What is Game Score? Definition, Formula, and Example
Game Score is a single-number metric created by Bill James that grades an individual starting pitcher's performance in one game, where 50 is average and 100+ is historically elite.
Plain-English Definition
Game Score is the cleanest way to answer "how good was that start?" in one number. Bill James invented it in the 1980s as a quick-and-dirty pitcher-of-the-day grade, and the framework has stuck around for forty years because it actually works. The original formulation centers on 50 as a league-average start; a Game Score of 70 is excellent, 80 is dominant, and anything above 90 belongs on the season highlight reel. Tom Tango later published an updated "Game Score v2" tuned to modern offense and three-true-outcomes baseball, and most public stat sites now display both numbers side by side.
How Game Score Is Calculated
Game Score v1 (Bill James, original):
- Start at 50
- +1 for each out recorded
- +2 for each full inning completed after the 4th
- +1 for each strikeout
- −2 for each hit allowed
- −4 for each earned run
- −2 for each unearned run
- −1 for each walk
Game Score v2 (Tom Tango):
- Start at 40
- +2 for each out
- +1 for each strikeout
- −2 for each walk
- −2 for each hit
- −3 for each run (earned or unearned)
- −6 for each home run
V2 punishes home runs explicitly (v1 doesn't) and treats unearned runs the same as earned, both of which better reflect modern run-scoring environments.
Worked Example
Max Scherzer's October 2015 no-hitter against the Mets: 9 IP, 0 H, 0 R, 0 BB, 17 K. Plugging into v1: 50 + 27 outs + (2 × 5 innings after the 4th) + 17 K − 0 hits − 0 ER − 0 BB = 104. The v2 version comes out to roughly 100. Both flag it as one of the most dominant starts in modern history. Compare to a typical six-inning, two-run, six-strikeout outing — your standard quality start — which scores around 58 in v1, exactly where you'd expect "slightly above average" to land.
Why Game Score Matters
Game Score is the single best one-number tool for comparing starts across games, seasons, or eras. It powers "Game Score 80+" leaderboards that broadcasters lean on during postseason coverage. Front offices use Game Score thresholds (often 65+) to define "dominant starts" when building bonus structures or evaluating swingmen. For DFS, a starter who consistently posts Game Scores of 60+ across his last five outings is a much safer cash-game play than one bouncing between 35 and 80. It also serves as a sanity check on flashier metrics — if a pitcher's FIP and ERA disagree, his Game Score distribution often reveals which is closer to the truth.
Limitations and Common Misconceptions
Game Score doesn't adjust for opponent, ballpark, weather, or defense behind the pitcher — a 75 against the 2003 Tigers means less than a 75 against the 2019 Astros. V1 over-rewards innings (a 7-inning, 4-run start can outscore a 5-inning, 0-run gem), which is exactly the gap v2 was built to close. It's also frequently confused with WPA — but WPA measures leverage-weighted win contribution, while Game Score is a context-neutral box-score grade.
Related Terms
In Legends Deck
Every pitcher card in Legends Deck carries a rolling Game Score distribution from the player's real MLB starts — not just an average, but the full shape (median, 25th, 75th percentile). That's what drives the variance in our start simulation: ace cards don't just have higher means, they have tighter distributions, so they bust less often. Build a rotation of high-floor 60-Game-Score starters and you'll see fewer 8-run blowups across a simulated season.