Player Stats and Metrics Tracked in Fantasy Databases

Fantasy player databases don't just store box scores — they warehouse dozens of interconnected data points that determine whether a roster wins or loses on any given week. This page maps the full landscape of player statistics and metrics that power fantasy databases, from the raw counting stats every platform tracks to the derived efficiency figures that separate serious research tools from simple leaderboards. Understanding what gets tracked, how it gets calculated, and where the disagreements live helps explain why two databases can disagree on the same player's value.

• Definition and scope
• Core mechanics or structure
• Causal relationships or drivers
• Classification boundaries
• Tradeoffs and tensions
• Common misconceptions
• Checklist or steps
• Reference table or matrix

Definition and scope

A fantasy player database, at its foundation, is a structured repository that maps real-world athletic performance onto a scoring framework. The scope of what gets tracked has expanded considerably as data providers and sports leagues have made underlying event-level data more accessible. The player statistics and metrics layer of any database is the engine under everything else — projections, rankings, auction values, and trade tools all trace back to it.

The broadest definition covers three tiers of data: raw game statistics (touchdowns, yards, home runs, assists), derived or calculated metrics (yards per carry, true shooting percentage, strikeouts per nine innings), and contextual metadata (snap count percentage, usage rate, air yards target share). Most consumer-facing fantasy platforms display the first tier. Research-grade databases at tools like Pro Football Reference or Baseball Reference make all three accessible, and the gap between what a casual user sees and what an analyst pulls from the API is substantial — sometimes more than 200 distinct tracked fields per player per season.

The key dimensions and scopes of fantasy player databases define which sports and which scoring systems shape the specific metrics a given database prioritizes. A database built for daily fantasy sports will weight ceiling and variance metrics differently than one designed for redraft season-long leagues.

Core mechanics or structure

Raw statistics are collected at the event level — each play, plate appearance, possession, or shift generates discrete data points. Data providers like Sportradar, Stats Perform, and Elias Sports Bureau serve as primary wholesalers for this event-level feed, which platforms then aggregate and transform.

The transformation pipeline generally follows four stages:

1. Event capture — pitch velocity, route run, shot location, shift positioning logged in real time.
2. Game aggregation — event data rolled into per-game totals (passing yards, rushing attempts, targets).
3. Season accumulation — game-level data stacked into season and career figures.
4. Derived metric calculation — formulas applied to produce efficiency stats, rate stats, and composite scores.

Derived metrics are where databases diverge most sharply. Expected points added (EPA) in football, weighted on-base average (wOBA) in baseball, and player efficiency rating (PER) in basketball all require specific methodological choices — which baseline population, which park factors, which down-and-distance weights. FanGraphs, for instance, defines wOBA using linear weights that are recalculated annually based on run environment (FanGraphs Glossary, wOBA), which means the same raw stat line produces a slightly different wOBA in different seasons.

For real-time data updates, the pipeline compresses dramatically — event data feeds into fantasy scoring engines within seconds during live games, though derived metrics typically update on a longer cycle (daily or weekly).

Causal relationships or drivers

The metrics that matter most in fantasy databases aren't simply the ones worth the most points — they're the ones that most reliably predict future point production. This distinction drives how advanced analytics for fantasy players gets built.

Snap count percentage in NFL data, for example, is a leading indicator of target opportunity, which itself precedes receiving yards and touchdowns in the causal chain. A receiver playing 85% of snaps on a pass-heavy team sits at the top of the opportunity funnel before a single target is thrown. Plate appearances and at-bats function similarly in baseball — volume metrics upstream of outcomes.

The research literature on fantasy-relevant predictability leans heavily on work from analysts at Football Outsiders, who developed DVOA (Defense-adjusted Value Over Average) as a team and player efficiency metric, and from the baseball analytics community's adoption of FIP (Fielding Independent Pitching) as a pitcher evaluation tool more stable than ERA. FIP removes defense from pitcher evaluation entirely, relying on strikeouts, walks, hit batters, and home runs — the four outcomes a pitcher controls directly (FanGraphs FIP definition).

