For most of football’s modern era, the home side has won roughly 45% of league matches, give or take a few points. That number has not held steady across the last six European seasons. It collapsed, recovered, and then split unevenly across the five major leagues in 2024-25.
What we measured
For each of the Premier League, La Liga, Serie A, Bundesliga, and Ligue 1, we tallied every home result across each of the last six published seasons. Each cell in the table below is the percentage of matches that ended in a home win, a draw, and an away win.
The data
| League | ’19-20 H/D/A | ’20-21 H/D/A | ’21-22 H/D/A | ’22-23 H/D/A | ’23-24 H/D/A | ’24-25 H/D/A |
|---|---|---|---|---|---|---|
| Premier League | 45.3 / 24.2 / 30.5 | 37.9 / 21.8 / 40.3 | 42.9 / 23.2 / 33.9 | 48.4 / 22.9 / 28.7 | 46.1 / 21.6 / 32.4 | 40.8 / 24.5 / 34.7 |
| La Liga | 45.8 / 27.6 / 26.6 | 41.6 / 28.7 / 29.7 | 43.4 / 29.2 / 27.4 | 47.9 / 23.4 / 28.7 | 43.9 / 28.2 / 27.9 | 44.5 / 25.5 / 30.0 |
| Serie A | 41.6 / 22.4 / 36.1 | 41.1 / 25.3 / 33.7 | 38.9 / 25.8 / 35.3 | 42.4 / 26.3 / 31.3 | 41.8 / 29.5 / 28.7 | 39.7 / 28.4 / 31.8 |
| Bundesliga | 40.2 / 22.2 / 37.6 | 42.2 / 26.5 / 31.4 | 46.7 / 23.9 / 29.4 | 47.4 / 24.5 / 28.1 | 43.8 / 26.5 / 29.7 | 38.6 / 25.2 / 36.3 |
| Ligue 1* | 48.0 / 25.1 / 26.9 | 37.4 / 25.0 / 37.6 | 42.6 / 26.8 / 30.5 | 42.9 / 24.2 / 32.9 | 39.2 / 26.5 / 34.3 | 46.7 / 20.3 / 33.0 |
* Ligue 1 ’19-20 is a partial season (279 of 380 matches played). The league was cancelled in April 2020 due to COVID and Paris Saint-Germain declared champions on a points-per-game basis.
Numbers tallied from each league’s published season results, cross-referenced against the home/away matrices on the corresponding per-season Wikipedia pages.
What the table shows
The headline is not a trend. It is three distinct phases.
The COVID collapse (2020-21)
Empty stadiums produced a sudden, immediate dip in home wins across every league. The Premier League fell from 45.3% to 37.9%, a drop of 7.4 percentage points. Ligue 1 fell from 48.0% to 37.4%, a drop of 10.6 points. La Liga dropped from 45.8% to 41.6%. The pattern is consistent with the broader crowd-effects literature. When the away side no longer has a hostile environment to absorb, the home side’s advantage shrinks.
The Bundesliga was the partial exception, nudging up from 40.2% to 42.2%. German sides appear to lean more on tactical shape than on crowd-driven momentum, and that shape held even without the stands.
The 2022-23 peak
The recovery was not gradual. It overshot. Every league hit its highest or near-highest home-raw accuracy of the six-season window in 2022-23. The Premier League reached 48.4%, its highest in the period. La Liga hit 47.9%. The Bundesliga hit 47.4%.
For one season, home-field advantage looked fully restored. The 2022-23 calendar is a useful baseline. This is what home-field advantage looks like in normal, full-stadium conditions.
The 2024-25 split
The recovery has not held evenly.
In 2024-25 the Premier League fell to 40.8%, down 7.6 points from its 2022-23 peak. The Bundesliga fell to 38.6%, down 8.8 points and its lowest in the dataset, with away wins jumping to 36.3%. La Liga and Serie A held roughly steady around their long-run averages. Ligue 1, the league that collapsed hardest in 2020-21, has rebounded sharply to 46.7% in 2024-25, almost back to its pre-COVID peak.
Why the split, not the trend
The Bundesliga-vs-Ligue 1 divergence is the most useful counter-example for anyone tempted to argue that home-field advantage is uniformly fading. It is not. Two leagues on the same continent, both playing with full stadiums, both operating under the same VAR rules, moved in opposite directions in a single season.
Whatever is driving the change is not a continental phenomenon. The split suggests at least three forces are working at once. Tactical shifts that favour the pressing away side. Fixture pile-ups that dilute the home side’s preparation. League-specific dynamics, Bundesliga’s away-goal intensity, Ligue 1’s home-dominant profile, that each league carries through its own cycle. None of them is large enough to dominate. Together they produce noise that looks like a trend.
What we tried that didn’t work
We first treated home advantage as a static constant in our model, a fixed bump added to every home match. The 2022-23 numbers validated that assumption. The 2024-25 numbers broke it.
We then tried a fixed decay, applying a uniform decay rate to the home boost over time. That overcorrected. Ligue 1’s 2024-25 rebound came in far above the decay curve, and the Bundesliga’s collapse did not match a smooth slope.
What we arrived at is a season-recency-weighted term. The home boost is anchored to the most recent two seasons, with smaller weight on the prior four. The Bundesliga’s 2024-25 collapse shows up immediately. Ligue 1’s rebound does too. The longer-history numbers still inform the prior, but they do not anchor it.
Known limitations
Six seasons is a short window. A pattern that looks visible over six data points can disappear once the next three seasons land.
The 2020-21 column is a structural outlier. Empty stadiums are a one-off event. The COVID column is informative as a natural experiment. It is not informative as a baseline.
Ligue 1 2019-20 is a partial season. The 48.0% figure is from 279 of 380 matches. It is consistent with Ligue 1’s structural home-dominance but is not directly comparable to a full-season figure.
We have not controlled for schedule asymmetry, which teams play at home after European midweek fixtures, which teams face freshly promoted sides in the opening weeks. Those effects are small in aggregate but they are real, and they vary league by league.
The 18-team format change in Ligue 1 (2024-25) and the Bundesliga (since 2003-04) reduces the total match count per season for those leagues, but does not affect the percentage interpretation.
What this means for an analysis
A 60% home-win call in 2024-25 lands in a narrower range than the same label in 2019-20. The model now applies a slightly weaker home boost than the 2019 baseline, anchored to recent data.
That is calibration, not a retreat. A probability that maps cleanly to the recent record is more useful than one that preserves a 2019 baseline at the cost of being wrong about 2024-25. The Bundesliga’s 2024-25 home-raw accuracy of 38.6% would have been treated as a noisy outlier under a static constant. Under a recency-weighted term it is the new baseline.
The five-league model is what makes this calibration possible. A single-league model would never surface the Bundesliga-vs-Ligue 1 split, and the home boost would stay anchored to whichever league it was built for.
For a reader following coverage across leagues, the same probability band now means different things in different competitions. A 60-70% home-win bucket in Ligue 1 will land over a narrower range than the same bucket in the Bundesliga. That is the model accurately capturing that the Bundesliga’s home advantage has weakened this season, while Ligue 1’s has not.
A read is a lens for separating signal from noise. Six seasons of data is enough to see that the signal is moving. The lens has to move with it.