About
What is QuantSports?
QuantSports is a sports analytics platform covering the NBA, MLB, and NFL. Every number on this site is computed from raw play-by-play and box-score data. No borrowed metrics from ESPN, FanGraphs, Baseball Savant, or anyone else. Each sport has its own pipeline, its own models, and its own set of proprietary statistics. The sections below explain what those statistics capture and, at a high level, how they work.
NBA
Traditional box-score stats tell you what happened but not how much it mattered. A player can average 20 points on a bad team and look identical in the box score to someone averaging 20 points while anchoring a top-five defense. Our NBA metrics try to close that gap by measuring impact: what actually changes when a player is on the court.
QWAR (Quant Wins Above Replacement)
An all-in-one wins metric that estimates how many additional wins a player contributes over a replacement-level player across a full season. Built on a proprietary plus-minus model that isolates each player's effect from possession-level lineup data, informed by play-type priors (isolation, pick-and-roll, spot-up, transition, etc.) and hustle inputs (deflections, contested shots, loose balls). Split into O-WAR and D-WAR.
Q Impact / QI per 100
A player's impact measured in points per 100 possessions, the rate-stat companion to QWAR. Where QWAR rewards both quality and playing time, Q Impact isolates per-possession value. Broken into O-Impact (offensive) and D-Impact (defensive).
Skill Breakdown (Scoring, Playmaking, Rebounding, Defense)
QWAR decomposed into four components so you can see where a player's value actually comes from. Each component is expressed in WAR units and sums to total QWAR.
Consistency
A 0-100 score capturing how stable a player's output is from game to game. Built from the statistical variance of per-game performance. A high score means you know what you're getting every night; a low score means boom-or-bust.
Durability
A 0-100 score projecting availability. Factors in injury history (weighted by injury type), age curves, body-stress indicators, and workload. A player with a torn ACL history and high usage will score differently than a low-mileage wing of the same age.
Trade Value
A 0-99 composite asset score designed to approximate real front-office valuations. Prioritizes offensive creation upside, then adjusts for age, contract, positional scarcity, and archetype demand. Tiered from MVP Candidate down to Below Replacement.
Surplus
The dollar difference between a player's on-court production value (derived from QWAR and the current market price per win) and their actual salary. Positive surplus means the team is getting more value than it's paying for.
MLB
Baseball analytics already has established WAR implementations from FanGraphs and Baseball Reference. Ours differs in a key way: we build entirely from a context-sensitive run-expectancy framework using raw event data. Every plate appearance, every pitch, and every fielding play is valued by the base-out state it occurs in, with event-level park adjustments rather than league-wide averages.
QWAR
Total wins above replacement combining batting, pitching, fielding, and baserunning value. Each component uses a run-expectancy matrix that scores every event by the base-out context it occurs in, with park factors applied per batted-ball type, spray direction, handedness, and exit-velocity bin. Replacement level is set empirically each season from actual fringe-roster performance.
Batting / Pitching / Fielding / Baserunning
The four components of QWAR. Batting captures offensive run production from plate appearances. Pitching is split into DI Runs (direct-impact events like strikeouts and walks) and BIP Runs (balls-in-play outcomes). Fielding uses a catch-probability model with Bayesian smoothing across positions. Baserunning scores stolen bases, advancement, and tagging decisions against the run-expectancy matrix.
Q-Stuff
A composite pitch-quality score for pitchers that captures how difficult a pitcher's arsenal is to hit, independent of outcomes. Synthesizes called-strike-plus-whiff rate, whiff rate, and velocity into a single percentile ranking against the pitcher population.
Q-Contact
The hitter counterpart to Q-Stuff: a composite measuring how dangerous a hitter's contact is. Blends power indicators (barrel rate, hard-hit rate, exit velocity), bat-to-ball ability, and plate discipline into one percentile score.
Consistency
Game-to-game reliability on a 0-100 scale, derived from the variance of per-game performance. Separate game-score formulas for hitters and pitchers ensure apples-to-apples comparisons within each group.
Durability
A 0-100 availability score weighting games-played rate, age risk, surgery history (including Tommy John), and recent health trends.
