Win Probability Added (WPA)
Also known as: win expectancy, win probability, win probability chart, what is WPA
Win Probability Added (WPA) measures how much a play changed a team's chance of winning. It is the team's win expectancy after the play minus its win expectancy before it.
WPA starts from win expectancy. MLB's glossary defines it as a team's chance to win at a given point in a game. It compares the score, inning, outs, runners and run environment with similar past games. WPA is the change in that chance on each play. A hitter gains WPA for helping and loses it for hurting. The pitcher gets the mirror image. Steals, caught stealing and pickoffs count too.
MLB's glossary gives a famous case. On May 26, 2015, the Blue Jays trailed by two in the bottom of the ninth. They had runners on second and third with nobody out, a 43% win expectancy. Josh Donaldson hit a walk-off home run to win the game. That added 57 percentage points of win expectancy, so his WPA for that plate appearance was 0.57.
FanGraphs calls WPA a good way to tell the story of a game. Season totals are not predictive of talent. In FanGraphs' rough chart, a season WPA of about +1.0 is average for an MLB regular and +6.0 is excellent.
Example
With two outs in the ninth and his team down a run, a batter doubles in two runners. His team's chance of winning jumps sharply. If win expectancy rose from 25% to 75%, his WPA for that at-bat would be 0.50.
Why it matters
WPA credits the hits that swing games. It explains who won a game, in a way box-score totals cannot.
Common mistakes
- Using season WPA to predict next season. FanGraphs says it is not predictive.
- Forgetting that a player can lose WPA, for example by being caught stealing.
- Comparing WPA between players with very different playing time. It is a cumulative stat.
Frequently asked questions
What is win expectancy?
MLB's glossary defines it as a team's chance to win at a given point, based on how often teams in similar past situations went on to win.
Is WPA a good measure of talent?
No. FanGraphs says season WPA is not predictive. It describes what happened and who helped win.
Sources
Related terms
- Leverage Index (LI)Leverage Index (LI) measures how much a game situation can swing a team's chance of winning. 1.0 is an average moment, and higher numbers mean more is at stake.
- Run Expectancy and RE24Run expectancy is the average number of runs scored from a given base-out state to the end of the inning. RE24 adds up how much each play changes it, plus any runs that score.
- Wins Above Replacement (WAR)Wins Above Replacement (WAR) estimates how many wins a player adds compared with a freely available replacement player. It combines hitting, baserunning, fielding and position into one number.
- Pythagorean Winning Percentage (Expected W-L)Pythagorean winning percentage estimates how often a team should win from its runs scored and allowed. Bill James's original formula is runs scored² ÷ (runs scored² + runs allowed²).
Available in other languages
Related guides & benchmarks
Ready to work on your own swing?
This term is background, not something video measures. Film one swing to find the one thing to work on first — free to start.