Table 1: ERA for American League Teams, 2011
ERA | |
Los Angeles | 3.57 |
Tampa Bay | 3.58 |
Oakland | 3.71 |
New York | 3.73 |
Texas | 3.79 |
Seattle | 3.91 |
Detroit | 4.04 |
Chicago | 4.10 |
Boston | 4.20 |
Cleveland | 4.24 |
Toronto | 4.33 |
Kansas City | 4.45 |
Minnesota | 4.60 |
Baltimore | 4.92 |
ERA has traditionally been used instead of Run Average (same as ERA, except runs replaces earned runs) because it supposedly isolates pitching from fielding by removing fielding errors from a pitcher's responsibility. However, there are a few problems with using ERA to separate pitching from fielding. The most obvious is potential official scorer bias. For example. if one team's official scorer is conservative in assigning errors, this might inflate that team's ERA in comparison to other teams.
A second shortcoming of ERA is that the only fielding events which are excluded from a pitcher's record are errors. A pitcher gets credit for everything else that may affect run scoring, such as great plays by teammates and balls that drop in front of fielders. In reality, some pitching staffs get more help from fielders than others and this is not reflected in ERA.
A third issue is that a pitcher can give up a bunch of runs in an inning and not get charged for any of them simply because one error was made. For example, suppose Max Scherzer retires the first two batters in an inning and the next one reaches on a throwing error by Brandon Inge. Then suppose that Scherzer proceeds to allow a double, a walk and a grand slam. That would be a poor job of pitching by Scherzer, but he would not be charged with any earned runs because Inge's error should have been the third out. The ERA statistic would be an inappropriate representation of Scherzer's performance in that case.
Finally, some believe that ERA unfairly favors ground ball pitchers compared to fly ball pitchers. This is because ground balls are more likely to be errors than fly balls and thus more likely lead to unearned runs.
It is my belief that pitching analysis should ideally start (but not end) with runs allowed because a pitcher's job is to limit runs as much as possible. The Tigers allowed 711 runs in 2011. The league average was 717 runs allowed, so the Tigers were six runs Saved Above Average (RSAA). Table 2 below illustrates that the Tigers were eighth in the AL by this measure.
Table 2: Runs Saved Above Average for American League Teams, 2011
Team | RSAA |
Tampa Bay | 103 |
Los Angeles | 84 |
New York | 60 |
Seattle | 42 |
Texas | 40 |
Oakland | 38 |
Chicago | 11 |
Detroit | 6 |
Boston | -20 |
Cleveland | -43 |
Toronto | -44 |
Kansas City | -45 |
Minnesota | -87 |
Baltimore | -143 |
The RSAA statistic is a good start, but it's not much better than ERA. We need to separate pitching from fielding. One way to do this is to calculate runs saved or cost by a team's fielding and then subtract that from RAA to determine runs allowed by pitchers. This is actually the basis of the Baseball-Reference Pitching Wins Above Replacement (WAR) statistic.
I'm going to create something similar to Baseball-Reference WAR using the team fielding runs saved/cost calculated in an earlier post. Keeping in mind that fielding measurement is uncertain, the Tigers fielding cost them an estimated 21 runs (the average of Total Zone and BIS Defensive Runs Saved) compared to the average team in 2011. I'll call this number Fielding Runs Saved (FRS) Subtracting FRS (-21) from RSAA (6), we estimate that the Tigers pitching saved 27 runs above the average team. So, in taking fielding into consideration, the Tigers pitching staff looks better than it did when we just looked at ERA or RSAA.
In addition to differences in fielding support, remember that some teams play in more pitcher-friendly parks than others. For example, Mariners hurlers have an easier time limiting runs in Safeco Field than the Rangers pitchers do in The Ballpark at Arlington. We can use park factors to estimate how many runs a park saved or cost a team's pitching staff. I'm going to use the five-year factors created by US Patriot, who writes the Walk Like a Sabermetrician blog. For those who don't know Patriot, he is one of the most respected statisticians in the hardcore saber community and I have learned a lot from his site.
According to Patriot's calculations, the Comerica Park Factor is 1.015. A Park Factor above one indicates a better than average run scoring park. So, Comerica Park plays slightly in favor of hitters: Tigers runs and runs allowed totals are inflated by an estimate 1.5 percent compared to a team playing in an average park. A detailed explanation of Park Factors can be found here.
If we multiply the league average runs allowed (717) by .015, we get 11 runs. So, Comerica Park cost the Tigers staff an estimated 11 runs above an average staff. I'll call this number Ballpark Runs (BPR). In the Tigers case, BPR= -11. Table 3 shows that this number ranged from + 28 runs saved at Tropicana Field in Tampa Bay to -36 runs cost at Fenway in Boston.
Table 3: Ballpark Runs Saved/Cost for AL Teams, 2011
Team | BPR |
Tampa Bay | 28 |
Seattle | 24 |
Oakland | 19 |
Los Angeles | 11 |
Minnesota | 11 |
Cleveland | 6 |
Toronto | 0 |
Kansas City | -10 |
Detroit | -11 |
Baltimore | -16 |
New York | -22 |
Chicago | -24 |
Texas | -34 |
Boston | -36 |
We can then subtract FRS and BPR from RSAA to get Pitching Runs Saved (PRS): RSAA - FRS - BPR = +6 - (-21) - (-11) = +38 for the Tigers. Table 4 below shows that the Tigers were 4th in the league in PRS.
Table 4: Pitching Runs Saved/Cost for AL Teams, 2011
Team | RSAA | BPR | FRS | PRS |
New York | 60 | -22 | -2 | 84 |
Chicago | 11 | -24 | -8 | 43 |
Texas | 40 | -34 | 34 | 40 |
Detroit | 6 | -11 | -21 | 38 |
Oakland | 38 | 19 | -17 | 36 |
Seattle | 42 | 24 | -1 | 19 |
Los Angeles | 84 | 11 | 54 | 19 |
Boston | -20 | -36 | -1 | 17 |
Tampa Bay | 103 | 28 | 60 | 15 |
Kansas City | -45 | -10 | 0 | -35 |
Toronto | -44 | 0 | 16 | -60 |
Cleveland | -43 | 6 | 15 | -64 |
Minnesota | -87 | 11 | -24 | -74 |
Baltimore | -143 | -16 | -42 | -85 |
So, according to this statistic, the Tigers pitching staff was not merely average (as ERA suggests), but significantly better than average. Other teams that look better according to PRS are the Yankees (4th in ERA, 1st on PRS) and White Sox (8th and 2nd). Some teams who look worse with PRS are the Angles (1st in ERA, 7th in PRS) and Rays (2nd and 9th).
Because of the uncertainty of the team fielding component, PRS should not be used as an end all of pitching analysis. Instead it should be viewed as another piece of evidence in pitching evaluation. Other pitching statistics will be discussed in later posts.
Note: Some of the data for this analysis was extracted from Baseball-Reference.com
No comments:
Post a Comment