Baseball Analytics Glossary

Your comprehensive guide to understanding baseball statistics, advanced metrics, and betting terminology

🔍

Showing 98 terms

A

American Odds

Sports Betting

American Odds Format

Odds shown as +/- values vs. $100 baseline.

Example: +200 returns $200 profit on $100 stake.

API

Tech

Application Programming Interface

Interface for programmatically requesting data or actions.

Example: statsapi.mlb.com endpoints for schedules and stats.

ARM

Fielding

Outfield Arm Runs

Runs gained via throwing out runners or deterring advances.

Example: Strong arms increase ARM and reduce extra bases.

AVG

Batting

Batting Average

Share of at-bats resulting in a hit.

Formula: H / AB
Example: A .300 AVG indicates 3 hits per 10 at-bats.

B

BABIP

Batting

Batting Average on Balls In Play

Measures how often a ball in play goes for a hit.

Formula: (H - HR) / (AB - K - HR + SF)
Example: A player with a .400 BABIP is likely experiencing good luck and may regress.

Bankroll

Sports Betting

Bankroll Management

Sizing bets to reduce risk of ruin.

Example: Flat 1–2% unit sizing is common.

Barrel%

Batting

Barrel Rate

Share of batted balls with ideal exit velocity/launch angle for damage.

Example: 10% Barrel% is very strong.

Baseline

Machine Learning

Baseline Model

Simple benchmark to beat before complex models.

Example: Predict league-average runs adjusted for park.

BB/9

Pitching

Walks per Nine

Walks scaled to nine innings.

Formula: BB * 9 / IP
Example: Under 2.0 BB/9 indicates strong command.

BB%

Batting

Walk Rate

Percentage of plate appearances ending in a walk.

Formula: BB / PA
Example: A 12% BB% indicates strong plate discipline.

BsR

Baserunning

Base Running Runs

Total baserunning value (steals, advancement, outs).

Example: +3 BsR means ~3 runs added on the bases.

C

CLV

Sports Betting

Closing Line Value

Difference between your bet’s odds and the market’s closing odds.

Example: Betting Over 8.0 that closes 8.5 shows positive CLV.

CSW%

Pitching

Called Strikes + Whiffs Percentage

Share of pitches resulting in called strikes or swinging strikes.

Example: 30%+ CSW% is strong.

CV

Machine Learning

Cross-Validation

Resampling method to estimate out-of-sample performance.

Example: K-fold CV with time-aware splits for season data.

D

Decimal Odds

Sports Betting

Decimal Odds Format

Total return per $1 stake.

Example: 1.80 returns $1.80 total on $1 stake.

DRS

Fielding

Defensive Runs Saved

Runs saved above/below average on defense.

Example: +10 DRS indicates strong defensive value.

E

Early Stopping

Machine Learning

Early Stopping

Stop training when validation metric stops improving.

Example: Patience of 100 rounds on eval_set.

ERA

Pitching

Earned Run Average

Average number of earned runs a pitcher allows per nine innings.

Formula: ER * 9 / IP
Example: An ERA under 3.00 is considered excellent.

ETL

Data Engineering

Extract, Transform, Load

Process of ingesting and preparing data for analytics.

Example: Pull from stats APIs, clean, and load to PostgreSQL.

EV

Batting

Average Exit Velocity

Average speed (mph) of a player's batted balls.

Example: 92 mph EV implies above-average pop.

EV

Sports Betting

Expected Value

Average profit/loss of a bet over many trials.

Example: A +EV edge exists when true probability exceeds implied probability.

F

FB%

Batting

Fly Ball Percentage

Share of batted balls hit in the air (fly balls).

Example: Higher FB% can boost HR potential.

Feature Engineering

Machine Learning

Feature Engineering

Creating informative predictors from raw data.

Example: Rolling 14-day wOBA or park-adjusted metrics.

Feature Importance

Machine Learning

Model Feature Importance

Relative contribution of features to model predictions.

Example: Gain-based importances highlight impactful stats.

