Bill James - Innovations

Innovations

Among the statistical innovations attributable to James are:

• Runs Created. A statistic intended to quantify a player's contribution to runs scored, as well as a team's expected number of runs scored. Runs created is calculated from other offensive statistics. James's first version of it was: Runs Created = (Total Bases * (Hits + Walks))/(Plate Appearances). Applied to an entire team or league, the statistic correlates closely to that team's or league's actual runs scored. Since James first created the statistic, sabermetricians have refined it to make it more accurate, and it is now used in many different variations.
• Range factor. A statistic that quantifies the defensive contribution of a player, calculated in its simplest form as RF = (Assists + Put Outs)/(Games Played). The statistic is premised on the notion that the total number of outs that a player participates in is more relevant in evaluating his defensive play than the percentage of cleanly handled chances as calculated by the conventional statistic Fielding Percentage.
• Defensive Efficiency Rating. A statistic that shows the percentage of balls in play a defense turns into an out. It is used to help determine a team's defensive ability. Calculated by: 1 - ((Opp. Hits + Reached on Error - Opp. Home runs) / (Plate appearances - Walks - Strikeouts - HitByPitch - Opp. Home runs)).
• Win Shares. A unifying statistic intended to allow the comparison of players at different positions, as well as players of different eras. Win Shares incorporates a variety of pitching, hitting and fielding statistics. One drawback of Win Shares is the difficulty of computing it.
• Pythagorean Winning Percentage. A statistic explaining the relationship of wins and losses to runs scored and runs allowed. In its simplest form: Pythagorean Winning Percentage equals Runs squared divided by the square of Runs plus the square of Runs Allowed. The statistic correlates closely to a team's actual winning percentage.
• Game Score is a metric to determine the strength of a pitcher in any particular baseball game.
• Major League Equivalency. A metric that uses minor league statistics to predict how a player is likely to perform at the major league level.
• The Brock2 System. A system for projecting a player's performance over the remainder of his career based on past performance and the aging process.
• Similarity scores. Scoring a player's statistical similarity to other players, providing a frame of reference for players of the distant past. Examples: Lou Gehrig comparable to Don Mattingly; Joe Jackson to Tony Oliva.
• Secondary average. A statistic that attempts to measure a player's contribution to an offense in ways not reflected in batting average. The formula is (Extra bases on hits+Walks+Stolen Bases)/At bats. Secondary averages tend to be similar to batting averages, but can vary widely, from less than .100 to more than .500 in extreme cases. Extra bases on hits is calculated with the formula (Doubles)+(Triplesx2)+(Homerunsx3) or more easily, (Total Bases)-(Hits).
• Power/Speed Number. A statistic that attempts to consolidate the various "clubs" of players with impressive numbers of both home runs and stolen bases (e.g., the "30/30" club (Bobby Bonds was well known for being a member), the "40/40" club (José Canseco was the first to perform this feat), and even the "25/65" club (Joe Morgan in the '70s)). The formula: (2x(Home Runs)x(Stolen Bases))/(Home Runs + Stolen Bases).
• Approximate Value. A system of cutoffs designed to estimate the value a player contributed to various category groups (including his team) to study broad questions such as "how do players age over time".
• "Temperature gauge" to determine how "hot" a player is, based on recent performance. The gauge has been used in NESN Red Sox telecasts and has provoked mixed reactions from critics.

Although James may be best known as an inventor of statistical tools, he has often written on the limitations of statistics and urged humility concerning their place amidst other kinds of information about baseball. To James, context is paramount: he was among the first to emphasize the importance of adjusting traditional statistics for park factors and to stress the role of luck in a pitcher's win-loss record. Many of his statistical innovations are arguably less important than the underlying ideas. When he introduced the notion of secondary average, it was as a vehicle for the then-counterintuitive concept that batting average represents only a fraction of a player's offensive contribution. (The runs-created statistic plays a similar role vis-à-vis the traditional RBI.) Some of his contributions to the language of baseball, like the idea of the "defensive spectrum", border on being entirely non-statistical.