WPA & leverage: the story stats
A 9th-inning walk-off and a 1st-inning solo shot are the 'same' homer. One flipped the game. Which stat sees that?
At every game state (inning, score, base-out, count) there's a historical win probability. Every play moves it. Credit the mover and you get Win Probability Added: the walk-off might be +0.45 wins, the first-inning shot +0.10. WPA is the stat that watches the game the way fans do.
Leverage: how much the moment mattered
Leverage Index measures how swingy the situation is — 1.0 average, 2.0+ is high drama, garbage time near 0. It's how we know closers pitch important innings and why 'he only pitches the 9th' cuts both ways.
The crucial caveat, and the perfect end to this course: clutch performance barely repeats. Year-to-year WPA-over-expected correlates near zero — great WPA seasons are mostly great hitters getting big moments, plus variance. Use WAR to judge talent; use WPA to tell the season's story. Knowing which stat answers which question — that's the whole skill you've built.
How this stat lies to you
- WPA is a narrative ledger, not a skill estimate — 'clutchness' mostly doesn't persist.
- Hitters can't choose their moments — WPA opportunity is luck-distributed.
- Relievers' WPA is inflated per inning by design (they're deployed at high leverage).
Check yourself
1. Which question does WPA answer well?
2. A hitter led the league in WPA. Next season you should expect…