A run of defeats, a worried board, a statement on the club website: sacking the manager mid-season is football's most common emergency measure. And it often seems to work. The new manager arrives, the team picks up points, and the decision looks justified. But does the change cause the improvement? Economists who have studied the question have a sobering answer.
The "new manager bounce"
The pattern is familiar. A team goes through a bad spell, the manager is sacked, and in the following weeks results get better. Fans and pundits call it the new manager bounce and credit fresh ideas, a lift in morale or players fighting for their places.
The problem is what the comparison leaves out: what would have happened if the manager had stayed?
What the research found
Jan van Ours and Martin van Tuijl studied in-season coaching changes in the top division of Dutch football over 14 seasons, published in Economic Inquiry in 2016. Their method was to compare teams that sacked their coach with teams that went through a similarly bad run but kept their coach.
Their finding: teams did improve after a sacking, but the comparison teams that kept their coach improved about as much. The replacement of head coaches did not improve performance. The better results that followed were mainly regression to the mean.
Key point: As Van Ours summarised it for Erasmus University Rotterdam, results often improve after a dismissal, but that improvement would most likely have come even if the manager had stayed.
Regression to the mean, explained
Every team's results are a mix of underlying quality and luck. A manager is usually sacked at the bottom of a run, when luck has been particularly bad: shots hitting the post, opponents scoring from their only chances, a penalty against. Bad luck does not last. As it evens out, results drift back towards the team's real level, whoever is in charge.
Example: Invented figures for illustration. A team that usually collects 1.3 points per game goes through eight matches at 0.6 per game, partly through bad luck. The manager is sacked. Over the next eight matches, the team averages 1.2 per game. That looks like a big improvement under the new manager, but it is close to the team's normal level: most of it would probably have come anyway.
Does that mean sacking never helps?
Not quite. The research shows that, on average, a mid-season change does not improve results beyond what regression to the mean would bring. Individual cases can differ:
- A manager who has lost the dressing room, or a team with a broken tactical system, may genuinely benefit from a change.
- A new manager can fix a specific problem, such as set-piece defending, quickly.
- Other changes made at the same time, such as a January signing, can make a difference.
The difficulty is telling those cases apart from simple bad luck, and boards rarely make that distinction under pressure.
How to judge a sacking with data
- Look at underlying numbers before the sacking. If the team's expected goals for and against were fine while results were bad, it was probably unlucky, and the new manager will inherit an improvement.
- Compare with the team's longer-term level, not with the worst run.
- Wait. Eight matches are too few to judge any manager, new or old.
Watch out: Short spells after a change are full of noise. A new manager who wins three of the first four matches has not proven anything yet, and neither has one who loses them.
To see how much a few chances can swing a match result, use the xG match calculator, and read How to Read Football Stats for the metrics that separate luck from performance.
- Results often improve after a mid-season sacking, but similar teams that kept their manager improved too.
- The main reason is regression to the mean: managers tend to be sacked at the low point of a run.
- Judge a team's form with underlying numbers such as xG, not with its worst run of results.
Sources
- In-season head-coach dismissals and the performance of professional football teams — Van Ours and Van Tuijl (CEPR discussion paper; published in Economic Inquiry, 2016)
- Sacking a manager mid-season usually pointless, says Jan van Ours — Erasmus University Rotterdam
Premier League and Champions League by the numbers.