A team loses four in five, the manager is sacked, and everyone talks about the "new manager bounce". Results often do pick up. But did the new manager cause it, or was the team always going to recover?
We collected every in-season manager change in the big five leagues and the Championship since the start of 2024/25: 119 changes. For the 99 with at least five league matches either side, we compared the team before and after, and then with teams on the same bad run that kept their manager.
Data: StatPair match database · Premier League, La Liga, Serie A, Bundesliga, Ligue 1, Championship · 2024/25 to 2026/27 · manager change dates from Wikipedia · Updated: 13 October 2026
TL;DR
- Teams do improve after a change: from 0.67 to 1.18 points per game over five league matches, and the win rate doubled from 14% to 30%.
- Teams that kept their manager improved just as much: from 0.67 to 1.21 on the same run of form. The difference is −0.03 points per game (standard error 0.05): no bounce.
- No first-match bounce either. New managers won 26, drew 27 and lost 46 of their first league games (1.06 points per game). Comparable teams averaged 1.25 in their next match.
- Team stats barely move. Shots, fouls, yellow cards and corners stayed within a few tenths of the control group. Over ten matches, shots conceded fell by 0.71 per game relative to controls, the only hint of a real effect.
- Caretakers and permanent appointments look the same on results.
- The line-up is what changes. In the first match, the new man kept 7.46 of the previous XI (8.46 for comparable teams) and made four or more changes 51% of the time (27% for controls). Regular starters lost about 6 minutes per match over the first five games.
How we measured it
The list of changes. We took the managerial changes tables from Wikipedia's season articles and kept only in-season departures: sackings, resignations, mutual-consent exits and managers poached by other clubs. An interim spell followed by a permanent appointment counts as one change.
Before and after. For each change, we compared the team's last five league matches before the departure with its first five afterwards (and ten with ten, where possible). If a team changed manager twice in quick succession, the windows stop at the other change.
The control group. This is the important part. A team that has taken 2 points from 5 games will usually improve whatever happens, because a run that bad is partly bad luck. So for each change, we found every other moment in the same three seasons where a team had:
- the same number of points in its previous five league matches,
- a season-to-date points per game within 0.25 of the sacking team's,
- reached a similar stage of the season (within six matches),
- and no manager change within ten league matches either side.
That gave a median of 45 control situations per change, from a pool of 5,195. What the control teams did next is what regression to the mean alone delivers. The new manager effect is whatever is left over.
Step 1: results after a change vs teams that stuck
| Five league matches | Before | After | Change |
|---|
| Changed manager: points per game | 0.67 | 1.18 | +0.51 |
| Kept manager: points per game | 0.67 | 1.21 | +0.54 |
| Changed manager: win rate | 14% | 30% | +16 pts |
| Kept manager: win rate | 13% | 31% | +18 pts |
| Changed manager: goals for / against | 0.89 / 1.83 | 1.25 / 1.48 | +0.36 / −0.35 |
| Kept manager: goals for / against | 0.92 / 1.78 | 1.21 / 1.45 | +0.29 / −0.32 |
99 manager changes; control figures are averaged across each change's matched teams.
Look only at the sacking teams and the bounce looks huge: almost double the points, double the wins, about a third of a goal more scored and a third fewer conceded per game. But the teams that didn't change anything had the same rebound. Over five games, sacking teams collected 5.91 points on average; the controls collected 6.06.
The pattern holds however we cut it:
| Group | Changes | PPG before → after | Control before → after | Effect vs control |
|---|
| All changes | 99 | 0.67 → 1.18 | 0.67 → 1.21 | −0.03 (±0.05) |
| Sackings only | 80 | 0.64 → 1.15 | 0.64 → 1.18 | −0.04 (±0.06) |
| Big five leagues | 74 | 0.67 → 1.18 | 0.67 → 1.19 | −0.01 (±0.06) |
| Championship | 25 | 0.67 → 1.18 | 0.67 → 1.28 | −0.10 (±0.12) |
| Caretaker took the first match | 38 | 0.67 → 1.21 | 0.67 → 1.25 | −0.04 (±0.09) |
| Permanent manager took the first match | 61 | 0.67 → 1.17 | 0.67 → 1.19 | −0.02 (±0.07) |
Figures in brackets are standard errors.
The share of teams that improved tells the same story: 73.7% of sacking teams picked up more points in the next five games than in the previous five, against 75.8% of teams that kept their manager. This matches the classic academic finding on sackings. For why extreme runs drift back, see our guide to regression to the mean in betting.
Step 2: the first match
The bounce is often sold as a first-game effect, with players keen to impress the new boss. Our data doesn't support it.
| First league match after the change | Matches | Won | Drew | Lost | Points per game |
|---|
| Teams that changed manager | 99 | 26 | 27 | 46 | 1.06 |
| Matched teams that didn't (next match) | – | – | – | – | 1.25 |
Sacking teams lost almost half of their first games. They did slightly worse than the controls, although the gap (0.19 points, standard error 0.12) is within chance. Caretakers (1.11 points per game) and new permanent managers (1.03) were both below their controls.
