clock menu more-arrow no yes mobile

Filed under:

Keuchel and Pitcher Fielding

Talking Astros Sabermetrics: Keuchel, Pitcher Fielding, FIP, and ERA

Peter G. Aiken-USA TODAY Sports

Astros' pitcher Dallas Keuchel had a break out season as one of the more productive starting pitchers in the majors. And on top of that, he is a finalist for the AL pitcher gold glove. As I discussed in this article last year, Rawlings has incorporated sabermetrics (25% weighting) into the gold glove selection process. My guess is that Keuchel made the cut as a finalist for the gold glove award because of the newly added influence of sabermetrics on the award.

Keuchel was the best fielding pitcher in the major leagues, according to the Fielding Bible's Defensive Runs Saved (DRS). And, he was best by a nice margin (Clayton Kershaw in the NL was second best). Keuchel was +10 runs saved, which is equal to more than a full win. Fangraphs' Wins Above Replacement (WAR) for pitchers is based on Fielding Independent Pitching (FIP), which excludes the effects of defense. If Keuchel's DRS was included, he would be a 5 WAR pitcher; only ten ML pitchers had more than 5 WAR last season.

Keuchel gained 1 DRS run due to his excellent ability to suppress stolen bases and another DRS run based on his above average handling of bunts. The biggest part of Keuchel's DRS score (7 runs) is based on his range in getting to balls hit to his left/right side.

When I examined the listing (by DRS) of top fielding pitchers, I noticed something that sparked my interest. It seemed like several of the best fielders were also leaders in producing ERAs below their FIP. Normally we think of these pitchers as prime candidates for regression. Johnny Cueto and R.A. Dickey, regarded by some as anomalies who repeatedly beat their FIPs, are numbers 4 and 5 on the DRS list. As I looked further, I found that the very top fielding pitchers, by DRS, tend to outperform their FIP.

The table below shows all qualified pitchers with DRS of 5 or greater. The "E - F" column refers to ERA minus FIP; a negative in this column means that ERA is less than FIP.

Dallas Keuchel 10 -0.28
Clayton Kershaw 7 -0.04
Henderson Alvarez 6 -0.93
Johnny Cueto 6 -1.05
R.A. Dickey 5 -0.61
Zack Greinke 5 -0.26
Adam Wainwright 5 -0.5
Kyle Kendrick 5 -0.04
Julio Teheran 5 -0.6

All of the pitchers who fielded at a rate of 5 runs saved or more have an ERA less than their FIP. This suggests that the pitchers who are elite at saving runs with their glove can achieve ERAs less than their FIP. The result for pitchers who are above average fielders but saved fewer runs is more mixed, in terms of E-F results.

The results for the 10 worst fielding pitchers, based on DRS, are shown below. A positive E-F means that the ERA was above the FIP.

Name DRS E - F
Phil Hughes -8 0.87
Scott Feldman -7 -0.37
A.J. Burnett -5 0.45
John Lackey -5 0.04
Colby Lewis -4 0.71
Max Scherzer -4 0.3
Chris Archer -4 -0.06
Drew Hutchison -4 0.64
Jon Lester -3 -0.34

The comparison to E-F is not as uniformly consistent as the previous table, but the result is in the expected direction. 7 of the 10 worst fielding pitchers experienced an ERA above their FIP. Feldman, Lester, and Archer fielded poorly but still kept their ERA below their FIP.

To some extent I'm not surprised by these results. Because FIP excludes batted ball results, we have always known that team defense is one of the more significant routes for suppressing ERA below FIP. Pitcher defense is one component of that defense, although it is a skill or trait confined to the pitcher, himself. It is also worth noting that pitcher fielding chances are less frequent, compared to position players. Thus, the sample size for balls fielded by the pitcher is relatively small; as a result the DRS results may be subject to regression in the future.

Some observations:

  • While the usual conclusion is that the ERA over-performers (relative to FIP) are likely to regress toward a higher ERA in the future, it's possible that the pitcher's inherent skill in fielding balls will mitigate the tendency for the ERA to regress.  Presumably, fielding ability is a repeatable skill (even if subject to some regression, itself), and it's possible that the pitcher can "help himself" through his fielding, thereby maintaining a tendency to suppress ERA below FIP.
  • If Keuchel can continue to field at a high level those balls hit toward the pitcher, it's possible that Keuchel isn't a significant candidate for upward regression in his ERA.
  • Scott Feldman can improve his future performance if he can find a way to improve his fielding results, even marginally.  Feldman has shown the capability to maintain an ERA below his FIP, despite his own poor fielding.  Feldman is notoriously slow to the plate and has shown a weakness in holding runners.  Most of Feldman's sub-par DRS performance is due to the negative 5 runs from stolen bases.  Only 17% of runners are unsuccessful in their steal attempts against Feldman.
  • Given the many possible factors which may cause ERA to deviate from FIP, it's surprising that pitcher fielding appears to have such a strong effect on whether ERA is above or below FIP. 
  • I wouldn't rule out the possibility that pitcher fielding DRS is indirectly influenced by the pitcher's ability to induce weak contact.  Given the location of balls fielded by pitchers, I would speculate that weak or strong contact is a major influence on whether the ball can be fielded.  While this may be generally true for all balls in play, I suspect it is a more significant effect for pitchers, given the proximity of the pitcher to the batter. DRS includes a subjective factor which is supposed to account for how hard the ball is hit, but I don't know if it is completely effective for the types of balls fielded by pitchers. Even if pitcher DRS is partially dependent on weak contact, the result should still be credited to the pitcher, since it is part of his skill set.