Loading…

Select a hitter to begin

Use the Hitter Name filter on the right to load the dashboard.

Zone%	Decisions	Zone Swing%	Chase%	xContact	Adj. Smash Factor	Contact%	Zone Contact%	Out-of-Zone Contact%	Vertical Plate Coverage	K%	BB%	Early-Count Swing%	Early-Count BIP%	2-Strike Chase%
—
League Averages
—

Run Value

—

Baseball Scouting Lab

Loading…

Select a hitter to begin

Use the Hitter Name filter on the right to load the dashboard.

xISO CON	xDamage	DHH%	Swing Acc (ft/s²)	Bat Speed 90	EV90	Barrel%	HH LA	Attack Direction	Spray Angle
—
League Averages
—

Run Value

—

Baseball Scouting Lab

Loading…

Select a pitcher to begin

Use the Pitcher Name filter on the right to load the dashboard.

Stuff	Location	Pitching	K%	BB%	K-BB%	Zone%	SwStr%	Chase%	Adj. GB+PU%	Exit Velo	Barrel%	Arm°	Rel Ext
—
League Averages
—

Run Value

—

Baseball Scouting Lab

Loading…

Select a pitcher to begin

Use the Pitcher Name filter on the right to load the dashboard.

Tunneling	Diversity (Movement)	Diversity (Velocity)	Repertoire	Fastball Usage%	Early Count FA%	Hitter Ahead FA%	2-Strk FA%
—
League Averages
—

Run Value

—

Baseball Scouting Lab

Loading…

—

Season

Metric

Group By

Hitter Hand

Pitcher Hand

Legend

BASEBALL SCOUTING LAB

Baseball analytics straight from Statcast, for fans and front offices alike.

Dashboards Per-player deep dives with heatmaps, histograms, and interactive charts

Models Custom 20-80 scale and raw models built from pitch-level Statcast data

Leaderboards Sortable, filterable player tables across 13 stat categories

Visuals Player and team-level plots showing value, pitch & swing mechanics, and more

Tools Play finders, pitch shapes, similarity scores, and an interactive strike zone.

Select filters to load plays

Select filters to load pitches

Loading…

Loading data…

Mode

Season

Pitcher

Pitcher Throws

Batter Stands

Count

Pitch Group

Pitch Type

Contact Quality

Pitch Result

Select a player to render the strike zone.

On this page

Model Documentation

How the Models Work

This site combines original machine-learning models with raw Statcast data to grade pitchers and hitters across multiple dimensions. This page explains what each model measures, how it was built, and how to interpret the scores — no data science background required.

LightGBM + XGBoost 2021–2025 Statcast Pitch-level training Batter-adjusted target + shrinkage 50 = league average ±10 pts = 1 SD, floor 20, ceiling 80

The Big Picture Overview

What is the 20-80 scouting scale?

The 20-80 scale is the universal language of professional baseball scouting. For decades, scouts have graded every tool — fastball, hit, power, speed, arm, field — on a scale from 20 (well below average) to 80 (generational), with 50 representing a league-average player. The grades follow a roughly normal distribution: a 70 or 80 is a true impact tool seen on a fraction of big-league players, 60 is a plus tool, 50 is average, 40 is below average, and 30 is fringe. Every model score on this site adopts that same convention so grades mean the same thing whether you're reading a scouting report or a stat table. 50 is exactly league average. Each point represents roughly one-tenth of a standard deviation — so a score of 70–80 is elite, 60 is above average, 40 is below average, and 30 is poor. Scores are hard-capped at 80 (ceiling) and 20 (floor), and are recalculated each season so the scale is always relative to that year's league. Higher is always better — for pitchers and hitters.

Pitcher Models

Three interlocking scores describe every pitcher: Stuff measures the raw nastiness of pitch physics, Location measures command, and Pitching combines both into an overall grade. Each score is calculated per pitch type — a pitcher's fastball Stuff is completely separate from his slider Stuff.

Hitter Models

Three models predict outcomes on every swing and take: Decision predicts the run-value quality of each swing-or-take choice, xContact predicts whether a swing will result in contact, and xDamage predicts whether a batted ball will be solid contact or a barrel. The scores reflect what the models estimate, not directly observed outcomes.

Raw Statcast

Pitch physics, location, count, handedness, bat tracking

→

LightGBM / XGBoost Models

Pitcher: LightGBM, random 60/40 split; Swing/Decision: XGBoost, 2023–24 train / 2025 val

→

Per-pitch Scores

Aggregated by player-season with shrinkage

→

Plus Scale

50 = avg, 10 pts = 1 SD, 20–80 scale, normalized per season

Pitcher Models Stuff · Location · Pitching

Target Variable: Count-Neutral Run Value

Every pitch in baseball has a run value — a number that says how much that pitch changed the expected run total of the game. A swing and miss at a 3-2 pitch is worth more than one at 0-0 because the count matters. To isolate pitch quality from count luck, the models use count-neutral run value: the expected run value of that pitch outcome averaged across all count states. This way a pitcher who happens to throw breaking balls in hitter's counts isn't penalized for things outside his control.

Architecture: Three Separate Models per Pitch Group

Fastballs behave differently than breaking balls, which behave differently than changeups. Rather than training one model on all pitch types, three separate LightGBM regressors are trained — one for fastballs (four-seamers, sinkers, and cutters), one for breaking balls (sliders, curveballs, sweepers, and knuckle-curves), and one for offspeed pitches (changeups and splitters). Each model has its own hyperparameters tuned specifically for that pitch family. All three pitch models use a random 60/40 train/validate split across all years (2021–2025), with batter-adjusted count-neutral RV as the target and James-Stein shrinkage applied to pitcher-season aggregates. Within each group, a location-only sub-model is also trained, capturing pure command independent of stuff.

0.061

Fastball RMSE

0.065

Breaking RMSE

0.063

Offspeed RMSE

r ≈ 0.82

Year-to-Year Stability (shrunk)

What the model sees

plate_x / plate_z Where the pitch crosses the plate, horizontally and vertically Location

ax / ay / az Acceleration vectors — the aerodynamic signature of the pitch. Top stuff feature after location for all three pitch groups. Stuff

flight_time Time from release to plate — derived from kinematic equations. Captures velocity and extension combined into a single timing signal. Stuff

arm_angle Arm slot derived from release position — the horizontal angle of the pitcher's arm at release Stuff

release_spin_rate Spin rate at release — affects magnitude and direction of movement Stuff

approx_spin_eff Approximate spin efficiency — ratio of transverse (Magnus-effect) spin to total spin, estimated from aerodynamic lift forces. 100% = pure backspin/topspin; 0% = pure gyrospin with no movement contribution. Stuff

vaa (vertical approach angle) The angle at which the pitch arrives at the plate. Steeper = harder to hit. Present in all three pitch group models. Stuff

flight_time_diff_fb / ax_diff_fb / az_diff_fb Flight time and horizontal/vertical acceleration differentials vs. the pitcher's own fastball profile. Breaking and offspeed models only — the primary deception signal; pitches that match the fastball's flight signature before breaking are hardest to read. Stuff

prev_pitch_group What the previous pitch was — sequencing context Context

p_throws / stand Pitcher handedness and batter stance — platoon splits Handedness

altitude_ft Ballpark elevation — affects ball movement at altitude (e.g. Coors Field) Environment

Note on location and context

Across all three pitch groups, location (plate_x and plate_z) is a strong predictor of pitch outcome. The the model includes arm angle, flight time, spin efficiency, and fastball differentials as core stuff features across all three pitch groups. Sequencing context (what the previous pitch was) is also a stronger signal than most stuff characteristics. The stuff qualities below describe what differentiates pitches within the same location.

