WARP - Pitcher WARP computed using DRA, roughly IP/9 * ([Replacement RA]-DRA) / [Runs Per Win].
CSAA - The total runs gained or lost from the net strike zone created by (1) batters faced, (2) umpires faced, and (3) the catcher receiving the pitcher; these are imported CSAA values for the season. Catcher framing is by far the biggest contributor to this.
FRAA - The total runs lost from good or bad defense (created by averaging the total FRAA/PA of the defenders on the field for each play).
Temp - The total runs gained or lost from temperature.
Stadium - The total runs gained or lost from the combination of right- and left-handed batters faced in the parks where the pitcher pitched.
Opposing Batter - The total runs gained or lost by the strength or weakness of opposing batters.
Role - Compensates for being a starter versus reliever (e.g., all full-time starters are compensated a few runs for this).
Home/Away - The total runs gained or lost from a pitcher's split of home versus away appearances. This includes a general home/away effect, as well as the additional effect of being at home on defense and of batting in the bottom of innings.
Adj Runs - The total number of plus / minus runs in the previous columns. There are other adjustments DRA makes as well: this is just the net effect of the factors in this particular spreadsheet, which are the ones people tend to ask about the most.
Technical Note: These run tallies are estimates derived by multiplying the average effect of each factor on each pitcher during a season times the number of batters the pitcher faced during that season. In other words, they are based off point estimates correlated to linear weights. However, run-scoring in baseball is not completely linear. Pitchers with fewer batters faced may derive less than the average benefit and pitchers who face more batters may derive more benefit than average. Each pitcher's final DRA compensates for this non-linearity; these run estimates do not because that adjustment is very complicated. However, it is accepted sabermetric practice to base run effects off the average impact of a baseball event, and we do so here. To the extent the "true" values for each differ at all from these, they are unlikely to substantially change the order of pitchers in each category, and at most might stretch the estimates out slightly for pitchers on the extremes. |
