A pitching tool that grades every arsenal in MLB against its exact peers and flags where a pitcher's usage doesn't match his results. Search any of 563 pitchers, compare all 30 staffs, and track arm-slot and pitch-shape changes from 2025 to 2026.
Starting pitchers, what they do best, and how the other lineup handles it
Games, probable pitchers and lineups load live from MLB. Pitcher and team grades use 2026 regular-season Statcast data. Lineups usually post 2–4 hours before first pitch.
Loading games…
Search any of 563 pitchers for his report card, what makes him nasty, movement, arm angle and what changed since 2025.
Every pitching staff graded pitch by pitch, next to what it costs.
Run value, whiffs, chase, velocity gains and arm slot changes.
How each shape moves, a common grip and who throws it best.
Pitching and prospect analysis published at Pitcher List.
Ten starters and ten relievers, 29 or younger, whose pitches grade better than their 2026 numbers. Tested on four past seasons first.
New to the numbers? Read the glossary · About Ben
Analysis for Pitcher List · 5 articles · select one to read
A 22-year-old who hit 48 homers across High-A and Double-A, in an organization built on contact. Does his power fit, or does the strikeout rate win out?
Read on Pitcher List →His results fade the second time through a lineup, and his stuff jumps in short bursts. The case for relief, and what still has to improve.
Read on Pitcher List →Tanner Bibee and Parker Messick win with middling Stuff+ grades. Contact management, sequencing and pitch mix as an organizational pattern.
Read on Pitcher List →His arm angle dropped from the steep slot he used in the minors, and the whiffs went with it. The new cutter as part of a bigger plan.
Read on Pitcher List →A new cutter, fewer four-seamers and a bullpen role. What Senzatela's turnaround says about how Colorado now builds pitchers.
Read on Pitcher List →Projects · Python · Statcast
Grades every pitch a pitcher throws on stuff (Whiff+), location (Location+) and results (run value) against the same pitch, hand and batter side, then flags where his usage and his results don't match. It powers Pitcher Search, Teams and the Leaderboards.
Read the notebook →View on GitHub →When a pitcher's arm slot moves, what happens to his movement and his results? 60 pitchers moved 4° or more from 2025 to 2026, starting with Emerson Hancock and Jonah Tong.
See the arm slot leaderboard →Pitch Mix Tool · every pitcher with 300+ pitches in a season, 2021 to 2026
Average break of each pitch, seen from behind the pitcher, so a lefty’s chart is a mirror image of a righty’s. Up is more rise. Dot size shows how often he throws it.
| Pitch | Aheadmore strikes | Even | Behindmore balls |
|---|
| Pitch | Usage 2025 → 2026 | Velo | IVB | HB | Runs/100 | Changes |
|---|
Graded against every 2026 MLB pitch of the same type, thrown with the same hand to the same batter side. Whiff+ and Location+ of 100 are league average. Location+ grades where he put each pitch compared with an average spot for that pitch type, batter side and count. Under Runs/100, the small line splits it into strikes (Str), balls (Ball) and contact (Con); the three add up to the total. Runs/100 is run value per 100 pitches versus league; positive saved runs. Location view: Zone% is the share of pitches in the strike zone, Chase% is the swing rate on pitches outside it, and in-zone whiff% is misses on swings at strikes, each shown against league. Velo is average release speed in mph. IVB is induced vertical break and HB is horizontal break, both in inches; positive HB breaks toward the pitcher’s arm side. xwOBAcon is expected damage on balls in play; negative is better for the pitcher, and it is grayed out under 20 balls in play. Pitches thrown fewer than 10 times to a side or in a count are left off. In the count table, usage is the share of his pitches in that count, and gray marks fewer than 30 swings.
Young arms whose stuff is ahead of their results
Breakouts usually show up in the stuff before they show up in the results. This list looks at 2026 pitchers who are 29 or younger, threw 600+ pitches, were average or worse (+0.3 runs per 100 pitches or less) and have never had a strong season since 2021, then ranks starters and relievers separately by projected 2027 run value. It's a watch list, not a promise: most of these pitchers won't break out.
It's a weighted blend, fit on every pitcher with 300+ pitches in back-to-back seasons from 2021 to 2026, that predicts next year's runs per 100 pitches. It leans most on Stuff+ (how good the pitches are, ignoring results) and Whiff+ (missed bats against the same pitch type), then Location+, a fastball velocity gain from the year before, and contact quality. This season's run value carries little weight, because one season of results is mostly noise: Stuff+ and Whiff+ each predicted the next season better than the results did.