Stuff characteristics that matter by pitch type

Fastballs

For pure stuff, vertical approach angle (VAA) is the standout characteristic — steeper VAA suppresses contact independent of location. Flight time, arm angle, spin efficiency, and acceleration vectors (ax/ay/az) round out the fastball feature set. Raw velocity matters less than the shape and angle of the delivery.

Breaking Balls

For breaking balls, deception is the primary driver. The model includes all nine fastball features plus three differentials vs. the pitcher's fastball: flight-time difference, ax differential, and az differential. A pitch that arrives later or moves differently from the same arm slot as the fastball is harder to pick up.

Offspeed

Offspeed pitches share the same nine-feature base as fastballs plus the three fastball differentials. The az differential (vertical acceleration gap from the fastball) is the primary deception signal — a changeup that mimics the fastball's flight line before dropping is hardest to time. Batter stance remains an important modifier for arm-side movement.

Year-to-year stability (shrunk Stuff xRV/100, all consecutive season pairs)

Pitch Group	YoY Pearson r	Sample Size	What it means
Fastball	r = 0.836	1802 pairs	Highly stable — fastball stuff is a true pitcher skill
Breaking Ball	r = 0.802	1387 pairs	Strong stability — breaking ball stuff is repeatable
Offspeed	r = 0.820	699 pairs	Stable across all consecutive season pairs

What Stuff is NOT

Stuff does not measure results. It also does not account for location or arsenal effects — a pitch with elite physical characteristics thrown in the middle of the plate will still grade poorly in Pitching. Stuff isolates the intrinsic physical quality of a pitch: what would happen on average if that pitch were thrown in a neutral count to a random hitter, independent of where it landed and what came before it. Full outcome prediction lives in Pitching.

Hitter Models Decision · xContact · xDamage

Decision — The Swing/Take Model

Every pitch a hitter sees is either swung at or taken. Decision grades that choice. For every pitch, the model calculates two things: what was the expected run value of swinging? And what was the expected run value of taking? Decision RV is the difference — did the batter choose the higher-value option? Positive means they made a good decision; negative means they would have been better off doing the opposite. Decision aggregates those choices across a season, shrunk toward the mean to account for sample size.

Called-Strike Probability Model

For taken pitches, the model needs to know: how likely was that pitch to be called a strike? This is answered by a separate XGBoost classifier trained exclusively on taken pitches. It achieved an AUC of 0.9845 and a Brier score of 0.0463 on 2024 validation data — meaning it's nearly perfectly calibrated. Edge pitches at ~50% CS probability are valued accordingly, rewarding hitters who lay off borderline pitches.

Swing RV Model

For swings, a second XGBoost regressor directly targets the run value of swinging at any location, in any count, against any handedness. Earlier iterations of the model approximated swing value using zone-based lookup tables — grouping pitches into in-zone vs. out-of-zone buckets with fixed run values. The current approach replaces that with a fully learned continuous model, which better captures how swing value fades smoothly from the heart of the plate outward rather than dropping off a cliff at the zone edge. The model also picks up count leverage and approach-angle interactions that a lookup table can't represent.

Decision model features

plate_x / plate_z Horizontal and vertical location — the dominant signals for both swing and take value Location

balls / strikes Count state — determines leverage of the swing-or-take decision. 3-0 vs. 0-2 changes everything Count

vaa / haa Vertical and horizontal approach angles — affect where the pitch looks vs. where it ends up Approach Angles

p_throws / stand Handedness — swing/take thresholds differ significantly for same-side vs. opposite-side matchups Handedness

0.9845

CS Model AUC

0.0463

CS Brier Score

r = 0.94

Decision YoY Stability

n ≈ 400

Qualifying hitters/yr

xContact — The Bat-to-Ball Model

xContact uses Statcast's bat-tracking data, which is only available from mid-2023 onward — xContact and xDamage cannot be calculated for earlier seasons. The model combines swing mechanics — bat speed, swing length, attack angle, attack direction, and swing path tilt — with pitch location and approach angles. It trains on 2023–2024 data and is validated on 2025. This captures genuine bat-to-ball skill: the same borderline pitch produces very different contact probabilities depending on a hitter's mechanics, even at identical pitch locations. AUC ranges from 0.84 (fastball) to 0.92 (breaking/offspeed) on 2025 validation data.

xDamage — The Quality-of-Contact Model

xDamage predicts the probability of solid contact or a barrel on each swing that produces a batted ball. It targets Statcast's solid contact and barrel designations (launch_speed_angle codes 5 and 6). It's worth noting that this is one reasonable definition of quality contact — not the only one. High exit velocity, optimal launch angle, and hard ground balls can all represent quality contact that doesn't register as a Statcast barrel. Bat speed is the dominant feature, followed by attack angle and contact depth. YoY stability: r = 0.956 contact, r = 0.956 damage on 2024→2025 OOS predictions.

xContact and xDamage model features

bat_speed Swing speed at contact — the most important xDamage feature across all pitch groups Swing Mechanics

attack_angle Vertical swing plane angle — determines if the bat path matches the pitch trajectory Swing Mechanics

attack_direction Horizontal swing direction — pull vs. oppo tendencies, critical for xContact on offspeed Swing Mechanics

swing_length / swing_path_tilt Compactness and tilt of the swing path — two-strike adjustments show up here Swing Mechanics

contact_depth_x / contact_depth_y Contact depth — where relative to the body the hitter makes contact (intercept_ball_minus_batter_pos_x/y_inches) Swing Mechanics

plate_x / plate_z Pitch location — controls for the difficulty of the pitch being swung at Location

vaa / haa Approach angles — how the pitch arrives affects how hard it is to square up Approach Angles

0.92

xContact AUC (Breaking)

0.80

xDamage AUC (Offspeed)

r = 0.97

xContact YoY Stability

r = 0.96

xDamage YoY Stability

How the three hitter scores fit together

Decision, xContact, and xDamage are independent — a hitter can be great at one and poor at another. A power hitter might have elite xDamage but average Decision. A contact hitter might have excellent Decision and xContact but low xDamage because they prioritize putting the bat on the ball over swinging hard. Together, the three scores give a complete picture of a hitter's plate approach.

Validation and Predictiveness Does it actually work?

Out-of-sample validation — the only number that matters

All accuracy statistics reported here are measured on data the models never saw during training. The pitcher models use a random 60/40 split across all seasons (2021–2025). The swing models train on 2023–2024 (bat tracking started mid-2023) and validate on 2025. The swing models train on 2023–2024 (bat tracking started mid-2023) and validate on 2025. This matters because any model will look good on its own training data — the real test is how well it generalizes to new players and new seasons.

Year-over-year stability — the predictiveness test

The gold standard question: does a high score in year N predict a high score in year N+1? If it does, the metric is measuring a real, repeatable skill. If scores bounce around randomly year to year, the metric is just noise. Decision has a YoY Pearson r of 0.924–0.952 across four consecutive season pairs — confirming that swing decision quality is a stable, repeatable skill. Stuff xRV/100 (shrunk pitcher-season estimates) stabilizes at r = 0.802–0.836 across pitch groups, meaningful predictive validity for a physics-based metric.