Step 3: what happens to the team stats
For team prop markets, the question is whether a new manager changes the way the team plays. Here are the per-match numbers over five games, with the effect measured against the control group.
| Per match (five games) | Before | After | Control before → after | Effect vs control |
|---|
| Shots | 11.69 | 11.67 | 11.18 → 11.32 | −0.16 |
| Shots conceded | 13.31 | 12.94 | 13.43 → 13.35 | −0.29 |
| Shots on target | 3.73 | 4.00 | 3.50 → 3.76 | 0.00 |
| Shots on target conceded | 4.80 | 4.33 | 4.93 → 4.61 | −0.16 |
| Corners won | 4.74 | 4.50 | 4.62 → 4.52 | −0.14 |
| Corners conceded | 5.07 | 5.12 | 5.16 → 5.29 | −0.08 |
| Fouls committed | 11.97 | 11.95 | 11.94 → 11.92 | +0.01 |
| Yellow cards | 2.03 | 2.07 | 1.99 → 2.02 | +0.01 |
Over five matches, nothing moves beyond noise. Fouls and cards are flat, whatever you've heard about a "new broom" bringing intensity. Standard errors are around 0.3 for shots and fouls and 0.1 to 0.2 for the rest, so none of these differences is meaningful.
The schedule wasn't a factor either: opponents' season points per game averaged 1.40 before and 1.36 after, against 1.43 and 1.38 for the controls.
Ten matches: a small defensive signal
For the 60 changes with ten clean matches either side, one thing does stand out:
| Per match (ten games) | Before | After | Control before → after | Effect vs control |
|---|
| Points per game | 0.94 | 1.30 | 0.96 → 1.23 | +0.08 (±0.06) |
| Shots conceded | 13.35 | 12.52 | 13.26 → 13.14 | −0.71 (±0.29) |
| Shots on target conceded | 4.68 | 4.22 | 4.71 → 4.49 | −0.24 (±0.14) |
| Corners conceded | 5.32 | 5.02 | 5.18 → 5.27 | −0.39 (±0.19) |
| Shots | 11.86 | 11.97 | 11.41 → 11.26 | +0.27 (±0.30) |
| Fouls / yellow cards | 11.53 / 2.05 | 11.67 / 1.98 | 11.95 / 2.06 → 11.90 / 1.97 | +0.19 / +0.02 |
New managers seem to make teams a little harder to shoot against: about 0.7 fewer shots and 0.4 fewer corners conceded per game than expected. That fits a new coach's usual first job, fixing the defence. The points edge (+0.08 per game, under one point over ten matches) is still within noise.
Treat this carefully. We tested many stats, and only shots conceded and corners conceded reach two standard errors; neither showed up over five matches. It's a hint, not a law.
Step 4: line-ups and rotation
This is where a change makes a clear difference, and it matters most for player props. We used 2025/26 line-ups (2024/25 start data is incomplete) for the 35 changes with full line-up data, against matched controls.
| Line-up measure | Changed manager | Matched controls | Difference |
|---|
| Starters kept from previous match, first game | 7.46 | 8.46 | −1.00 (±0.24) |
| Four or more changes, first game | 51.4% | 26.7% | +24.7 pts |
| Starters kept per match, games 1-5 | 8.42 | 8.63 | −0.21 (±0.17) |
| Starters kept per match, last 5 games before | 8.09 | 8.53 | −0.44 (±0.14) |
| Regular starters who started the first game | 72.5% | 78.9% | −6.4 pts |
| Regular starters' start rate, games 1-5 | 67.2% | 74.4% | −7.2 pts |
| Regular starters' minutes per match, games 1-5 | 59.2 | 65.4 | −6.3 (±1.7) |
"Regular starters" started every one, or all but one, of the team's previous five league matches. Across all 2025/26 matches, teams kept 8.42 starters on average and made four or more changes 26.8% of the time.
Three things stand out:
- The first XI gets shaken up. One extra change on average, and a coin flip for four or more changes, against roughly one in four normally.
- It settles quickly. Averaged over the first five games, the new manager keeps 8.42 starters per match against 8.63 for controls, so after game one he rotates only slightly more than usual.
- The old manager was already tinkering. Teams rotated more in the five games before the change (8.09 kept against 8.53).
Regular starters lose out most: about six fewer minutes per match over the first five games than regulars at comparable teams. Minutes fall in both groups, because "regular" is defined on the previous five games, but the drop is bigger after a change.
Caretakers changed the most in game one: 6.80 starters kept and four or more changes in 8 of 10 cases, against 7.72 and 40% for permanent managers. Ten cases is indicative at best. For rotation in congested weeks, see our fixture congestion study.