FIP

Pitching

Fielding Independent Pitching

Estimates pitcher performance using HR, BB, HBP, and K, independent of defense.

Example: FIP can better predict future ERA than past ERA alone.

First-Pitch Strike%

Pitching

First-Pitch Strike Percentage

How often a pitcher gets ahead 0–1.

Example: Higher first-pitch strike% often correlates with lower walks.

Fractional Odds

Sports Betting

Fractional Odds Format

Profit : Stake ratio.

Example: 5/2 returns $5 profit per $2 stake.

FRM

Fielding

Catcher Framing Runs

Runs a catcher adds via receiving and framing borderline pitches.

Example: Elite framers can add multiple runs over a season.

Futures

Sports Betting

Futures Market

Season-long outcomes (e.g., World Series winner).

Example: Team to win division at +350.

G

GB%

Batting

Ground Ball Percentage

Share of batted balls hit on the ground.

Example: High GB% can suppress home runs.

Grid Search

Machine Learning

Grid Search Hyperparameter Tuning

Exhaustive search over specified parameter grids.

Example: Try depths {4,6,8} × learning rates {0.03,0.1}.

H

HardHit%

Batting

Hard-Hit Rate

Percentage of batted balls hit 95+ mph.

Example: 45% HardHit% suggests consistent quality contact.

Hedge

Sports Betting

Hedging

Taking an opposing bet to reduce variance or lock profits.

Example: Hedge a longshot future after team reaches the finals.

HR/9

Pitching

Home Runs per Nine

Home runs allowed scaled to nine innings.

Formula: HR * 9 / IP
Example: Keeping HR/9 low is critical in hitter-friendly parks.

HR/FB

Batting

Home Run per Fly Ball

Percentage of fly balls that become home runs.

Example: Unusually high HR/FB can regress.

I

Implied Prob

Sports Betting

Implied Probability

Probability implied by the odds.

Formula: For -A: A/(A+100); for +A: 100/(A+100)
Example: -150 ⇒ 60.0% implied; +150 ⇒ 40.0% implied.

ISO

Batting

Isolated Power

Power-only measure, stripping out singles.

Formula: SLG - AVG
Example: .200 ISO indicates above-average power.

K

K-BB%

Pitching

Strikeout Minus Walk Percentage

Plate-discipline dominance metric.

Formula: K% - BB%
Example: 15%+ is solid; 20%+ is excellent.

K/9

Pitching

Strikeouts per Nine

Strikeouts scaled to nine innings.

Formula: SO * 9 / IP
Example: 10.0 K/9 shows swing-and-miss stuff.

K%

Batting

Strikeout Rate

Percentage of plate appearances ending in strikeout.

Formula: SO / PA
Example: A 30% K% is high swing-and-miss.

Kelly

Sports Betting

Kelly Criterion

Bankroll staking formula maximizing long-term growth.

Example: Bet fraction = (bp - q)/b for decimal odds b with win prob p.

L

LA

Batting

Average Launch Angle

Average vertical angle of batted balls.

Example: 12–15° LA often yields more extra-base hits.

LD%

Batting

Line Drive Percentage

Share of batted balls classified as line drives.

Example: Line drives have the highest hit probability.

Leakage

Machine Learning

Target Leakage

Using future or label-derived info in training features.

Example: Including final score in a pregame model is leakage.

Learning Rate

Machine Learning

Boosting Learning Rate

Step size for boosting updates; lower often needs more trees.

Example: eta = 0.05 with 2000 estimators.

LI

Team

Leverage Index

Measures pressure of a game situation (1.0 = average).

Example: Closers often pitch in LI > 1.5 spots.

Live Betting

Sports Betting

In-Game Betting

Wagering while the game is in progress.

Example: Take the over after a low-scoring first inning at a better number.

LOB%

Pitching

Left On Base Percentage

Share of base runners a pitcher strands.

Example: Extreme LOB% often regresses toward league norms.

M

MAE

Machine Learning

Mean Absolute Error

Average absolute difference between predictions and targets.