Decision YoY Stability by Season Pair

Season Pair	Pearson r	n (hitters)	Interpretation
2021 → 2022	r = 0.924	394	Very high repeatability across all pairs
2022 → 2023	r = 0.952	397	Peak stability — swing decision is a true skill
2023 → 2024	r = 0.935	407	Consistent year to year
2024 → 2025	r = 0.945	401	OOS (2025 not in training data)

What "predicts" what — cross-metric correlations

Decision (year N) correlates at r = 0.94 with Decision (year N+1) and at r = 0.67 with next-year BB%. It correlates at r = −0.80 with next-year O-Swing% — confirming that Decision predicts future chase rate, not just same-season behavior. xDamage in year N correlates at r = 0.67 with next-year xwOBACON, meaning it meaningfully predicts future quality of contact on balls in play. These relationships confirm the models are capturing real, observable baseball skills.

Limitations and FAQ Where to be careful

Small samples

All scores use reliability shrinkage — a hitter with 250 plate appearances will have their score pulled toward 100 more than a hitter with 600. Even so, treat any score from a player with fewer than ~300 pitch decisions as preliminary.

Bat tracking data limitations

Statcast bat tracking data only became available from mid-2023 onward. xContact and xDamage cannot be calculated for seasons before that. The models also occasionally have missing bat tracking readings for a subset of swings — those pitches are excluded from aggregation rather than imputed.

Stuff ≠ results

A pitch with elite Stuff can still get hit hard due to poor sequencing, predictability, or a well-timed guess. Stuff measures average outcome across random contexts. It isolates the pitch's intrinsic quality. For combined outcomes, check Pitching, which incorporates the location and context.

Decision is not plate discipline

Decision measures run-value-weighted decision quality, not raw chase rate or walk rate. A hitter who swings at borderline strikes is not penalized the same as one who chases clear balls. The model accounts for count leverage — a bad swing decision at 3-0 is worse than the same swing at 0-2.

Human + AI collaboration

A human chose the input feature sets for every model and manually implemented hyperparameter tuning — iterating on model architecture, regularization, and feature engineering without AI guidance. Claude Code handled the rest: writing and debugging all Python model scripts, data processing pipelines, visualization code, and the full site codebase. The pitcher models run on LightGBM with a random 60/40 train/validate split, batter-adjusted run value target, and James-Stein shrinkage on aggregates. Swing and decision models run on XGBoost. All models are trained on Statcast parquet data loaded via PyBaseball. The site uses DuckDB-WASM for in-browser analytics. Questions, feedback, or something looks wrong? Reach out using the contact links in the footer.

On this page

Glossary — Pitchers

Pitcher Stat Definitions

Definitions for every pitcher metric on this site. All source data is owned by Statcast, which provides most of the underlying definitions. Terms marked Model are original machine-learning model outputs. Terms marked Statcast are pulled and calculated directly from MLB. Terms marked Original are designed and developed by this site’s author. Terms marked External are derived from methods or formulae credited to others.

Release Stats * Credit to Alex Chamberlain (FanGraphs) for starting formulae on vertical and horizontal release angles.

Term	Definition
Arm Angle	A horizontal line extending from the location of the pitcher’s throwing shoulder and the location of the ball at the point of release. Lower values indicate more sidearm delivery; higher values indicate more over-the-top. Statcast
Extension	Release distance in feet from the front of the pitching rubber. Greater extension shortens the effective distance to home plate and increases perceived velocity. Statcast
HRA	Horizontal Release Angle — the horizontal angle at which the ball leaves the pitcher’s hand, measured in degrees. External
Rel Height	Height of the pitcher’s release point in feet, measured from the ground. A higher release point generally produces steeper vertical approach angles on fastballs. Statcast
Rel Horiz	Width of the pitcher’s release point in feet from the catcher’s point of view. Negative values indicate the arm side of the plate; positive values indicate the glove side. Statcast
VRA	Vertical Release Angle — the vertical angle at which the ball leaves the pitcher’s hand, measured in degrees. External

Stuff Stats * Credit to Eli Ben-Porat for spin efficiency estimates.

Term	Definition
HAA AA	Horizontal Approach Angle Above Average — the horizontal angle at which a pitch crosses home plate, above expectation based on that pitch’s horizontal location at the plate. External
HB Arm-Side	Horizontal arm-side break of the ball measured in inches. Negative values mean the pitch moved toward the glove side instead. Statcast
HB Arm-Side AA	Horizontal arm-side break above expectation based on arm angle — how much the pitch breaks relative to what a pitcher with the same release slot would typically produce. External
iVB	Induced Vertical Break — vertical movement a pitcher generates by spinning the ball, measured in inches. A four-seam fastball with high iVB appears to “rise” relative to a spinless pitch thrown at the same velocity. Statcast
iVB AA	Induced Vertical Break above expectation based on arm angle — how much carry or drop the pitch generates relative to what a pitcher with the same release slot would typically produce. External
MPH	Pitch velocity in miles per hour, measured at release. Statcast
pMPH	Perceived MPH — adjusted velocity that accounts for a pitcher’s extension. Because greater extension shortens the effective distance to home plate, pMPH better approximates the batter’s actual reaction-time window. Statcast
Spin Axis	Direction of the ball’s spin, displayed as a clock position (e.g., 12:00 = pure backspin, 6:00 = pure topspin). Raw Statcast measurement is in degrees (0–360°). Statcast
Approx. Spin Eff%	Approximate spin efficiency — the percentage of a pitch’s total spin that directly contributes to its movement. Estimated from aerodynamic lift forces derived from Statcast trajectory data. 100% indicates all spin is transverse (pure backspin or topspin), while 0% indicates pure gyrospin with no Magnus effect. External
Spin Rate	Revolutions per minute (RPM) of the ball upon release. Higher spin generally amplifies movement, though the direction of that movement depends on spin axis. Statcast
SSW	Seam-Shifted Wake — differential between the spin axis and the approximate break axis. A high SSW indicates the ball’s seam orientation creates aerodynamic drag beyond what spin alone would predict, causing movement in unexpected directions. External
Stuff	Predicted count-neutral expected run value of a pitch based on its physical characteristics — arm angle, flight time, acceleration vectors, spin rate, spin efficiency, VAA, and altitude (fastball); plus flight-time, ax, and az differentials from the pitcher's primary fastball (breaking/offspeed). Presented on the 20-80 scouting scale: 50 = average, ±10 per SD, floor 20, ceiling 80. Does not account for location, sequencing, or arsenal context. For methodology, see Models → About. Model
VAA AA	Vertical Approach Angle — the flatness or steepness at which a pitch enters the hitting zone — above expectation based on pitch height at the plate. Positive values indicate a flatter approach than typical for that pitch height, which generally correlates with more swing-and-miss. External

Command Stats * Credit to Lance Brozdowski (Marquee Sports) for starting formulae on CL Zone. Credit to Michael Rosen (FanGraphs) for mechanics repeatability and command theory.

