The Pythagorean Table — How Goal Difference Reveals Football's True Best Teams

July 20, 2026 · 10 min read

A team sits top of the league with 85 points, but their goal difference suggests they should have 72. Another club languishes in sixth with 68 points, yet their goals scored and conceded say they belong in the top three. The Pythagorean table — a formula borrowed from baseball analytics — cuts through the noise of results to reveal which teams are genuinely dominant and which are riding their luck.

A Baseball Formula Finds Its Way to Football

In 1980, American baseball statistician Bill James introduced a deceptively simple formula. He observed that a team's win-loss record correlated strongly with the ratio of runs scored to runs allowed, raised to a power. He called it the Pythagorean expectation because the formula's structure — a squared term divided by the sum of two squared terms — resembled the Pythagorean theorem.

The original formula was elegant in its simplicity:

Win Ratio = Runs Scored² / (Runs Scored² + Runs Allowed²)

Baseball statisticians later refined the exponent to 1.83, which proved more accurate across a wider range of scoring environments. The key insight was revolutionary: a team's scoring margin predicts their future performance better than their actual win-loss record. Teams that outscore their opponents consistently win consistently — the formula simply quantifies "consistently."

For decades, the Pythagorean expectation stayed within baseball's analytics community. Then football data analysts noticed something remarkable: the same principle applied to the beautiful game, with one critical adjustment. Football's lower scoring rate — an average of 2.6 goals per match compared to baseball's 9+ runs — required a different exponent. Research across European leagues found that an exponent between 1.3 and 1.5 produced the most accurate predictions for football.

The Football Pythagorean Formula

The adapted formula for football works like this:

Expected Points = Total Points × (GF^1.3 / (GF^1.3 + GA^1.3))

Where GF is goals scored, GA is goals conceded, and 1.3 is the empirically derived exponent for football. Some analysts use 1.35 or 1.4, but 1.3 provides the best fit across multiple leagues and seasons.

A simpler version that many analysts prefer focuses on the Pythagorean points directly:

Pythagorean Points = 3 × Matches × (GF^1.3 / (GF^1.3 + GA^1.3))

This version estimates how many points a team "should" have earned based purely on their goal record, assuming the standard three-points-for-a-win system. The formula does not account for draws directly — instead, it models the probability of winning each match and multiplies by three.

Why the Exponent Matters

The exponent controls how sensitive the formula is to goal difference. In baseball, where teams routinely score 4-8 runs per game, an exponent of 1.83 works. In football, where a 1-0 result is common and 5-0 is a blowout, a lower exponent is necessary. Too high an exponent and the formula over-rewards big victories; too low and it treats a 1-0 win the same as a 4-0 win.

The exponent of 1.3 was derived by fitting the formula to multiple seasons of Premier League, La Liga, Bundesliga, Serie A, and Ligue 1 data. It produces expected points that correlate with actual points at a rate of approximately r = 0.93 — meaning goal difference explains about 86% of the variance in league points.

Reading the Pythagorean Table: What It Tells You

The Pythagorean table ranks teams not by their actual points, but by their expected points based on goal difference. The gap between actual and expected points — often called the Pythagorean residual — reveals which teams are overperforming or underperforming their underlying numbers.

Positive Residual (Actual > Expected)

A team with more actual points than Pythagorean points is winning more games than their goal difference suggests they should. This can indicate several things: exceptional efficiency in tight matches, strong performance in one-goal games, outstanding goalkeeping in key moments, or simply good luck. Research across European leagues shows that teams with a positive residual of more than 5 points tend to regress the following season — their luck runs out, and their results converge toward their underlying performance.

Negative Residual (Actual < Expected)

A team earning fewer points than their goal difference deserves is typically one that loses too many close matches while winning comfortably. They might have a habit of conceding late goals, struggling in penalty shootouts, or failing to convert dominant performances into victories. These teams are prime candidates for improvement — their underlying quality is better than their league position suggests.

A Worked Example: The Premier League 2024-25 Season

To see the Pythagorean table in action, consider the final Premier League standings from the 2024-25 season. Liverpool won the title with 84 points, scoring 86 goals and conceding 41. Applying the formula:

Liverpool: 86^1.3 / (86^1.3 + 41^1.3) = 220.4 / (220.4 + 80.1) = 0.733. Expected points: 0.733 × 114 = 83.6. Actual: 84. Residual: +0.4. Liverpool's results matched their underlying performance almost exactly — a hallmark of a deserved champion.

Now compare that with a team like Arsenal, who finished second with 79 points, scoring 74 and conceding 36. Their Pythagorean calculation: 74^1.3 / (74^1.3 + 36^1.3) = 187.6 / (187.6 + 65.8) = 0.740. Expected points: 84.4. Actual: 79. Residual: -5.4. Arsenal's goal difference suggested they were actually the better team — they conceded far fewer goals — but their results didn't reflect that dominance. This kind of analysis is exactly what Pythagorean tables reveal.

