Game score
Game Score is a baseball statistic created by sabermetrician Bill James to evaluate a starting pitcher's performance in a single game.[1] Introduced in his 1988 Baseball Abstract, it provides a numerical summary—typically ranging from 40 to 70 for average starts, with scores above 80 indicating exceptional outings—based on basic pitching lines like innings pitched, strikeouts, hits, runs, and walks.[2] The metric aims to capture overall effectiveness beyond wins or ERA, though it does not adjust for ballpark factors or opponent quality.[1] James developed Game Score as a simple "garbage stat" to quickly compare pitching starts across history.[3] The original formula is:Start with 50 points.
Add 1 point for each out recorded (or 3 points per complete inning).
Add 2 points for each inning completed after the fourth.
Add 1 point for each strikeout.
Subtract 2 points for each hit allowed.
Subtract 4 points for each earned run allowed.
Subtract 2 points for each unearned run allowed.
Subtract 1 point for each walk or hit by pitch.[1] This weights longevity and dominance while penalizing mistakes, resulting in scores from 0 upward, with rare perfect games or no-hitters often exceeding 90. An updated version by Tom Tango in 2014 incorporates home run penalties and adjusts the baseline to better reflect modern baseball.[1] Widely used in baseball analysis, Game Score is tracked by sites like Baseball-Reference.com, which list historical highs such as Kerry Wood's 105 on May 6, 1998—the highest in a nine-inning game.[4] It facilitates cross-era comparisons of pitching feats but has limitations, such as underpenalizing home runs in the original formula, making it a supplementary tool alongside metrics like ERA+ or FIP.[1]
Overview
Definition
Game Score (often abbreviated as GmSc) is a basketball statistic developed by analyst John Hollinger to measure an individual player's productivity and overall contribution in a single game, primarily using basic box score data from the National Basketball Association (NBA).[5] It serves as a simplified, linear extension of Hollinger's more complex Player Efficiency Rating (PER), providing a quick numerical summary without adjustments for team pace or minutes played.[5] The metric weights positive actions like points, field goals, rebounds, steals, assists, and blocks while penalizing inefficiencies such as missed shots, turnovers, and personal fouls.[5] Scores generally range from around 0 for poor performances to over 40 for exceptional ones, with a typical outing averaging approximately 10 points.[5][6]Purpose and Interpretation
Game Score was introduced by John Hollinger in the early 2000s as part of his pioneering work in basketball analytics, aiming to capture a player's impact through a straightforward, at-a-glance evaluation of their box-score line.[5] Its purpose is to synthesize key performance elements into a single number, facilitating rapid assessment of individual contributions without requiring deeper contextual analysis.[5] In interpretation, scores below 5 typically indicate subpar outings, while 10 represents an average performance; values of 15 or higher signify strong games, 20 or above denote excellent dominance, and 30 or more mark historic efforts.[6][7] Higher scores generally reflect greater productivity and correlate with positive team outcomes, though not causally, as they depend on factors like offensive support.[7] Despite its accessibility, Game Score has limitations, including no adjustments for playing time, game pace, opponent strength, or advanced defensive impacts, positioning it as a descriptive rather than predictive tool compared to metrics like PER or Real Plus-Minus.[5] As of 2025, it remains widely used in fan discussions, media recaps, and statistical databases for highlighting standout player performances.[7]Calculation
Original Formula
The original game score formula provides a numerical summary of a starting pitcher's performance by starting from a baseline and applying incremental adjustments for positive and negative outcomes. Introduced by Bill James in his 1988 Baseball Abstract, it aims to quantify effectiveness in a single metric scaled roughly from 0 to 100, with 50 representing an average outing.[8] The computation follows these steps:- Begin with a base score of 50 points.
- Add 1 point for each out recorded by the pitcher.
- Add 2 points for each full inning completed after the fourth.
- Add 1 point for each strikeout.
- Subtract 2 points for each hit allowed.
- Subtract 4 points for each earned run allowed.
- Subtract 2 points for each unearned run allowed.
- Subtract 1 point for each walk issued.