Term	Definition
Chase	Rate of pitches in the chase zone — the band just outside the strike zone. Defined as pitches crossing the plate between ±13.3 to ±20 inches horizontally and 6–14 or 46–54 inches vertically. Statcast
CL Zone	Count Leverage Zone Rate — the rate of pitches thrown in the strike zone in count-leverage counts (0-0 and 1-1), the two counts where neither the pitcher nor hitter holds a clear advantage. Used as a measure of a pitcher’s ability to command the strike zone before any count leverage has accumulated. External
Heart	Rate of pitches in the heart of the strike zone — the most hittable area of the plate. Defined as pitches crossing within ±6.7 inches horizontally and 22–38 inches vertically. Statcast
Location	Predicted count-neutral expected run value of a pitch based on its vertical and horizontal location at the plate. Presented on the 20-80 scouting scale: 50 = average, ±10 per SD, floor 20, ceiling 80. For methodology, see Models → About. Model
SD HRA	Standard deviation in horizontal release angles across all pitches of a given type. Lower values indicate more consistent mechanics, based on the idea that repeatability of mechanics leads to better command. External
SD Rel Height	Standard deviation in release height across all pitches of a given type. Lower values indicate more consistent mechanics, based on the idea that repeatability of mechanics leads to better command. External
SD Rel Horiz	Standard deviation in horizontal release width across all pitches of a given type. Lower values indicate more consistent mechanics, based on the idea that repeatability of mechanics leads to better command. External
SD VRA	Standard deviation in vertical release angles across all pitches of a given type. Lower values indicate more consistent mechanics, based on the idea that repeatability of mechanics leads to better command. External
Shadow	Rate of pitches in the shadow zone — the border region around the strike zone. Defined as pitches crossing ±6.7 to ±13.3 inches horizontally at strike-zone heights, or at the top and bottom edges (14–22 or 38–46 inches vertically). The most contested call region and most valuable put-away locations. Statcast
Strike Zone	The rulebook strike zone as tracked by Statcast — pitches crossing the plate within approximately ±8.5 inches horizontally and between the hitter’s knees and the midpoint of their torso vertically, mapped dynamically per batter. Zone rate on this site uses Statcast’s Gameday zone designations (zones 1–9). Statcast
Waste	Rate of pitches in the waste zone — far outside the strike zone. Defined as pitches crossing beyond ±20 inches horizontally or below 6 inches or above 54 inches vertically. Statcast

Arsenal and General Pitching Stats

Term	Definition
BB Group	Breaking ball group usage — combined usage rate of a pitcher’s sliders, sweepers, curveballs, knuckle-curves, slurves, slow curves, and knuckleballs, as a percentage of total pitches. Original
Diversity Mov	Standard deviation of pythagorean break across the full arsenal. Presented on the 20-80 scouting scale: 50 = average, ±10 per SD, floor 20, ceiling 80 — indicating a more varied movement profile across pitch types. Original
Diversity Velo	Standard deviation of perceived MPH across the full arsenal. Presented on the 20-80 scouting scale: 50 = average, ±10 per SD, floor 20, ceiling 80 — indicating a wider velocity spread between pitch types. Original
FB Group	Fastball group usage — combined usage rate of a pitcher’s four-seamers, sinkers, and/or cutters, as a percentage of total pitches. Original
Pitching	Predicted count-neutral expected run value of a pitch based on its stuff and location characteristics combined. Presented on the 20-80 scouting scale: 50 = average, ±10 per SD, floor 20, ceiling 80. For methodology, see Models → About. Model
Primary	Primary pitch usage — usage rate of the pitcher’s most-used pitch type as a percentage of total pitches. Original
Repertoire	Average of Diversity Movement and Diversity Velocity. Presented on the 20-80 scouting scale: 50 = average, ±10 per SD, floor 20, ceiling 80 — indicating a more varied arsenal in both shape and speed. Original
SD HB	Standard deviation of horizontal break across all pitches thrown, regardless of pitch type. Higher values indicate a wider spread of horizontal movement profiles across the arsenal. Original
SD iVB	Standard deviation of induced vertical break across all pitches thrown, regardless of pitch type. Higher values indicate a wider spread of vertical movement across the arsenal. Original
SD pMPH	Standard deviation of perceived MPH across all pitches thrown, regardless of pitch type. Higher values indicate a wider velocity spread across the arsenal. Original
SD Pythag Break	Standard deviation of pythagorean break — calculated as √(HB² + iVB²) — across all pitches. Captures total movement variety across the arsenal by combining horizontal and vertical break into a single distance metric. Original
Tunneling	Standard deviation of the horizontal and vertical position of the ball 150 milliseconds into its flight time, measured in feet. Lower raw values indicate pitches that look more similar early in flight. Presented on the 20-80 scouting scale: 50 = average, ±10 per SD, floor 20, ceiling 80. Original
Usage	Rate at which a specific pitch type is thrown — that pitch type’s count divided by the pitcher’s total pitch count in the filtered sample. Statcast

On this page

Glossary — Hitters

Hitter Stat Definitions

Definitions for every metric on this site. All source data is owned by Statcast, which provides most of the underlying definitions. Terms marked Model are original machine-learning model outputs. Terms marked Statcast are pulled and calculated directly from MLB. Terms marked Original are designed and developed by this site’s author. Terms marked External are derived from methods or formulae credited to others.

Hitter Models

Term	Definition
xContact	Predicted probability of making contact on a given swing, based on swing mechanics — bat speed, swing length, attack angle, attack direction, and swing path tilt — as well as pitch location and approach angles. Presented on the 20-80 scouting scale: 50 = average, ±10 per SD, floor 20, ceiling 80. Requires bat-tracking data available from mid-2023 onward. For methodology, see Models → About. Model
xDamage	Predicted probability of producing solid contact or a barrel on a given batted ball, based on the same swing mechanics and pitch context. Statcast’s barrel and solid contact designations are one reasonable definition of quality contact — not the only one. Presented on the 20-80 scouting scale: 50 = average, ±10 per SD, floor 20, ceiling 80. Requires bat-tracking data available from mid-2023 onward. For methodology, see Models → About. Model
Decision	Predicted run-value quality of each swing-or-take decision. For every pitch, the model estimates the expected run value of swinging and of taking, then computes the difference. Positive values mean the batter chose the higher-value option. Aggregated over a season with reliability shrinkage. Presented on the 20-80 scouting scale: 50 = average, ±10 per SD, floor 20, ceiling 80. For methodology, see Models → About. Model

Decision Strategy

Term	Definition
2K Chase%	Rate of pitches out of the strike zone in two-strike counts (0-2, 1-2, 2-2, 3-2) that the hitter swings at. Original
Early BIP%	Early count ball-in-play rate — rate of pitches in early counts (0-0, 1-0, 0-1) that are put in play. Original
Early Sw%	Early count swing rate — rate of pitches in early counts (0-0, 1-0, 0-1) that are swung at. Original
Green Light%	Swing rate in 3-0 and 3-1 counts. Original
Hitter Ahead%	Rate of pitches seen with the hitter ahead in the count — counts 1-0, 2-0, 2-1, 3-0, and 3-1. Original
Pitcher Ahead%	Rate of pitches seen with the pitcher ahead in the count — counts 0-1, 0-2, 1-2, and 2-2. Original
Pitcher Count Lev%	Pitcher count leverage — Pitcher Ahead% minus Hitter Ahead%. Positive values indicate the hitter saw a greater share of pitches in pitcher-friendly counts. Original

Swing Mechanics* Credit to Kyle Bland for swing acceleration and swing time formulae. Credit to Lau Sze Yui for vertical and horizontal offset formulae.

