Stuff+ is a pitch grade built from how the ball flies. It estimates how much damage a pitch with that speed, movement and release would allow on average, then scales the result so 100 is league average [1][2]. It ignores where the pitch was thrown and what pitch came before it. That narrow focus is both its biggest strength and its biggest weakness.
What goes into Stuff+?
Several public and private versions exist. Driveline's 2024 update is one of the best documented, and it uses these inputs [1]:
- Velocity and movement. Speed plus vertical and horizontal break.
- Release traits. Arm angle and extension.
- Approach angles. Since 2024, location-adjusted vertical and horizontal approach angles, the angle at which the pitch arrives at the plate.
- Differences from your fastball. Breaking balls and off-speed pitches are judged against your primary fastball, which is now defined separately for right-handed and left-handed hitters.
A few design choices matter for interpreting the number. Scores are scaled within three buckets (fastballs, breaking balls, off-speed), so, in Driveline's words, "A 110 slider is not the same as a 110 sinker." Velocity is treated as exponential, so elite velocity earns a widening bonus. In that model, sinkers grade above four-seamers until about 97 mph [1].
Approach angle can rescue a pitch that looks plain on a movement chart. Driveline's example is Alexis Díaz: about 13 inches of vertical break and 9.5 inches of horizontal break, a shape often called a "dead zone" fastball, still graded above average because of 7.7 feet of extension and a low release [1].
Why Stuff+ is useful: it settles fast
The best argument for a stuff model is speed. FanGraphs found its Stuff+ becomes reliable about 80 pitches into a season, while its Location+ grade needs about 400 [2]. Shape is sticky from start to start and season to season; location bounces around. That makes a stuff grade a strong early signal for a new pitch, long before you have enough game data to judge results.
Bannister frames the value well for developing pitchers: better stuff increases your margin for error while command catches up through repetition [3]. A pitch with good shape survives more mistakes than a pitch without it.
Stuff+ vs. Location+ vs. Pitching+
FanGraphs publishes three related grades, and knowing the difference prevents a lot of bad arguments [2]:
| Grade | What it measures | Best use |
|---|---|---|
| Stuff+ | The physical pitch: speed, movement, release | Early read on a new or changed pitch |
| Location+ | Where pitches are thrown, given the count and pitch type | Judging command over a longer sample |
| Pitching+ | Both combined | Overall pitch quality once enough pitches are in |
For relievers, FanGraphs found Pitching+ beats preseason projections after about 250 pitches; for starters, after about 400, roughly four or five starts [2]. In other words, a stuff grade is the first number worth reading on a new pitch, and the last one you should rely on alone.
What Stuff+ leaves out
The blind spots are where pitchers get into trouble.
- Location and command. Locations are excluded by design. Driveline itself notes that stuff is only one component of pitching [1].
- Count and game state. An average grade ignores when the pitch is thrown. Driveline's newer research grades pitches in each of the 288 combinations of count, base runners and outs, and finds some pitches that win specific situations despite a modest average grade. In some situations, a ground-ball pitch beats a whiff pitch [4][5].
- The rest of your arsenal. A pitch's value depends partly on what it is paired with, which a single-pitch grade cannot see [6].
- Hitter adaptation. Hitters adjust as shapes spread, so the models are retrained regularly. Sweepers had the second-highest whiff rate of any pitch type in 2021, and by 2024 whiffed at rates closer to changeups and curveballs [1]. Driveline's coaches describe the newest models as favoring hard breaking balls [7].
Rare pitches are another weak spot. A model learns from pitches it has seen, so a truly unusual shape has few comparisons. Tatsuya Imai's arm-side "reverse" slider carried only an average stuff grade before he came to MLB, which analyst Chris Langin argued badly understated it [8]. In 2026 that slider produced a 39.7% whiff rate over 660 pitches [9].
Plot your arsenal
See your pitches side by side in Pitch Lab
Pitch AI's Pitch Lab plots your arsenal on a movement chart from the velo, IVB, spin, axis and release you log, flags pitches that overlap, and ranks what to develop next.
Logan Webb: grade down, results up
The clearest example of a stuff grade missing the bigger picture is Logan Webb. Driveline reported that Webb's overall Stuff+ fell to a career low in 2025 [6]. That same season he set a career high with 224 strikeouts, up from 172 in 2024, and his strikeout rate climbed from 20.5% to 26.2% [10].
What changed was his fastball menu. Webb had added a cutter in 2024 and raised its usage to 7.9% in 2025 while bringing his four-seamer back at 7.8%. Driveline found his contact rate on pitches outside the zone fell 10.8 points, the second-largest drop among qualified pitchers [6]. More fastball looks made his sinker and changeup harder to sit on. The grade, which scores each pitch on its own, could not see that.
Shape alone can also swing the grade a long way. A Driveline coach describes Dylan Cease's fastball grade jumping from 150 to 182 in Driveline's model in 2021 even though he threw almost 1 mph slower, because he added about three inches of ride [11]. That is the company's own model and has not been independently checked, but it shows how much shape alone can move the number.