What this means for betting
Don't pay for a bounce
Once you account for regression, the new manager effect on results is about zero. If a team's odds shorten sharply just because of a new name in the dugout, the market is pricing something our data can't find. The rebound is real, but it happens at teams that keep their manager too.
Price the team on its season, not its last five
Before the change, sacking teams averaged 0.99 points per game for the season, against 0.67 in their last five. For the rest of the season they averaged 1.21. The longer sample was the better guide, the same lesson as our 10-match rule study: short runs exaggerate.
Team props: no adjustment for fouls, cards or corners
Fouls, yellow cards, corners and shots were essentially unchanged. The one adjustment worth considering is a small move down on shots conceded after the first few weeks, and even that is tentative.
Player props: wait for the team sheet
- The first game is the risky one. Half of new managers made four or more changes, and regular starters started 72.5% of first games against 78.9% at comparable teams. Pre-line-up prices on minutes-dependent props carry extra risk.
- Check the first two line-ups before backing anyone's shots or tackles line. Dropped regulars lose minutes over the following weeks.
- Caretakers are the least predictable. In our small sample, they changed the most in game one.
Method notes
- Source of change dates: the "Managerial changes" tables in the English Wikipedia articles for the 2024–25, 2025–26 and 2026–27 Premier League, La Liga, Serie A, Bundesliga, Ligue 1 and EFL Championship seasons, accessed 13 October 2026. Match data comes from the StatPair database.
- Sample: 119 in-season changes; 100 had five league matches either side and 99 of those had at least one control match. 61 had ten matches either side (60 with controls). All eight changes so far in 2026/27 happened too recently to analyse, as did early-season exits such as Nottingham Forest's in September 2025. Lyon's 2025 suspension of Paulo Fonseca is excluded, because he stayed in post.
- Caretaker or permanent: "caretaker" means the first match after the departure was played before the permanent successor's appointment date (or no permanent manager was appointed that season).
- Controls were matched on recent points (within one point for the ten-match version), season-to-date form and stage of the season, not on league position or budget. A team can appear in several control groups.
- League matches only, with Championship play-offs excluded. A handful of league matches are missing from our database, which shifts a few windows by one game.
- Line-ups: 2025/26 only; about one team-match in eight lacked a full recorded XI and was skipped. 35 changes is a small sample.
- No xG. StatPair doesn't provide xG, so underlying performance here means shots and shots on target.
- Many comparisons: with more than a dozen stats tested, an occasional result near two standard errors is expected by chance. This is a descriptive study, not a causal model.
How to check this with StatPair
- Open the team's match-by-match history and compare the last five games with the whole season. Use the home/away filter to check the bad run wasn't just a string of away trips.
- Use the opponent filter to see whether the run came against strong sides.
- For player props, check each player's minutes and starts in the first games under the new manager before trusting his season averages.
- Use the Trends Finder to spot players whose shots or tackles lines have changed since the switch.
- Save players the new manager has dropped or promoted in your Notebook, and ask the AI Analyzer for a pre-match read.
Try StatPair free for 7 days → · Regression to the mean → · Spotting real form changes →
FAQ
Q: Is the new manager bounce real?
A: The improvement is real, but the manager doesn't cause it. Across 99 in-season changes, teams went from 0.67 to 1.18 points per game over five matches. Teams on the same bad run that kept their manager went from 0.67 to 1.21. The rebound is regression to the mean.
Q: Do teams win their first game after sacking the manager?
A: Usually not. In our sample, teams won 26, drew 27 and lost 46 of their first league match after a change, an average of 1.06 points per game. Comparable teams that kept their manager averaged 1.25 in their next match.
Q: Is a caretaker manager better or worse than a permanent appointment?
A: There was no difference in results. Over five matches, caretaker-led teams were 0.04 points per game below their controls and teams with a new permanent manager were 0.02 below, both well within noise. Caretakers did change the line-up more in their first game.
Q: Do teams commit more fouls or get more cards under a new manager?
A: No. Fouls went from 11.97 to 11.95 per game and yellow cards from 2.03 to 2.07 over five matches, the same as the control group. There's no "new manager intensity" effect on fouls or cards.
Q: How does a new manager affect player props?
A: Mainly through line-ups. In the first match, new managers kept 7.46 starters from the previous game against 8.46 for comparable teams, and made four or more changes in 51% of cases. Regular starters played about six fewer minutes per match over the first five games than regulars at comparable teams.
Bottom line
Sacking a manager after a bad run is usually followed by better results, but teams that kept their manager recovered by the same amount. Over 99 changes, the effect on points is indistinguishable from zero, there's no first-game boost, and fouls, cards, corners and shots don't change. What does change is the line-up, especially in game one. Bet on the team's season-long level, not the run that got the manager sacked, and wait for the team sheet before backing player props.