RadarDoes it matter?
Events that are often said to change a team's results, measured on every case in our database, next to teams in the same situation without that event.
Coach change during the season n = 109
We took every coach change during a season and compared the 5 league matches after it with the 5 before. Coaches are usually replaced after a bad run, and a bad run tends to improve anyway, so the right comparison is with teams in an equally bad run that kept their coach. “Vs expectation” = the result (1 win, ½ draw, 0 loss) minus what Elo expected before the match: −0.30 means the team took, on average, almost a third of a win less per match than the strength of its opponents suggested.
No measurable effect: teams in an equally bad run recover just as much without a coach change.
| Group | Cases | Points per match, last 5 before | Points per match, next 5 after | Vs expectation, before | Vs expectation, after |
|---|---|---|---|---|---|
| Coach changed after a bad run | 55 | 0.38 | 1.15 | −0.30 | −0.03 |
| No change, equally bad run | 140 | 0.48 | 1.20 | −0.29 | −0.02 |
Difference after the change, vs expectation: −0.02 per match (95% interval −0.07 to +0.04). The interval contains 0.
How we measured
How we chose the cases and the comparison, so anyone can check them. The 95% interval shows how much the difference could move by chance alone: if it includes 0, we cannot say the coach change made any difference.
A bad run = the last 5 league matches at or below the median of all coach changes (−0.21 vs expectation per match). Comparison moments: the same seasons, no coach change within 60 days. Only team seasons where the coach is known for every match (359 of 432). Coach dates: Wikidata; the current coach is shown only when football-data.org names the same person.
Leagues measured: Bundesliga, La Liga, Liga 1, Ligue 1, Premier League, Serie A.
Latest coach changes n = 12
The latest coach changes we measured, with the team's points in the 5 league matches before and the 5 after (a win = 3 points, a draw = 1). Example: 4 before and 9 after = 4 points out of a possible 15 before, 9 after.
| Team | League | New coach | From | Pts, 5 before | Pts, 5 after |
|---|---|---|---|---|---|
| 1. FC Union Berlin | Bundesliga | Marie-Louise Eta | 11 Apr 2026 | 4 | 4 |
| Tottenham Hotspur FC | Premier League | Roberto De Zerbi | 31 Mar 2026 | 1 | 8 |
| Sevilla FC | La Liga | Luis García Plaza | 25 Mar 2026 | 5 | 6 |
| 1. FC Köln | Bundesliga | René Wagner | 23 Mar 2026 | 3 | 6 |
| US Cremonese | Serie A | Marco Giampaolo | 18 Mar 2026 | 1 | 4 |
| FC Nantes | Ligue 1 | Vahid Halilhodžić | 10 Mar 2026 | 3 | 3 |
| Deportivo Alavés | La Liga | Quique Sánchez Flores | 3 Mar 2026 | 5 | 6 |
| RCD Mallorca | La Liga | Martín Demichelis | 26 Feb 2026 | 3 | 7 |
| Torino FC | Serie A | Roberto D'Aversa | 23 Feb 2026 | 4 | 9 |
| Paris FC | Ligue 1 | Antoine Kombouaré | 22 Feb 2026 | 4 | 9 |
| Nottingham Forest FC | Premier League | Vítor Pereira | 15 Feb 2026 | 6 | 5 |
| Tottenham Hotspur FC | Premier League | Igor Tudor | 14 Feb 2026 | 2 | 1 |
One case alone does not show an effect: compare it with the groups above.
Updated 25 Oct 2026