Trade Value
A 0-99 asset score with a nonlinear production curve. Superstar production is worth disproportionately more than replacement-level production. Layered with age-projected decline, years of team control, contract surplus, and positional scarcity. Relievers are capped at 60.
Surplus Value
Production value minus salary. Converts QWAR to dollars at the current market rate per win, then subtracts the player's contract. Positive means the team is getting a bargain.
NFL
Football is harder to model than baseball or basketball. Play outcomes depend on 22 players at once and sample sizes are small (17 games). Our NFL metrics take two complementary approaches: an EPA-based efficiency view and a counting-stats volume view, then combine them with a position-specific composite grade.
Q Impact
A player's estimated impact on team point differential using seasonal Expected Points Added. Captures per-play efficiency and accumulates it over the season, then ranks each player against their position peer group.
QWAR
The counting-stats companion to Q Impact: wins above replacement derived from volume production. Position-specific formulas weight yards, touchdowns, turnovers, and defensive disruptions into a single wins number.
Q Grade
A composite 0-99 efficiency grade that synthesizes per-play efficiency, opportunity quality, and advanced metrics into one position-specific score. Different positions weight different inputs: QBs lean on EPA per dropback and completion percentage over expected, receivers lean on EPA per target and weighted opportunity rating, defenders lean on disruption and pass-rush rates.
Consistency
Game-to-game variance on a 0-100 scale. Built from the coefficient of variation across regular-season performances. In a 17-game season, separating the reliable producers from the week-to-week coin flips matters more than in other sports.
Durability
A 0-100 availability score factoring in games played against a 17-game season, two years of injury history, and age-based risk tiers.
Trade Value
A 0-99 asset score incorporating current production, age decay by position, and years of contract control. Tiered into Franchise, Blue-chip, Starter-level, Depth, and Fringe.
Surplus ($/WAR)
Production value minus cap hit. Converts QWAR to dollars at a position-specific market rate, then subtracts the player's cap number. Positive surplus means the player is outperforming their deal.
Fantasy
Every player page has a dedicated Fantasy tab covering projections, usage trends, and reliability scores calibrated to each sport's scoring formats. The goal is to surface the signal that actually predicts weekly output, not just what looked good last week.
Projected Fantasy Points
A per-game projection based on role, recent usage, matchup, and pace. Updated continuously through the season as lineup and opportunity data changes. Available for roto and points-league formats across NBA, MLB, and NFL.
Usage Trend
Tracks whether a player's opportunity is expanding or contracting. Shot attempts, target share, innings pitched, plate appearances per game: whichever inputs drive fantasy scoring for that position are tracked and flagged when the trend changes materially.
Consistency Score
The same 0-100 game-to-game reliability metric used in the core player profile, re-expressed relative to fantasy scoring. High consistency means a player delivers a predictable floor; low consistency means high variance upside with matching downside risk.
Injury Risk
The durability score reframed for fantasy: how likely is this player to miss games, and how does their injury history pattern affect weekly start decisions. Factors in position-specific injury frequency alongside the player's own history.
Betting
Six proprietary metrics are computed for every player-prop line across all three sports. Each metric uses Bayesian inference with position-specific population priors, tuned per sport. Rolling windows, recency half-lives, and cross-season rollover are all calibrated to the length and rhythm of each league's schedule.
Over Rate
The Bayesian probability a player exceeds a given prop threshold, informed by a position-level population prior and recency-weighted to emphasize recent form.
Average Margin
How far above or below the line a player typically lands. Uses a trimmed mean to remove outliers, with a full percentile distribution underneath for context.
Confidence
The strength of statistical evidence that a real edge exists on a given prop. Evaluates the Bayesian posterior against the breakeven threshold for standard -110 odds and classifies evidence from anecdotal to decisive.
Consistency
How repeatable a player's performances are relative to the line. Uses a two-state model to detect whether a player's game-to-game variance reflects genuine performance modes or just noise.
Trend
Whether a player's production is statistically rising, falling, or flat. Combines change-point detection with a non-parametric trend test to separate real shifts from random fluctuation.
Streak
The probability and expected duration of a hot or cold streak. Models consecutive overs and unders as a Markov process and tests whether observed streaks are statistically real or just chance.