Term	Definition
Attack Angle	Vertical angle of the direction the sweet spot of the bat is travelling at the contact point, relative to the ground. Statcast
Attack Direction	Horizontal angle of the direction the sweet spot of the bat is travelling at the contact point, relative to an imaginary line from home plate to straightaway center field. Statcast
Bat Speed	Speed of the swing in miles per hour, measured at the sweet spot of the bat. Statcast
Bat Spd 90th	90th-percentile bat speed for the hitter in the filtered sample. Statcast
Contact Pt X	Horizontal distance from the hitter’s center of mass that he makes contact with the ball, or passes closest to it on a swing and miss, measured in inches. Statcast
Contact Pt Y	Distance in front of the hitter’s center of mass that he makes contact with the ball, or passes closest to it on a swing and miss, measured in inches. Statcast
Adj. Smash Factor	Driveline’s collision efficiency formula — measuring how much of a swing’s speed was converted to exit velocity. Presented on the 20-80 scouting scale: 50 = average, ±10 per SD, floor 20, ceiling 80. Original
Swing Acceleration	Approximate acceleration of the swing in ft/s². External
Swing Length	Length of the swing path in feet from hand start position to contact. Statcast
Swing Path Tilt	The angular orientation of the swing path compared to the ground, measured 40 milliseconds before contact. Statcast
Swing Time	Approximate duration of the swing in seconds. External
Vert Offset	The space between the center of the ball and the center of the bat at contact, in inches. A value of 0 indicates perfect vertical contact. External
Horiz Offset	The space between the ball and the horizontal center of the bat at contact, in inches. Lower values indicate contact closer to the end of the bat; higher values indicate contact closer to the handle. External

Swing Strategy

Term	Definition
BSpdΔ Heart	A hitter’s average bat speed on pitches in the heart zone (±6.7 in. horizontally, 22–38 in. vertically) minus his overall average bat speed. Original
BSpdΔ Hit Count	A hitter’s average bat speed on swings in hitter-ahead counts (1-0, 2-0, 3-0, 2-1, 3-1) minus his overall average bat speed. Original
BSpdΔ 2 Strikes	A hitter’s average bat speed on swings in two-strike counts minus his overall average bat speed. Original
SwLenΔ 2 Strikes	A hitter’s average swing length on swings in two-strike counts minus his overall average swing length. Original
TiltΔ High	A hitter’s average swing path tilt on pitches above the vertical midpoint of the strike zone (plate_z > 2.5 ft) minus his overall average swing tilt. Original
TiltΔ Low	A hitter’s average swing path tilt on pitches below the vertical midpoint of the strike zone (plate_z < 2.5 ft) minus his overall average swing tilt. Original
Vert Plate Cov	Vertical plate coverage — TiltΔ Low minus TiltΔ High. Measures the total range of swing-tilt adjustment between high and low pitches. Original

Pitch Shapes

Pitch shape categories used in the Hitters vs Pitch Shapes tool. Each group is defined by pitch type, velocity, spin rate, and movement characteristics (iVB = induced vertical break in inches; arm-side HB = horizontal break toward the pitcher’s arm side in inches). Stats shown are Contact%, Chase%, xBACON, or xSLGCON against each shape.

Shape	Definition
85+ Breaking	Breaking balls thrown 85 MPH or faster. Consists mostly of gyro sliders and knuckle curves. Statcast
95+	Any pitch thrown 95 MPH or faster, regardless of pitch type or movement profile. Statcast
Drop	Breaking balls with at least 6 inches of downward break. Predominantly includes curveballs, especially knuckle curves. Statcast
Gyro	Breaking balls with both iVB and arm-side HB between −3 and +3 inches. Pitches with minimal true break in any direction; includes most traditional sliders with gyro/bullet spin. Statcast
No-Spin	Any pitch thrown with 1,500 RPM or fewer. Comprised predominantly of changeups, splitters, and knuckleballs. Produces unpredictable or minimal break. Statcast
Ride-Cut	Fastballs with 16 or more inches of iVB and 7 or fewer inches of arm-side HB. High-carry with tight horizontal movement. Statcast
Ride-Run	Fastballs with 16 or more inches of iVB and 7 or more inches of arm-side HB. High-carry fastballs that also run back toward the arm side. Statcast
Running OS	Offspeed pitches with 12 or more inches of arm-side HB. Mostly comprised of changeups with significant tail. Statcast
Sink-Cut	Fastballs with 16 or fewer inches of iVB and 7 or fewer inches of arm-side HB. Steeper fastballs with tight horizontal movement; includes most cutters. Statcast
Sink-Run	Fastballs with 16 or fewer inches of iVB and 7 or more inches of arm-side HB. Pitches that drop and tail; includes most sinkers and running fastballs. Statcast
Sinking OS	Offspeed pitches with 3 or fewer inches of iVB. Low-spin offerings with drop; includes most splitters, forkballs, and some kick changeups. Statcast
Spin	Any pitch thrown with 2,500 RPM or more, regardless of pitch type or velocity. Often associated with cutting action. Statcast
Sweep	Breaking balls with at least 6 inches of glove-side break. Comprised mostly of sweepers and traditional curveballs. Statcast
Two-Plane	Breaking balls with both −6 or fewer inches of iVB and −6 or fewer inches of arm-side HB simultaneously. Includes most traditional curveballs. Statcast

On this page

Glossary — General

General Stat Definitions

Definitions for shared metrics across pitcher and hitter leaderboards. All source data is owned by Statcast. Terms marked Model are original machine-learning model outputs. Terms marked Statcast are pulled and calculated directly from MLB. Terms marked Original are designed and developed by this site’s author. Terms marked External are derived from methods or formulae credited to others.

Quick Stats

Term	Definition
AVG	Batting average — hits divided by at-bats. Statcast
BB%	Walk rate — percentage of plate appearances ending in a walk. Statcast
HBP%	Hit-by-pitch rate — percentage of plate appearances ending in a hit-by-pitch. Statcast
HR	Home runs. Statcast
K%	Strikeout rate — percentage of plate appearances ending in a strikeout. Statcast
K-BB%	Strikeout rate minus walk rate. Statcast
OBP	On-base percentage — rate of reaching base via hit, walk, or hit-by-pitch. Statcast
PA	Plate appearances. Statcast
SLG	Slugging percentage — total bases divided by at-bats. Statcast
wOBA	Weighted On-Base Average — an offensive statistic that assigns different weights to singles, doubles, triples, home runs, walks, hit-by-pitches, and strikeouts based on the observed run values of those events. Scaled to resemble OBP. Statcast
xBACON	Expected batting average on contact — estimated batting average based on exit velocity and launch angle on balls in play. Statcast
xISOCON	xSLGCON minus xBACON — expected isolated power on contact. Statcast
xSLGCON	Expected slugging percentage on contact — estimated slugging based on exit velocity and launch angle on balls in play. Statcast
xwOBACON	Expected wOBA on contact — estimated wOBA based on exit velocity and launch angle on balls in play. Statcast