Example: MAE of 0.65 runs on total-runs prediction.

Max Depth

Machine Learning

Tree Maximum Depth

Controls interaction complexity and overfitting.

Example: Depth 6–8 is a common starting range.

Middle

Sports Betting

Middling

Holding positions on both sides at different numbers to win both if final lands in between.

Example: Under 9.5 early and Over 8.0 late creates a middle at 9.

Moneyline

Sports Betting

Moneyline

Bet on which team wins the game straight up.

Example: -150 favorite implies ~60% implied probability.

MSE

Machine Learning

Mean Squared Error

Average squared error; penalizes large misses.

Example: Used for hyperparameter tuning alongside MAE.

N

n_estimators

Machine Learning

Number of Trees

Total boosting iterations or forest trees.

Example: 500 trees with early stopping on validation loss.

O

O-Swing%

Pitching

Chase Rate

Percentage of swings at pitches outside the zone.

Example: High O-Swing% shows a pitcher induces chase.

OAA

Fielding

Outs Above Average

Statcast-based estimate of outs saved above average.

Example: +8 OAA marks very good fielding.

OBP

Batting

On-Base Percentage

How often a batter reaches base.

Formula: (H + BB + HBP) / (AB + BB + HBP + SF)
Example: .360 OBP is above league average.

Oppo%

Batting

Opposite Field Percentage

Share of batted balls hit to the opposite field.

Example: High Oppo% can indicate all-fields approach.

OPS

Batting

On-Base Plus Slugging

Simple combo of OBP and SLG to summarize production.

Formula: OBP + SLG
Example: .800 OPS is good; .900+ is elite.

OPS+

Batting

OPS Plus

OPS adjusted for park and league (100 = average).

Example: An OPS+ of 140 is 40% better than league average.

P

Parlay

Sports Betting

Parlay Bet

Combines multiple legs; all must win for payout.

Example: 3-leg moneyline parlay increases payout and variance.

Pipeline

Machine Learning

ML Pipeline

Chained steps for preprocessing and modeling.

Example: Scale → encode → model with cross-validation.

PostgreSQL

Data Engineering

PostgreSQL Database

Relational database used for storing structured data.

Example: Games, markets, and features stored in normalized tables.

Props

Sports Betting

Proposition Bets

Wagers on player/team events (e.g., strikeouts, hits).

Example: Over 6.5 strikeouts for a starting pitcher.

Pull%

Batting

Pull Percentage

Share of batted balls hit to the pull side.

Example: Pull-heavy hitters may be shift-vulnerable (rules permitting).

Pythag

Team

Pythagorean Expectation

Expected winning percentage from runs scored/allowed.

Formula: RS^x / (RS^x + RA^x), x≈1.83
Example: Used to spot teams over/underperforming record.

R

R²

Machine Learning

Coefficient of Determination

Share of variance explained by the model.

Example: R² of 0.25 is typical for noisy sports outcomes.

Random Search

Machine Learning

Randomized Hyperparameter Search

Sample parameter combinations randomly for efficiency.

Example: Use distributions for max_depth and min_child_weight.

RandomForestRegressor

Machine Learning

Random Forest Regressor

Ensemble of decision trees averaging predictions to reduce variance.

Example: Used to predict runs scored using many correlated features.

Rate Limiting

Tech

API Rate Limiting

Restrictions on request frequency.

Example: Throttle calls to theoddsapi.com to avoid 429 errors.

RE24

Team

Run Expectancy 24

Runs above/below average across the 24 base/out states.

Example: Captures value of sequencing and situational hitting.

Regularization

Machine Learning

L1/L2 Regularization

Penalties to reduce overfitting by shrinking weights.

Example: XGBoost’s lambda and alpha terms.

RMSE

Machine Learning

Root Mean Squared Error

Square root of MSE; error in original units.

Example: RMSE of 0.85 runs is easier to interpret than MSE.

Round Robin

Sports Betting

Round Robin Parlay

Creates multiple smaller parlays from a list of selections.