The test: for each season from 2022 to 2025, the model was built using only earlier seasons and picked its top 10 starters and top 10 relievers for the next year. A breakout means the next season was worth 10+ runs above average and at least 8 runs better than the year before. "Never had a strong season" means no season since 2021 worth 10+ runs above average, at +1.0 runs per 100 pitches or better, or with an arsenal grade of 60+. That keeps veterans bouncing back from a bad year off the list.
Age decides who's eligible (29 or younger) rather than counting as a signal. Adding age and years in the majors to the model didn't improve it consistently in testing, so they're shown but not weighted.
What it can't see: injuries and health history, role changes, a new team or a pitch added in the offseason. The reliever list has tested better than the starter list. In 2021, MLB's mid-season crackdown on sticky substances changed some pitchers' spin. Check back next October to see how this list did.
| List | Pitcher | Age | Runs/100 that year | Next year | Broke out? |
|---|
What each pitcher could change · how often to throw each pitch, and how to reshape it
Pitch mix. Each pitch gets a value against lefties and righties: its own 2026 run value, trusted more as the sample grows, blended with its Stuff+ grade. The optimizer moves usage toward the most valuable pitches until the extra value runs out, because a pitch gets less effective the more hitters see it (each extra point of usage is charged against that pitch). Guardrails: no pitch moves more than 20 points, a pitch needs 50+ thrown to that side before it's suggested more, and every pitcher keeps a real fastball. Season runs use his 2026 workload to that side.
Did the mix optimizer work in the past? Run on 2025 data only, it suggested real changes in 452 pitcher-vs-side matchups. Pitchers who moved toward its advice in 2026 held steady (+0.03 runs/100, 53% improved); pitchers who moved away got worse (−0.24, 37% improved). Most of that gap is regression to the mean, though: after controlling for how good each mix was in 2025, following the advice was worth only about +0.03 runs/100. The honest read: it points the right way, and one season isn't enough to prove by how much.
The Bryan Baker test. Given only his 2025 numbers, the optimizer said to throw righties fewer sliders (39% → 26%) and more changeups (17% → 32%). The Rays went further in the same direction: 6% sliders and 38% changeups in 2026, and he became an All-Star closer. Against lefties it agreed on more changeups but wanted more sliders, too; the Rays nearly scrapped the slider.
Diminishing returns. The data alone says pitches got better when thrown more (+0.03 runs/100 per usage point). That's backwards: teams throw a pitch more once it's already working, so cause and effect are tangled. The optimizer uses a cautious charge instead: +10 points of usage costs that pitch 0.30 runs/100.
Pitch fixes. One small change at a time (1 mph, 2 inches of break, 150 rpm or 0.3 ft of extension), re-scored with Stuff+ and allowed only if pitchers with a similar arm slot already throw that version, plus the typical version of a pitch he doesn't throw. Backtested on 2025→2026 pitch changes: the model predicted which pitches gained or lost whiffs (r = 0.23; 62% right on meaningful changes), but year-to-year run value was too noisy to confirm.
What it can't know. Whether his body can make a change, how pitches play off each other in sequence, or what a hitter is sitting on. Treat these as ideas to test with video and a coach.
All 30 pitching staffs, ranked by run value
| Pitcher | Throws | Pitches | Run value | Best pitch |
|---|
Run value is runs saved per 100 pitches compared with league average; positive is good. Pitchers are grouped by their team as of October 2026. Payroll from Spotrac; Cleveland's wasn't available. Select a pitcher to open his card.
Single pitches against one batter side · 2026 · select a row to open the report card
| # | Pitcher | Pitch | Usage | Value |
|---|
Every 2026 pitch, sorted into 16 shapes by velocity and movement
Every term used on the site
Pitching analysis and tools
I work in ticket sales for the Brooklyn Cyclones at Maimonides Park and write pitching analysis for Pitcher List's Going Deep section. I also consult independently for a junior college program, working with its TrackMan data.
Kafkaesque Baseball started as one question: is a pitcher using his pitches the right way? It grew into a tool that answers that for every pitcher in MLB. I designed the analysis and built it in Python with pybaseball, learning the language as I went and working with Claude as a coding partner. Every method, threshold and limit is written out in the notebook.