Value Stats

Term	Definition
Run Value	Average change in pre-pitch run expectancy caused by a pitch, based on count, baserunners, and outs. Negative values benefit the pitcher; positive values benefit the batter. Statcast
RV/100	Run value per 100 pitches seen. Statcast
wPA	Win Probability Added — total change in win probability generated by all plate appearances, per MLB’s win probability model. Statcast
wPA/100	Win Probability Added per 100 pitches seen. Statcast

Contact Ability Stats

Term	Definition
BIP%	Ball-in-play rate — percentage of plate appearances ending with a ball put in play. Statcast
Contact%	Percentage of total swings that result in contact — foul balls, fair balls, or foul tips. Statcast
CSW%	Called strikes plus swinging strikes as a percentage of total pitches. Statcast
CStr%	Called strike rate — percentage of all pitches resulting in a called strike. Statcast
Heart SwStr%	Swinging strike rate on pitches in the heart zone (±6.7 in. horizontally, 22–38 in. vertically). Statcast
O Contact%	Percentage of swings at pitches outside the strike zone that result in contact. Statcast
SwStr%	Swinging strike rate — percentage of all pitches resulting in a swinging strike. Statcast
Z Contact%	Percentage of swings at pitches in the strike zone that result in contact. Statcast

Contact Distribution Stats

Term	Definition
Adj. GB+PU%	Adjusted ground ball plus pop-up rate — percentage of batted balls with a launch angle below −3° or above 38°. These outcomes averaged a wOBA of 0.300 or less from 2022–2025. Original
BBE	Batted ball events — total balls put in play. Statcast
Cent%	Approximate percentage of batted balls hit up the middle, defined as a spray angle between −15° and +15°. Statcast
FB%	Fly ball rate — percentage of batted balls classified by Statcast as fly balls. Statcast
GB%	Ground ball rate — percentage of batted balls classified by Statcast as ground balls. Statcast
LA StdDev	Standard deviation of launch angle on all batted balls. Lower values tend to correlate with higher batting averages. Statcast
LD%	Line drive rate — percentage of batted balls classified by Statcast as line drives. Statcast
Oppo%	Approximate percentage of batted balls hit opposite field, defined as a spray angle above +15°. Statcast
Pull%	Approximate percentage of batted balls pulled, defined as a spray angle below −15°. Statcast
Pull FB%	Percentage of all batted balls that were pulled fly balls — spray angle below −15° on a fly ball. Statcast
PU%	Pop-up rate — percentage of batted balls classified by Statcast as pop-ups. Statcast
Spray Angle	Approximate average batted ball direction relative to a line from home plate to straightaway center field. Negative values indicate pull tendency; positive values indicate opposite-field tendency. Statcast
wOBACON	wOBA on contact — calculated only on balls put in play. Statcast

Contact Quality Stats * Credit to Jonathan Metzelaar (Pitcher List) for the Ideal Contact formula.

Term	Definition
Barrel%	Percentage of batted balls resulting in a Statcast barrel — a dynamic combination of exit velocity and launch angle considered very likely to produce an extra-base hit. Statcast
DHH%	Dynamic Hard-Hit Rate — percentage of batted balls at or above the 75th-percentile exit velocity for that batted ball’s combination of bat speed, launch angle, and spray angle bucket (pull, center, or opposite field). Original
EV90	90th-percentile exit velocity in miles per hour. Statcast
Flare/Burn%	Percentage of batted balls resulting in a Statcast flare or burner — hard ground balls and medium liners very likely to produce a base hit. Statcast
HH LA	Hard-hit launch angle — average launch angle on batted balls with an exit velocity of 95 MPH or greater. Statcast
Ideal Con%	Ideal contact rate — barrels plus solid contact plus flares and burners as a percentage of batted balls. External
Max EV	Maximum exit velocity in miles per hour in the sample. Statcast
Solid%	Percentage of batted balls resulting in Statcast solid contact — high-end exit velocity given the launch angle. Statcast
Topped%	Percentage of batted balls resulting in Statcast topped contact — weakly hit on the ground. Statcast
Under%	Percentage of batted balls resulting in Statcast under contact — weakly hit in the air. Statcast
Weak%	Percentage of batted balls with an exit velocity of 60 MPH or less. Statcast

Plate Discipline Stats

Term	Definition
Ball%	Percentage of pitches resulting in a ball. Statcast
Chase Sw%	Percentage of pitches in the chase zone (±13.3–20 in. horizontally, 6–14 or 46–54 in. vertically) that were swung at. Statcast
CSW-Ball%	CSW% minus Ball% — favorable pitch outcomes minus unfavorable ones. Statcast
Heart Sw%	Percentage of pitches in the heart zone (±6.7 in. horizontally, 22–38 in. vertically) that were swung at. Statcast
O Swing%	Percentage of pitches outside the strike zone that were swung at. Statcast
Shadow Sw%	Percentage of pitches in the shadow zone (±6.7–13.3 in. at strike-zone heights, or 14–22 and 38–46 in. vertically) that were swung at. Statcast
Waste Sw%	Percentage of pitches in the waste zone (beyond ±20 in. horizontally or below 6 in. or above 54 in. vertically) that were swung at. Statcast
Z Swing%	Percentage of pitches in the strike zone that were swung at. Statcast
Zone%	Percentage of pitches landing within the strike zone (Gameday zones 1–9). Statcast

A–Z Index

Glossary — Full Alphabetical

All Stats, A to Z

Every metric on this site in alphabetical order. Use the letter index to jump to a section. For full context and methodology, see the Pitchers, Hitters, and General tabs.

Term	Definition
Adj. GB+PU%	Adjusted ground ball plus pop-up rate — percentage of batted balls with a launch angle below −3° or above 38°. Original
Arm Angle	A horizontal line extending from the pitcher’s throwing shoulder to the ball at release. Lower values indicate more sidearm; higher values indicate over-the-top. Statcast
Attack Angle	Vertical angle of the bat’s sweet spot direction at contact, relative to the ground. Statcast
Attack Direction	Horizontal angle of the bat’s sweet spot direction at contact, relative to a line from home plate to center field. Statcast
AVG	Batting average — hits divided by at-bats. Statcast

Term	Definition
Ball%	Percentage of pitches resulting in a ball. Statcast
Barrel%	Percentage of batted balls resulting in a Statcast barrel — a combination of exit velocity and launch angle very likely to produce an extra-base hit. Statcast
Bat Spd 90th	90th-percentile bat speed. Statcast
Bat Speed	Speed of the swing in miles per hour, measured at the sweet spot. Statcast
BB Group	Breaking ball group usage — combined usage of sliders, sweepers, curveballs, knuckle-curves, slurves, slow curves, and knuckleballs. Original
BB%	Walk rate — percentage of plate appearances ending in a walk. Statcast
BBE	Batted ball events — total balls put in play. Statcast
BIP%	Ball-in-play rate — percentage of plate appearances ending with a ball in play. Statcast
BSpdΔ 2 Strikes	Bat speed in two-strike counts minus overall bat speed. Original
BSpdΔ Heart	Bat speed on heart-zone pitches minus overall bat speed. Original
BSpdΔ Hit Count	Bat speed in hitter-ahead counts minus overall bat speed. Original