Why Goal Difference Is a Better Predictor Than Points

The intuition behind the Pythagorean table rests on a fundamental statistical principle: goal difference is less noisy than match results. A single match result is binary — win, draw, or loss — and can be heavily influenced by a single moment of brilliance, a refereeing decision, or plain luck. Goal difference, accumulated over a season, captures the underlying pattern of performance more accurately.

Consider two teams: Team A wins 1-0 in 20 matches and loses 0-4 in 18 matches. Team B wins 3-1 in 10 matches, draws 1-1 in 18 matches, and loses 0-1 in 10 matches. Both earn 60 points, but Team A has a goal difference of -52 while Team B has +10. The Pythagorean table would rank Team B far above Team A — and rightly so. Team A's record is built on fragile one-goal wins punctuated by heavy defeats, while Team B's record reflects sustained competitive performance.

How to Use the Pythagorean Table for Predictions

The Pythagorean table is not just an analytical curiosity — it is one of the most reliable prediction tools available to football analysts. Here are three practical applications:

1. Identify regression candidates: Teams with a positive residual greater than 5 points are likely to see their results decline. Their goal difference does not support their league position, and historically, such teams lose an average of 6-8 points the following season. If a team is sitting in a Champions League spot with a negative Pythagorean residual, they may be at risk of dropping out.
2. Spot dark horses for improvement: Teams with a large negative residual are undervalued by the league table. A club in 12th with the goal difference of a top-8 team has the underlying performance to climb. These are the teams worth watching in the transfer window — one or two targeted signings can convert those dominant performances into actual points.
3. Predict head-to-head outcomes: When two teams meet, their Pythagorean expected points per match provides a baseline for the expected outcome. A team averaging 2.1 Pythagorean points per match against a team averaging 1.4 would be expected to win, and the formula can generate approximate win/draw/loss probabilities.

The Limits of the Pythagorean Approach

No model is perfect, and the Pythagorean table has well-documented limitations:

• It ignores match context: A team that scores 5 goals in dead-rubber matches and 1 goal in must-win games will have an inflated Pythagorean expectation. The formula treats all goals equally, regardless of when or against whom they are scored.
• It assumes consistency: The formula works best over large sample sizes. In a World Cup group stage of just 3 matches, a single 6-0 result can massively distort the expected points. The Pythagorean table is a league-season tool, not a short-tournament tool.
• Draws are not modeled directly: Football's three-outcome system (win/draw/loss) is more complex than baseball's two-outcome system. The Pythagorean formula models win probability and multiplies by three, which implicitly assumes draws don't exist. More sophisticated versions adjust for draw probability, but the basic formula remains the most widely used.
• Home and away differences: The formula does not account for home advantage, which in football is worth approximately 0.4-0.5 goals per match. Teams with a strong home record and weak away record may have their Pythagorean expectation distorted.

Beyond the Basic Formula: Advanced Variants

Football analysts have developed several refinements to the basic Pythagorean formula to address its limitations:

Adjusted Exponent by Scoring Environment

Just as baseball analysts developed the Pythagenpat and Pythagenport formulas to adjust the exponent based on runs per game, football analysts can adjust the exponent based on goals per match in a specific league. In the Bundesliga, where the average is approximately 3.1 goals per match, an exponent of 1.35 may be more accurate than 1.3. In Ligue 1, where the average is closer to 2.5, an exponent of 1.25 might fit better.

Expected Goals (xG) Pythagorean Table

The most powerful modern variant replaces actual goals with expected goals (xG). An xG-based Pythagorean table uses the quality of chances created and conceded rather than the actual goals scored. This eliminates the noise of finishing luck — a team that creates 2.5 xG per match but only scores 1.5 is genuinely underperforming their chance creation, and the xG Pythagorean table will identify them as regression candidates even if their actual goal difference looks poor.

Key Takeaways

• Goal difference predicts future performance better than points. The Pythagorean formula, adapted from baseball with an exponent of 1.3 for football, translates goals scored and conceded into expected points that correlate at r = 0.93 with actual results.
• The residual reveals what the table hides. A gap of more than 5 points between actual and expected points signals a team that is either overperforming (positive residual) or underperforming (negative residual) their true quality.
• Regression to the mean is the strongest prediction signal. Teams with large positive residuals historically lose 6-8 points the following season. Teams with large negative residuals gain a similar amount. The Pythagorean table identifies these candidates before the market adjusts.
• The formula works best over full seasons. In short tournaments like the World Cup, a single blowout can distort the calculation. Use it for league predictions, not knockout-stage analysis.
• xG-based variants are the cutting edge. Replacing actual goals with expected goals in the Pythagorean formula eliminates finishing luck and provides an even more accurate measure of underlying team quality.