Data for Thought Data for Thought

Sports Forecasts and Analyses

Sports are a perfect lab for learning what data can and cannot predict. They produce a steady stream of measurable outcomes, and they make it hard to hide from a bad forecast. Looking for patterns in historical data also makes the games more fun to watch and seasons more fun to follow. I build models for the NWSL, Premier League, NFL, and World Cup. These dashboards are where I publish forecasts, compare them with the market, track how the models perform, and explore how teams, players, and competitions change over time. Along the way, I explain what the numbers can and cannot tell us.

Championship probability trends for the NWSL contenders
NWSL

National Women’s Soccer League

Match forecasts, the championship race, and the scorecard.

Games · Table · Scorecard
Live · Updated dailyOpen dashboard →
NFL teams mapped by offensive and defensive strength
NFL

National Football League

Weekly win probabilities and spreads, graded against the closing line.

Games · Standings · Scorecard
Season begins September 10Open dashboard →
A Premier League game card with win probabilities from several models
Premier League

English Premier League

Match probabilities, the table, and the Golden Boot race.

Games · Table · Scorecard · Golden Boot
Season begins August 21Open dashboard →
The World Cup Golden Boot race, won by Mbappé
World Cup 2026

2026 World Cup Archive

Every game, bracket, and grade from the 2026 tournament. Spain, champions.

Games · Bracket · Scorecard · Golden Boot
Complete · Tournament archiveOpen dashboard →