Term	Definition
Cent%	Approximate percentage of batted balls hit up the middle (spray angle −15° to +15°). Statcast
Chase	Rate of pitcher pitches in the chase zone (±13.3–20 in. horizontally, 6–14 or 46–54 in. vertically). Statcast
Chase Sw%	Percentage of pitches in the chase zone that were swung at. Statcast
CL Zone	Count Leverage Zone Rate — rate of pitches in the strike zone in 0-0 and 1-1 counts. External
Contact Pt X	Horizontal distance from the hitter’s center of mass to the contact point, in inches. Statcast
Contact Pt Y	Distance in front of the hitter’s center of mass to the contact point, in inches. Statcast
Contact%	Percentage of total swings that result in contact. Statcast
xContact	Predicted contact probability per swing. Presented on the 20-80 scouting scale: 50 = average, ±10 per SD, floor 20, ceiling 80, higher is better. For methodology, see Models → About. Model
CStr%	Called strike rate — percentage of all pitches resulting in a called strike. Statcast
CSW%	Called strikes plus swinging strikes as a percentage of total pitches. Statcast
CSW-Ball%	CSW% minus Ball%. Statcast

Term	Definition
xDamage	Predicted solid contact or barrel probability per batted ball. Presented on the 20-80 scouting scale: 50 = average, ±10 per SD, floor 20, ceiling 80, higher is better. For methodology, see Models → About. Model
Decision	Predicted run-value quality of swing-or-take decisions, aggregated over a season. Presented on the 20-80 scouting scale: 50 = average, ±10 per SD, floor 20, ceiling 80, higher is better. For methodology, see Models → About. Model
DHH%	Dynamic Hard-Hit Rate — percentage of batted balls at or above the 75th-percentile EV for that bat speed, launch angle, and spray angle bucket. Original
Diversity Mov	Standard deviation of pythagorean break across the full arsenal. Presented on the 20-80 scouting scale: 50 = average, ±10 per SD, floor 20, ceiling 80, higher is better. Original
Diversity Velo	Standard deviation of perceived MPH across the full arsenal. Presented on the 20-80 scouting scale: 50 = average, ±10 per SD, floor 20, ceiling 80, higher is better. Original
Drop	Breaking balls with at least 6 inches of downward break. Predominantly includes curveballs, especially knuckle curves. Original

Term	Definition
Early BIP%	Ball-in-play rate in early counts (0-0, 1-0, 0-1). Original
Early Sw%	Swing rate in early counts (0-0, 1-0, 0-1). Original
EV90	90th-percentile exit velocity in mph. Statcast
Extension	Release distance in feet from the front of the pitching rubber. Statcast

Term	Definition
FB Group	Fastball group usage — combined usage of four-seamers, sinkers, and/or cutters. Original
FB%	Fly ball rate — percentage of batted balls classified as fly balls. Statcast
Flare/Burn%	Percentage of batted balls resulting in a Statcast flare or burner — hard grounders and medium liners very likely to be base hits. Statcast

Term	Definition
GB%	Ground ball rate — percentage of batted balls classified as ground balls. Statcast
Green Light%	Swing rate in 3-0 and 3-1 counts. Original
Gyro	Breaking balls with both iVB and arm-side HB between −3 and +3 inches. Pitches with minimal true break in any direction; includes most traditional sliders with gyro/bullet spin. Original

Term	Definition
HAA AA	Horizontal Approach Angle above expectation based on horizontal plate location. External
HB Arm-Side	Horizontal arm-side break in inches. Negative values indicate glove-side movement. Statcast
HB Arm-Side AA	Horizontal arm-side break above expectation based on arm angle. External
HBP%	Hit-by-pitch rate — percentage of plate appearances ending in a HBP. Statcast
Heart	Rate of pitcher pitches in the heart zone (±6.7 in., 22–38 in. vertically). Statcast
Heart Sw%	Percentage of pitches in the heart zone that were swung at. Statcast
Heart SwStr%	Swinging strike rate on pitches in the heart zone. Statcast
HH LA	Average launch angle on batted balls with an exit velocity of 95 MPH or greater. Statcast
Hitter Ahead%	Rate of pitches seen in hitter-ahead counts (1-0, 2-0, 2-1, 3-0, 3-1). Original
HR	Home runs. Statcast
Horiz Offset	The space between the ball and the horizontal center of the bat at contact, in inches. Lower values indicate contact closer to the end of the bat; higher values indicate contact closer to the handle. External
HRA	Horizontal Release Angle — horizontal angle at which the ball leaves the hand. External

Term	Definition
iVB	Induced Vertical Break — vertical movement generated by spin, in inches. Statcast
iVB AA	Induced Vertical Break above expectation based on arm angle. External

Term	Definition
K%	Strikeout rate — percentage of plate appearances ending in a strikeout. Statcast
K-BB%	Strikeout rate minus walk rate. Statcast

Term	Definition
LA StdDev	Standard deviation of launch angle. Lower values tend to correlate with higher batting averages. Statcast
LD%	Line drive rate — percentage of batted balls classified as line drives. Statcast
Location	Predicted count-neutral run value based on pitch location. Presented on the 20-80 scouting scale: 50 = average, ±10 per SD, floor 20, ceiling 80, higher is better. For methodology, see Models → About. Model

Term	Definition
Max EV	Maximum exit velocity in mph. Statcast
MPH	Pitch velocity in miles per hour, measured at release. Statcast

Term	Definition
No-Spin	Any pitch thrown with 1,500 RPM or fewer. Comprised predominantly of changeups, splitters, and knuckleballs. Produces unpredictable or minimal break. Original

Term	Definition
O Contact%	Percentage of swings outside the strike zone that result in contact. Statcast
O Swing%	Percentage of pitches outside the strike zone that were swung at. Statcast
OBP	On-base percentage — rate of reaching base via hit, walk, or HBP. Statcast
Oppo%	Approximate percentage of batted balls hit opposite field (spray angle > +15°). Statcast

Term	Definition
PA	Plate appearances. Statcast
Pitcher Ahead%	Rate of pitches seen in pitcher-ahead counts (0-1, 0-2, 1-2, 2-2). Original
Pitcher Count Lev%	Pitcher Ahead% minus Hitter Ahead%. Original
Pitching	Predicted count-neutral run value combining stuff and location. Presented on the 20-80 scouting scale: 50 = average, ±10 per SD, floor 20, ceiling 80, higher is better. For methodology, see Models → About. Model
pMPH	Perceived MPH — velocity adjusted for release extension. Statcast
Primary	Primary pitch usage — usage rate of the pitcher’s most-used pitch type. Original
PU%	Pop-up rate — percentage of batted balls classified as pop-ups. Statcast
Pull FB%	Percentage of batted balls that were pulled fly balls. Statcast
Pull%	Approximate percentage of batted balls pulled (spray angle < −15°). Statcast

Term	Definition
Rel Height	Height of the pitcher’s release point in feet. Statcast
Rel Horiz	Width of the pitcher’s release point in feet from the catcher’s point of view. Statcast
Repertoire	Average of Diversity Movement and Diversity Velocity. Presented on the 20-80 scouting scale: 50 = average, ±10 per SD, floor 20, ceiling 80, higher is better. Original
Run Value	Average change in pre-pitch run expectancy based on count, baserunners, and outs. Statcast
RV/100	Run value per 100 pitches. Statcast
Ride-Cut	Fastballs with 16 or more inches of iVB and 7 or fewer inches of arm-side HB. High-carry with tight horizontal movement. Original
Ride-Run	Fastballs with 16 or more inches of iVB and 7 or more inches of arm-side HB. High-carry fastballs that also run back toward the arm side. Original
Running OS	Offspeed pitches with 12 or more inches of arm-side HB. Mostly comprised of changeups with significant tail. Original

