Pitch-grade leaderboard
Stuff+ grades a pitch only on physical traits the pitcher controls: velocity, movement, spin, release point, extension, arm angle, and how the pitch differs from his primary fastball. 100 is an average qualified MLB pitcher from the season before (2025), and every 10 points is one standard deviation. The scale is frozen for the whole season, so a grade means the same thing in April as in September. Location+ grades only where each pitch crossed the plate, in the count it was thrown, and Pitching+ grades both together; they use the same scale. RV+ is actual run value on the same kind of scale, so you can see who got results the grades don’t explain, and vice versa. Pick a grade to sort, chart, and track it. Move the date slider to see the board as it stood on any week, or sort by the change column to find who’s improving. How the models work and how well they predict: Stuff predicts next season better than results do and Stuff or command?
The league over the season
2026 regular season, weekly snapshots (the last game on or before each Sunday, plus the final day). Triple-A pitchers are graded on the same MLB-trained model, so their numbers are directly comparable with big leaguers’. Triple-A RV+ is measured against weaker hitters, so expect it to run high. Location+ and Pitching+ are frozen to the same 2025 pitchers. Roles (SP, RP, Swing) are as of the selected date and use the role-based pitch minimums in the glossary; earlier in the season the minimums are prorated to games played. How Stuff+ is built; how Location+ and Pitching+ are built.