The causal chain runs: opportunity → efficiency → outcomes → fantasy points. Databases that track only outcomes miss the upstream signals that explain variance and predict regression.

Classification boundaries

Not all tracked metrics belong in the same analytical bucket. The classification breaks down along two axes: input vs. output and rate vs. counting.

Input metrics measure opportunity and usage — targets, carries, air yards, time on ice, plate appearances. Output metrics measure results — touchdowns, yards, home runs, saves. Rate metrics express the ratio (yards per carry, target share percentage, hard-hit rate). Counting metrics express the total (total targets, total carries, total assists).

A secondary classification distinguishes official statistics — maintained by league bodies like MLB's official stats office or the NFL's statistical service — from proprietary or analytical metrics generated by third-party researchers. Air yards, route participation, and separation distance at the catch point are not official NFL stats; they're tracked by private data vendors and vary in methodology across providers.

The data sources and provider standards question determines which classification system a given database adopts and how it handles conflicting figures from different vendors.

Tradeoffs and tensions

The central tension in metric selection is stability versus descriptiveness. A metric can be highly descriptive — telling exactly what happened — while being nearly useless for prediction because it's driven largely by luck. Batting average on balls in play (BABIP) is one famous example: a pitcher's BABIP against in a single season carries roughly 0.15 correlation with the following season, which is weak enough that Baseball Reference displays it prominently as a regression warning flag for analysts, not as a performance indicator.

On the other side, highly stable metrics (like strikeout rate or yards per route run) are predictive but can feel abstract to users who intuitively anchor to outcomes. A receiver with an excellent yards-per-route-run figure but low touchdowns looks worse in a scoring summary than a player who scored twice on limited targets, even if the underlying efficiency tells a reversed story.

A second tension exists between granularity and accessibility. Play-by-play data for a full NFL season generates millions of discrete rows. Most fantasy managers don't query raw play-level data — they use aggregated surfaces. The aggregation choices made by database builders (which splits to surface, which filters to apply, which thresholds to require for minimum sample size) shape what conclusions are even possible at the user layer. The database search and filtering tools a platform provides directly reflects these backend decisions.

Common misconceptions

Misconception 1: More metrics always means better analysis.
A database with 400 tracked fields is not inherently more useful than one with 60. Metric proliferation without clear definitions creates noise. Analysts at Football Outsiders have noted that some "next-generation" tracking stats show lower predictive validity than simple market-derived proxies like snap count and target share, which have years of validation behind them.

Misconception 2: Official stats are neutral and complete.
Official statistics are defined by administrative decisions, not analytical ones. The official NFL rushing statistic, for example, counts yards after contact the same as yards before contact — a methodological choice that obscures whether a running back or an offensive line is responsible for production. Databases that track yards before contact as a separate field (via vendors like Pro Football Focus) tell a materially different story about the same players.

Misconception 3: Historical statistics are stable.
Historical performance data is periodically revised. MLB retroactively corrects errors in official scorekeeping, which changes career totals. The Hall of Fame voting implications of retroactive stat corrections — such as the career hit total adjustments made for several 19th-century players documented by Retrosheet — illustrate that even the most settled-seeming numbers carry revision risk.

Misconception 4: Advanced metrics are sport-agnostic.
True shooting percentage in basketball and adjusted OPS+ in baseball require completely different underlying data and answer different questions. Crossover application of frameworks from one sport to another (applying "usage rate" logic from basketball directly to NFL receiver analysis, for instance) requires significant methodological translation and can produce misleading comparisons.

Checklist or steps

Evaluating the metric coverage of a fantasy player database:

Reference table or matrix

Stat Category Map by Sport and Metric Type

The player rankings methodology used by any platform ultimately reflects which metrics appear in the table above — and which ones get weighted most heavily in aggregation. The full fantasy player database home provides context on how these components connect across sports.