’ Dismissal

Background: Literature has probed the effect of coach dismissal on a team’s performance. All studies measured the change in performance, in one way or another, as the difference in league points before and after the dismissal.

Aims: In this study we have three main aims: 1) To investigate whether winning is a prerequisite for ensuring the continued employment of the coach, or whether it is sufficient to avoid losses; 2) To check if the effect of league points on the probability of the coach’s dismissal is monotonic or not; and 3) To distinguish between the effect of lack of success and the effect of failure on the likelihood of the coach to be fired.

Method: We ran logit regressions according to 8,820 observations from all the English premier league match results since season 2004/5.

Results, Discussion and Conclusions: We found that to ensure the continued employment of the coach it is sufficient to avoid losses. This finding may be explained by loss aversion (Kahneman & Tversky, 1984). While the number of league points is biased in favor of a win result (3 points for a win, 1 point for a draw), the decision of dismissal is taken according to losses. Therefore, the number of league points cannot monotonically predict coaches’ dismissal. We also found that in order to keep your job, you don’t have to succeed – you just have to avoid failure.