Example: A 3-team RR builds three 2-leg parlays.

Run Diff

Team

Run Differential

Runs scored minus runs allowed; proxy for team strength.

Example: +100 run differential typically correlates with strong records.

Run Line

Sports Betting

Run Line (Spread)

Baseball’s spread, typically ±1.5 runs.

Example: +1.5 underdog must lose by ≤1 or win.

S

SB%

Baserunning

Stolen Base Percentage

Success rate on steal attempts.

Formula: SB / (SB + CS)
Example: 75%+ is generally break-even or better.

SHAP

Machine Learning

SHapley Additive exPlanations

Game-theory method to attribute prediction contributions.

Example: Explains why a game total was projected high.

SIERA

Pitching

Skill-Interactive ERA

Advanced estimator of a pitcher's ERA using strikeouts, walks, and batted-ball mix.

Example: Often a strong predictor of future run prevention.

SLG

Batting

Slugging Percentage

Average total bases per at-bat.

Formula: TB / AB
Example: .500 SLG suggests strong power.

Spin Rate

Pitching

Pitch Spin Rate

Revolutions per minute (rpm) of a pitch impacting movement and shape.

Example: High-spin four-seamers can play well up in the zone.

Stuff+

Pitching

Stuff Plus

Index of pitch quality using velocity, movement, and release; 100 = average.

Example: 120 Stuff+ indicates well above-average raw stuff.

SwStr%

Batting

Swinging Strike Percentage

Share of pitches on which the batter swings and misses.

Example: Lower SwStr% correlates with better contact skills.

T

TimeSeriesSplit

Machine Learning

Time Series Cross-Validation

Sequential folds that respect temporal ordering.

Example: Walk-forward validation for daily MLB games.

Total

Sports Betting

Over/Under

Bet on combined runs scored by both teams.

Example: Over 8.5 cashes if 9+ total runs are scored.

Train/Test Split

Machine Learning

Holdout Validation

Partition data to assess generalization.

Example: Train on past seasons, test on current season.

U

UZR

Fielding

Ultimate Zone Rating

Runs saved using batted-ball location and difficulty.

Example: High UZR reflects range and sure-handedness.

V

Vig

Sports Betting

Juice / Vigorish

Bookmaker’s commission embedded in odds.

Example: -110 sides imply ~4.5% hold if lines are balanced.

W

WAR

Team

Wins Above Replacement

All-in-one value vs. a replacement player; scale varies by source.

Example: 5+ WAR ≈ All-Star-level season.

WHIP

Pitching

Walks plus Hits per Inning Pitched

Base runners allowed per inning.

Formula: (BB + H) / IP
Example: 1.10 WHIP is excellent.

wOBA

Batting

Weighted On-Base Average

Combines offensive events using run-value weights to measure overall batting quality.

Example: League average wOBA is roughly .320.

WPA

Team

Win Probability Added

Change in win probability credited to a player’s events.

Example: Clutch hits late can spike WPA.

wRC

Batting

Weighted Runs Created

Estimated runs a hitter contributes based on wOBA and context.

Example: Higher wRC indicates more total offensive value.

wRC+

Batting

Weighted Runs Created Plus

Park- and league-adjusted run creation index (100 = league average).

Example: 120 wRC+ means 20% better than league average.

X

xERA

Pitching

Expected ERA

Model-based expected runs allowed using contact quality and strikeouts/walks.

Example: xERA below ERA can hint at positive regression.

xFIP

Pitching

Expected Fielding Independent Pitching

Normalizes home run rate by using league-average HR/FB.

Example: If HR/FB was fluky, xFIP may differ from FIP.

XGBoost

Machine Learning

Extreme Gradient Boosting

Gradient-boosted trees optimized for speed and regularization.

Example: Commonly outperforms single models on tabular sports data.

Z

Zone%

Pitching

Zone Percentage

Share of pitches thrown in the strike zone.

Example: Combining Zone% with chase rate (O-Swing%) informs command vs. deception.