How should a pitcher use Stuff+?
Use it the way it was built: as a fast early read on whether a shape has promise. Then let command and real hitters decide.
- Compare within one model. Grades from different providers, versions or pitch buckets are not interchangeable. Keep the model name and version with every score [12].
- Use it to rank candidates. When testing two grips, a stuff-style comparison of velocity, movement and release tells you which one deserves more reps.
- Do not scrap a pitch only because it grades average. Webb's added fastballs and Imai's slider both outperformed their grades.
- Do not keep a pitch only because it grades well. If you cannot locate it, the grade does not matter in a game.
One more caution for amateurs: these models are trained on professional pitchers and hitters. A 14-year-old's numbers do not translate directly, and chasing a pro grade can push a young pitcher toward shapes he cannot command yet. The principles transfer better than the numbers.
What to do this week
- If you have tracking data, sort your pitches by how different each one is from your main fastball in speed and movement.
- Pick the pitch with the best shape that you also command reasonably well, and plan to use it more.
- If a new pitch grades well in bullpens, start logging where it lands before you call it a weapon.
For the fastball inputs these models start from, see how to throw a four-seam fastball. For the arsenal effect the grade misses, read why adding another fastball shape can help and our pitch design workflow. Earlier in this series, we covered how to tell if a new pitch is actually good; next, we look at whether you should lower your arm slot.
Frequently asked questions
What is a good Stuff+?
By design, 100 is league average, and higher is better. Each provider builds and scales its model differently, and Driveline scales fastballs, breaking balls and off-speed pitches separately, so compare scores only within the same model and pitch category.
Does Stuff+ include location or command?
No. Stuff models leave location out on purpose and grade the physical characteristics of the pitch. FanGraphs publishes a separate Location+ grade, and its combined Pitching+ uses both.
Why did Logan Webb's Stuff+ drop while his strikeouts went up?
Driveline reported that Webb's overall Stuff+ fell to a career low in 2025, yet he struck out 224 batters, up from 172, after adding more fastball looks. A pitch-by-pitch grade cannot see how the whole arsenal works together or how pitches are used by count.
Should high school pitchers care about Stuff+?
Lightly. The models are trained on professional hitters and pitchers, so the numbers do not transfer directly to amateur baseball. The useful idea is the same at any level: shape gives an early read on a pitch, while command and results take far longer to judge.
Use the grade the right way
Track shape early, then let results decide
Pitch AI's Pitch Lab plots your arsenal and ranks what to develop next. Log each pitch's metrics and every grip change, then judge the new shape against your own baseline as you move from bullpens to live hitters.
References
- Lambert J (2024). Revisiting Stuff+: An Update on Driveline's Methodology to Quantifying Pitch Design. Driveline Baseball, May 31, 2024. https://www.drivelinebaseball.com/2024/05/revisiting-stuff-plus/
- McGrattan O (2023). Stuff+, Location+, and Pitching+ Primer. FanGraphs Library, March 10, 2023. https://library.fangraphs.com/pitching/stuff-location-and-pitching-primer/
- Tread Athletics (2022). Brian Bannister: Mastering Pitch Design (podcast). YouTube, December 22, 2022. https://youtu.be/E0yoPt42A5Q
- Hejka J, Lambert J, Ramilo M; Driveline Baseball (2025). Projecting Pitch Design, Saberseminar 2025 presentation. YouTube, December 31, 2025. https://youtu.be/HeBvjYrU_vk
- Driveline Baseball (2026). The Evolution of Pitch Design, Saberseminar 2025 (R&D Podcast Ep. 111). YouTube, January 19, 2026. https://youtu.be/2fjpL_BraVU
- Sawchik T (2025). How many fastballs should a pitcher have on the menu? Driveline Baseball, November 25, 2025. https://www.drivelinebaseball.com/2025/11/how-many-fastballs-should-a-pitcher-have-on-the-menu/
- Driveline Baseball (Lambert J, White C) (2026). What's the Next Sweeper? YouTube podcast, July 28, 2026. https://youtu.be/QPzFR4GNxdk
- Langin C (2025). Tatsuya Imai Deserves $200M+: The Inflection Point for Modern Pitcher Evaluation. Unfiltered Labs. https://unfilteredlabs.com/blog/tatsuya-imai-deserves-200m-the-inflection-point-for-modern-pitcher-evaluation
- Baseball Savant, MLB Advanced Media. Statcast pitch-level search and Pitch Arsenal Stats leaderboards, 2021 to 2026 seasons; retrieved and summarized by Pitch AI on October 14, 2026. https://baseballsavant.mlb.com
- MLB Stats API. Season, monthly split and transaction data for the pitchers named; retrieved by Pitch AI on October 14, 2026. https://statsapi.mlb.com
- Driveline Baseball (Langin C) (2022). How to Throw a Four Seam Fastball. YouTube, May 26, 2022. https://youtu.be/YQkUrg4obaA
- Pitch AI (2026). Pitch Design Playbook, with Pitch Trial Log and Pitch Decision Log templates. Internal coaching synthesis, September 2026.