Term	Definition
SD HB	Standard deviation of horizontal break across all pitches. Original
SD HRA	Standard deviation in horizontal release angles. Lower values indicate more consistent mechanics, based on the idea that repeatability of mechanics leads to better command. External
SD iVB	Standard deviation of induced vertical break across all pitches. Original
SD pMPH	Standard deviation of perceived MPH across all pitches. Original
SD Pythag Break	Standard deviation of pythagorean break — √(HB² + iVB²) — across all pitches. Original
SD Rel Height	Standard deviation in release height. Lower values indicate more consistent mechanics, based on the idea that repeatability of mechanics leads to better command. External
SD Rel Horiz	Standard deviation in horizontal release width. Lower values indicate more consistent mechanics, based on the idea that repeatability of mechanics leads to better command. External
SD VRA	Standard deviation in vertical release angles. Lower values indicate more consistent mechanics, based on the idea that repeatability of mechanics leads to better command. External
Shadow	Rate of pitcher pitches in the shadow zone (±6.7–13.3 in., or 14–22/38–46 in. vertically). Statcast
Shadow Sw%	Percentage of pitches in the shadow zone that were swung at. Statcast
SLG	Slugging percentage — total bases divided by at-bats. Statcast
Adj. Smash Factor	Collision efficiency — bat speed converted to exit velocity. Presented on the 20-80 scouting scale: 50 = average, ±10 per SD, floor 20, ceiling 80, higher is better. Original
Solid%	Percentage of batted balls resulting in Statcast solid contact — high-end EV given the launch angle. Statcast
Spin Axis	Direction of the ball’s spin as a clock position. Raw measurement in degrees (0–360°). Statcast
Approx. Spin Eff%	Approximate spin efficiency — the percentage of a pitch’s total spin that directly contributes to its movement. External
Spin Rate	Revolutions per minute (RPM) of the ball upon release. Statcast
Spray Angle	Average batted ball direction relative to straightaway center field. Negative = pull, positive = oppo. Statcast
SSW	Seam-Shifted Wake — differential between spin axis and approximate break axis. External
Strike Zone	The rulebook strike zone as tracked by Statcast (Gameday zones 1–9). Statcast
Stuff	Predicted count-neutral run value based on pitch physics only — not location or arsenal context. Presented on the 20-80 scouting scale: 50 = average, ±10 per SD, floor 20, ceiling 80, higher is better. For methodology, see Models → About. Model
Swing Acceleration	Approximate swing acceleration in ft/s². External
Swing Length	Length of the swing path in feet. Statcast
Swing Path Tilt	Angular orientation of the swing relative to the ground, measured 40ms before contact. Statcast
Swing Time	Approximate swing duration in seconds. External
SwLenΔ 2 Strikes	Swing length in two-strike counts minus overall swing length. Original
SwStr%	Swinging strike rate — percentage of all pitches resulting in a swinging strike. Statcast
Sink-Cut	Fastballs with 16 or fewer inches of iVB and 7 or fewer inches of arm-side HB. Steeper fastballs with tight horizontal movement; includes most cutters. Original
Sink-Run	Fastballs with 16 or fewer inches of iVB and 7 or more inches of arm-side HB. Pitches that drop and tail; includes most sinkers and running fastballs. Original
Sinking OS	Offspeed pitches with 3 or fewer inches of iVB. Low-spin offerings with drop; includes most splitters, forkballs, and some kick changeups. Original
Spin	Any pitch thrown with 2,500 RPM or more, regardless of pitch type or velocity. Often associated with cutting action. Original
Sweep	Breaking balls with at least 6 inches of glove-side break. Comprised mostly of sweepers and traditional curveballs. Original

Term	Definition
TiltΔ High	Swing tilt on pitches above plate_z 2.5 ft minus overall swing tilt. Original
TiltΔ Low	Swing tilt on pitches below plate_z 2.5 ft minus overall swing tilt. Original
Topped%	Percentage of batted balls resulting in Statcast topped contact — weakly hit on the ground. Statcast
Tunneling	Standard deviation of pitch position 150ms into flight, in feet. Presented on the 20-80 scouting scale: 50 = average, ±10 per SD, floor 20, ceiling 80, higher is better. Original
Two-Plane	Breaking balls with both −6 or fewer inches of iVB and −6 or fewer inches of arm-side HB simultaneously. Includes most traditional curveballs. Original

Term	Definition
Under%	Percentage of batted balls resulting in Statcast under contact — weakly hit in the air. Statcast
Usage	Rate at which a specific pitch type is thrown as a percentage of total pitches. Statcast

Term	Definition
VAA AA	Vertical Approach Angle above expectation based on pitch height at the plate. External
Vert Offset	The space between the center of the ball and the center of the bat at contact, in inches. A value of 0 indicates perfect vertical contact. External
Vert Plate Cov	Vertical plate coverage — TiltΔ Low minus TiltΔ High. Original
VRA	Vertical Release Angle — vertical angle at which the ball leaves the hand, in degrees. External

Term	Definition
Waste	Rate of pitcher pitches in the waste zone (beyond ±20 in. horizontally or below 6/above 54 in. vertically). Statcast
Waste Sw%	Percentage of pitches in the waste zone that were swung at. Statcast
Weak%	Percentage of batted balls with an exit velocity of 60 MPH or less. Statcast
wOBA	Weighted On-Base Average — weights singles, doubles, triples, HRs, walks, HBP, and strikeouts by observed run values. Statcast
wOBACON	wOBA on contact — calculated only on balls put in play. Statcast
wPA	Win Probability Added per season, per MLB’s win probability model. Statcast
wPA/100	Win Probability Added per 100 pitches seen. Statcast

Term	Definition
xBACON	Expected batting average on contact based on exit velocity and launch angle. Statcast
xISOCON	xSLGCON minus xBACON. Statcast
xSLGCON	Expected slugging percentage on contact based on EV and LA. Statcast
xwOBACON	Expected wOBA on contact based on EV and LA. Statcast

Term	Definition
Z Contact%	Percentage of swings at pitches in the strike zone that result in contact. Statcast
Z Swing%	Percentage of pitches in the strike zone that were swung at. Statcast
Zone%	Percentage of pitches landing within the strike zone. Statcast

Term	Definition
2K Chase%	Rate of pitches out of the strike zone in two-strike counts (0-2, 1-2, 2-2, 3-2) that the hitter swings at. Original
85+ Breaking	Breaking balls thrown 85 MPH or faster. Consists mostly of gyro sliders and knuckle curves. Original
95+	Any pitch thrown 95 MPH or faster, regardless of pitch type or movement profile. Original

Hitter Plate Dashboards

—

Hitter Damage Dashboards

—

Pitcher Dashboards

—

Pitch Arsenal Dashboards

—

BASEBALL SCOUTING LAB

How the Models Work

Pitcher Stat Definitions

Hitter Stat Definitions

General Stat Definitions

All Stats, A to Z