Pythagorean Win Expectation

Overview

This pillar analyzes a team's point differential to estimate how many games they should have won, identifying 'lucky' overachievers and 'unlucky' underachievers. It provides a more accurate measure of a team's true strength than their simple win-loss record.

What It Does

Pythagorean Win Expectation uses a formula, originally developed for baseball but adapted for football, to convert a team's points scored and points allowed into an expected winning percentage. This reveals teams whose records are inflated by close wins or deflated by narrow losses. It helps forecast future performance by assuming teams will eventually regress to their statistical mean.

Why It Matters

A team's point differential is a more stable and predictive indicator of future success than its current win-loss record. This pillar provides an edge by spotting teams that the market has mispriced based on a misleading record, creating opportunities for betting on their future progression or regression.

How It Works

First, we gather a team's total points scored (PF) and points allowed (PA) over a specific period, usually the current season. Next, we apply the NFL-specific formula: Expected Win % = (PF^2.37) / (PF^2.37 + PA^2.37). This percentage is then multiplied by the number of games played to determine the team's expected wins. Finally, we compare this to their actual wins to find a 'luck factor'.

Methodology

The core calculation is Win % = (Points For ^ 2.37) / ((Points For ^ 2.37) + (Points Against ^ 2.37)). The exponent of 2.37 is an empirically derived constant optimized for the scoring environment of the NFL. The 'Luck Factor' is calculated as (Actual Wins - Expected Wins). Analysis is typically run on season-to-date data.

Edge & Advantage

This model uncovers value by identifying teams whose betting odds are based on a misleading win-loss record, allowing for trades before the market corrects.

Key Indicators

• Expected Wins

  high

  The number of wins a team's point differential suggests they should have.

• Luck Factor

  high

  The difference between actual wins and expected wins. A high positive number suggests luck; a high negative number suggests unluckiness.

• Point Differential

  medium

  The net difference between total points scored and total points allowed.

Data Sources

• Pro-Football-Reference

  Provides comprehensive historical and current season statistics for all NFL teams, including points for and against.

• TeamRankings

  Offers various NFL stats including points per game data and other advanced metrics.

Example Questions This Pillar Answers

• → Will the Las Vegas Raiders win over 8.5 games this season?
• → Which team will win the AFC North division?
• → Will the Atlanta Falcons make the playoffs this year?

Tags