Components and Adjustments
The Game Score formula begins with a baseline of 50 points, providing a neutral starting point calibrated to reflect an average performance around that value.[1] This base allows the metric to scale positive and negative contributions symmetrically, typically resulting in scores between 40 and 70 for most starts, with exceptional outings exceeding 80.[9] Positive components emphasize a pitcher's ability to record outs and demonstrate dominance. Each out recorded contributes 1 point, equating to 3 points per complete inning, which rewards longevity and control in the game.[1] Additionally, 2 points are added for each full inning completed after the fourth, incentivizing pitchers to go deeper into contests beyond the typical early workload; this adjustment acknowledges the increasing value of sustained performance as games progress, and applies only to completed innings (partial innings do not qualify).[3] Strikeouts receive 1 point each, highlighting swing-and-miss ability as a marker of overpowering hitters, though the modest weighting reflects that outs via any means contribute to success.[9] Together, these elements weight endurance and strikeout-heavy dominance positively—for instance, a scoreless inning with three strikeouts and no baserunners yields +3 points for outs and +3 for strikeouts, totaling +6 points—while the post-fourth-inning bonus further boosts scores for complete games or deep outings. Game Score is typically applied to starting pitchers.[1] Negative components penalize events that allow baserunners and scoring, underscoring the importance of preventing opportunities for the offense. Hits deduct 2 points each, and walks deduct 1 point each, creating a heavier penalty for solid contact over free passes and emphasizing the control needed to limit traffic on the bases.[3] Runs allowed are penalized more severely: 4 points per earned run and 2 points per unearned run, differentiating the pitcher's direct responsibility from defensive miscues while still docking for overall game impact.[9] There is no direct penalty for errors committed by fielders, as these are indirectly addressed through the lighter unearned run deduction, avoiding over-punishment of the pitcher for team-wide failures.[1] Hit by pitches are not factored into the original formula, treating them as unpenalized events akin to unavoidable contact without the deduction applied to walks.[9] This structure prioritizes conceptual dominance by rewarding efficient outs and strikeouts while heavily weighting the prevention of baserunners and runs, ensuring the score captures both volume and quality of performance.[8]History and Variants
Introduction by Bill James
Bill James, a pioneering sabermetrician, first introduced the game score metric in the 1988 edition of his annual Baseball Abstract.[8] He presented it as a simple, composite statistic intended to provide a quick summary of a starting pitcher's performance in a single game, emphasizing its role as an entertaining rather than rigorously analytical tool.[3] James explicitly downplayed its seriousness, calling it "a kind of garbage stat that I present not because it helps us understand anything in particular but because it is fun to play around with."[11] James's philosophy behind game score reflected his broader approach to baseball statistics during the 1980s, favoring accessible, intuitive measures that could be grasped at a glance over more intricate metrics like earned run average (ERA) or walks plus hits per inning pitched (WHIP).[12] He designed it to capture the essence of a pitcher's outing through a single numerical value, allowing casual fans and analysts alike to evaluate starts without delving into box score minutiae.[2] This lighthearted intent aligned with James's self-published roots in sabermetrics, where he often experimented with fun, experimental stats to engage readers beyond traditional evaluations.[8] Upon its debut, game score quickly gained popularity among baseball fans for its straightforward appeal and ease of use in comparing individual performances.[13] However, some analysts critiqued it for its simplicity, arguing that the metric's arbitrary weighting overlooked nuances in defensive support, ballpark effects, and opponent quality.[3] James himself made no significant revisions to the original framework in subsequent works, maintaining its status as an unchanged, whimsical contribution to pitcher evaluation.[12]Alternative Versions
