How Often Does the Team with Higher Possession Actually Win?

Does having more possession mean you’re more likely to win? We looked at data from Europe’s top leagues and major tournaments to find out. The answer: not always. While teams with higher possession win slightly more often, the link is weak. Stats like expected goals (xG) are far better predictors of match outcomes.

Measuring the Core Question: What Percentage of Matches Are Won by the Team With Higher Possession?

We looked at five top men’s European leagues – the Premier League (England), La Liga (Spain), Serie A (Italy), Bundesliga (Germany) and Ligue 1 (France). We looked at matches where one team dominated possession. For this, we looked at three full seasons (2020–2023), using FBref data and the official league websites.

Teams with more possession win about 57% of the time. But that also means 43% of the time, they don’t.

We also looked at women’s leagues and in the FIFA Women’s World Cup 2022, teams with more possession won 38 out of 64 games. That’s about 59%. Those teams also got 1.98 points per game. Teams with less possession won 13 games (20%) and got 0.81 points per game.

Possession Alone: A Weak Predictor of Match Outcome

From the stats above, most teams with more possession won. But there are still instances where those with less possession won. That means possession alone is not enough to predict the match outcome.

We also used the Pearson correlation coefficient (r) to measure the relationship and direction of possession (x) and winning (y). Here’s the formula:

r = (n * Σxy – Σx * Σy) / √[n * Σx² – (Σx)²] [n * Σy² – (Σy)²]y

Our results showed that:

• In the Premier League, r = 0.24
• In La Liga, r = 0.29
• In Bundesliga, r = 0.20

If r is less than 0.3 (as in all three leagues we analysed), there’s a low correlation between the two variables. That means possession is weakly linked to success. In simple terms, just keeping the ball doesn’t mean you’ll win. There are other factors, like:

• Chance quality: Are you making good shots or just passing?
• Defensive setup: Can the other team defend and counter-attack?
• Game state: Are you chasing the game or protecting a lead?

Regression Analysis: Does Possession Improve the Odds of Winning?

Facts alone aren’t enough to draw the correct conclusion. So we ran a regression analysis to see how increasing possession changes winning chances. Here’s our simple model:

P(Win) = 1 / (1 + e^-(β₀ + β₁ × Possession))

β₁ = 0.045 in our Premier League dataset (p < 0.01, meaning it’s statistically significant)

Regression models show that the coefficient for possession (β1) is usually positive but small, meaning a small increase in win probability as possession rises. For example:

• 45% possession: Win probability is around 30-35%.
• 55% possession: It goes up to 40-45%.
• 65% possession: It can get to 60-65%, but only when combined with good chance creation.

As we mentioned earlier, higher possession helps because when you have the ball, the opponent can’t score. But it won’t work on its own. Winning isn’t about having the ball. It’s about goals and if you don’t score, it doesn’t matter how long you had the ball.

Case Study Comparison: Teams That Dominate the Ball but Lose

Looking at different possession-dominant teams, it’s clear that possession alone isn’t enough. Teams like Manchester City, Brighton and Tottenham had the highest possession rates in the 2023-2024 EPL season but only City created the most chances when in ball control. The rest didn’t do as much and that’s why Arsenal and Aston Villa beat them to the top spots.

Let’s look at a few more examples:

• Brighton 2021/22 (vs Burnley): 68% possession, 0–3 loss.
• Spain vs Russia (2018 World Cup): Spain had 74% possession. They lost on penalties.
• Arsenal before Arteta: They had 60%+ possession in most matches, but they struggled to win.

These teams controlled the game but struggled to convert control into goals, especially against harder defenses. Coaches call this sterile domination, which simply means keeping the ball but not doing anything to generate results.

Possession vs Expected Goals (xG): Which Is the Better Predictor?

xG tells us how likely a shot is to score and therefore has a higher correlation to match results than possession (0.58). Also, scatterplots show a clearer relationship between xG and wins than between possession and wins. From here, you’ll see that possession is descriptive while xG is predictive.

To help drive the point home, here’s an example:

• Team A: 65% possession,  0.8 xG
• Team B: 35% possession, 2.1 xG
• Likely result: Team B wins even when out-possessed.

So, xG gives a clearer picture of who’s likely to win. Possession shows how the game looks.

Sources

1. https://fbref.com/en/comps/9/Premier-League-Stats