In 2003, Dayn Perry introduced Game Score 2.0 in a Baseball Prospectus article, aiming to better reflect defense-independent aspects of pitching performance by emphasizing controllable outcomes like strikeouts and unintentional walks while penalizing home runs more heavily. The formula begins with a base of 50 points, adds 1 point per batter retired, adds 2 points per inning completed after the fifth, adds 2 points per strikeout, subtracts 3 points per hit allowed, subtracts 5 points per non-home run run allowed, subtracts 4 points per unintentional walk, and subtracts 7 points per home run allowed; intentional walks are excluded from the walk penalty to avoid unfairly docking pitchers for strategic team decisions.[14] In 2014, statistician Tom Tango developed another revision, also termed Game Score Version 2.0, to align more closely with contemporary baseball dynamics, including increased emphasis on home run prevention and walk rates. This version starts at a baseline of 40 points (lowered from 50 to better represent replacement-level outings), awards 2 points per out recorded, adds 1 point per strikeout, subtracts 2 points per walk or hit allowed (doubling the original walk penalty to match hits), subtracts 3 points per run allowed regardless of earned status, and imposes an additional 6-point penalty per home run to account for their outsized impact.[1][11] Comparisons between versions highlight trade-offs in applicability: the original formula maintains stronger correlations with historical team winning percentages and ERAs due to its simpler structure suited to pre-2000s data, while Tango's iteration shows improved predictive power for modern win probabilities and ERA outcomes by incorporating era-specific weights like enhanced home run penalties.[15][11] Tango's version has seen widespread adoption, powering the game score metric on MLB.com and FanGraphs pitcher game logs (as GSv2), whereas Perry's 2.0 and the original remain referenced in sabermetric analyses for contextual comparisons across eras.[1][2]Notable Performances
All-Time Highest Scores
The all-time highest game score in Major League Baseball history is 153, achieved by Boston Braves pitcher Joe Oeschger on May 1, 1920, in a 26-inning tie against the Brooklyn Robins, where he allowed 9 hits, 1 unearned run, 4 walks, and 7 strikeouts while pitching the entire contest.[16] This extraordinary performance, though ending without a decision, exemplifies how extended games can dramatically elevate game scores due to the heavy weighting of innings pitched in the formula. In the same game, Brooklyn's Leon Cadore recorded the second-highest score of 140, surrendering 7 hits, 1 unearned run, 3 walks, and no strikeouts over the full 26 innings.[16] Among standard 9-inning games, the record stands at 105, set by Chicago Cubs pitcher Kerry Wood on May 6, 1998, against the Houston Astros, with 1 hit, 0 runs, 0 walks, and a major-league-record 20 strikeouts.[17] This mark highlights the metric's emphasis on dominance through strikeouts and minimal baserunners, often aligned with shutouts or near-no-hitters. Washington Nationals pitcher Max Scherzer posted the second-highest 9-inning score of 104 on October 3, 2015, against the New York Mets, completing a no-hitter with 0 runs, 0 walks, and 17 strikeouts.[18] Other notable high scores include those from extra-inning shutout bids, such as St. Louis Cardinals reliever Jose DeLeon's 103 on August 30, 1989, against the Cincinnati Reds (11 innings, 1 hit, 0 runs, 0 walks, 8 strikeouts).[16] In recent seasons, game scores remain impressive but have not surpassed these benchmarks, with St. Louis Cardinals pitcher Sonny Gray tying his career high of 96 on June 27, 2025, against the Cleveland Guardians (9 innings, 1 hit, 0 runs, 0 walks, 11 strikeouts).[19] Similarly, Detroit Tigers ace Tarik Skubal achieved 96 on May 25, 2025, versus the Guardians (9 innings, 2 hits, 0 runs, 0 walks, 13 strikeouts).[20] These outings underscore ongoing elite performances, often featuring efficient pitch counts under 100 in complete-game shutouts.| Rank | Pitcher | Date | Team vs. Opponent | IP | H | R | BB | SO | Game Score |
|---|---|---|---|---|---|---|---|---|---|
| 1 | Joe Oeschger | 5/1/1920 | Braves vs. Robins | 26 | 9 | 1 | 4 | 7 | 153 |
| 2 | Leon Cadore | 5/1/1920 | Robins vs. Braves | 26 | 7 | 1 | 3 | 0 | 140 |
| 3 | Dean Chance | 6/6/1964 | Angels vs. Yankees | 14 | 3 | 0 | 2 | 12 | 116 |
| 4 | Chris Short | 9/3/1966 | Phillies vs. Mets | 15 | 9 | 0 | 3 | 18 | 114 |
| 5 | Kerry Wood | 5/6/1998 | Cubs vs. Astros | 9 | 1 | 0 | 0 | 20 | 105 |
| 6 | Max Scherzer | 10/3/2015 | Nationals vs. Mets | 9 | 0 | 0 | 0 | 17 | 104 |
| 7 | Jose DeLeon | 8/30/1989 | Cardinals vs. Reds | 11 | 1 | 0 | 0 | 8 | 103 |
| 8 | Sandy Koufax | 9/9/1965 | Dodgers vs. Cubs | 9 | 0 | 0 | 0 | 14 | 101 |
| 9 | Nolan Ryan | 5/1/1991 | Rangers vs. Athletics | 9 | 0 | 0 | 2 | 16 | 101 |
| 10 | Randy Johnson | 5/18/2004 | Diamondbacks vs. Braves | 9 | 0 | 0 | 0 | 13 | 100 |
9-Inning Games Reaching 100
In Major League Baseball history, starting pitchers have achieved a game score of 100 or higher in exactly 16 nine-inning regular-season games through the 2025 season.[21] These rare performances highlight exceptional dominance, typically involving complete-game shutouts with minimal baserunners, often in the form of no-hitters or one-hitters. All 16 instances resulted in zero runs allowed, underscoring the threshold's association with scoreless outings; no nine-inning game reaching 100 has allowed any runs.[21] Eleven of the games were no-hitters, while the other five permitted exactly one hit. No such performances occurred in the 2024 or 2025 seasons.[21] The following table details each instance, ordered by descending game score. Key statistics include innings pitched (all 9.0), hits allowed (H), earned runs (ER, all 0), walks (BB), and strikeouts (SO).| Rank | Pitcher | Date | Team | Opponent | Result | H | BB | SO | Game Score | Notes |
|---|---|---|---|---|---|---|---|---|---|---|
| 1 | Kerry Wood | 1998-05-06 | CHC | HOU | W 2-0 | 1 | 0 | 20 | 105 | One-hitter |
| 2 | Max Scherzer | 2015-10-03 | WSN | NYM | W 2-0 | 0 | 0 | 17 | 104 | No-hitter |
| 3 | Clayton Kershaw | 2014-06-18 | LAD | COL | W 8-0 | 0 | 0 | 15 | 102 | No-hitter |
| 4 | Matt Cain | 2012-06-13 | SFG | HOU | W 10-0 | 0 | 0 | 14 | 101 | No-hitter |
| 5 | Nolan Ryan | 1991-05-01 | TEX | TOR | W 3-0 | 0 | 2 | 16 | 101 | No-hitter |
| 6 | Sandy Koufax | 1965-09-09 | LAD | CHC | W 1-0 | 0 | 0 | 14 | 101 | No-hitter |
| 7 | Nap Rucker | 1908-09-05 | BRO | BSN | W 6-0 | 0 | 0 | 14 | 101 | No-hitter |
| 8 | Justin Verlander | 2019-09-01 | HOU | TOR | W 2-0 | 0 | 1 | 14 | 100 | No-hitter |
| 9 | Gerrit Cole | 2018-05-04 | HOU | ARI | W 8-0 | 1 | 1 | 16 | 100 | One-hitter |
| 10 | Max Scherzer | 2015-06-14 | WSN | MIL | W 4-0 | 1 | 1 | 16 | 100 | One-hitter |
| 11 | Brandon Morrow | 2010-08-08 | TOR | TBR | W 1-0 | 1 | 2 | 17 | 100 | One-hitter |
| 12 | Randy Johnson | 2004-05-18 | ARI | ATL | W 2-0 | 0 | 0 | 13 | 100 | No-hitter |
| 13 | Curt Schilling | 2002-04-07 | ARI | MIL | W 2-0 | 1 | 2 | 17 | 100 | One-hitter |
| 14 | Nolan Ryan | 1973-07-15 | CAL | DET | W 6-0 | 0 | 4 | 17 | 100 | No-hitter |
| 15 | Nolan Ryan | 1972-07-09 | CAL | BOS | W 3-0 | 1 | 1 | 16 | 100 | One-hitter |
| 16 | Warren Spahn | 1960-09-16 | MLN | PHI | W 4-0 | 0 | 2 | 15 | 100 | No-hitter |
Career Totals for Key Pitchers
Among pitchers who achieved at least five starts with a game score of 90 or higher, Walter Johnson leads with 20 such outings during his career from 1907 to 1927.[22] Nolan Ryan recorded 15, spanning his 27-season tenure from 1966 to 1993.[23] Pedro Martinez amassed 10, primarily during his peak years with the Boston Red Sox in the late 1990s and early 2000s.[24] For even rarer feats, game scores of 100 or higher represent extraordinary dominance. Both Walter Johnson and Nolan Ryan tallied four such performances each, while other notable pitchers like Justin Verlander have achieved it once (in 2019). These career accumulations highlight the interplay of longevity, skill, and era-specific conditions in achieving high game scores. Dead-ball era pitchers like Johnson often posted higher totals due to lower offensive outputs and fewer earned runs, contrasting with modern hurlers facing elevated run environments despite advanced analytics and training.| Rank | Pitcher | Number of 90+ Starts | Career Span |
|---|---|---|---|
| 1 | Walter Johnson | 20 | 1907–1927 |
| 2 | Nolan Ryan | 15 | 1966–1993 |
| 3 | Cy Young | 14 | 1890–1911 |
| 4 | Christy Mathewson | 13 | 1900–1916 |
| 5 | Grover Alexander | 12 | 1911–1930 |
| 6 | Lefty Grove | 11 | 1925–1941 |
| 7 | Pedro Martinez | 10 | 1992–2009 |
| 8 | Roger Clemens | 9 | 1984–2007 |
| 9 | Randy Johnson | 8 | 1988–2009 |
| 10 | Warren Spahn | 7 | 1942–1965 |