J.D. Martinez can't make the catch on fly ball off the wall by Ryan Howard at MMP last week. (Photo by Bob Levey/Getty Images)
We know that the Astros' pitching has been a problem this year. That is a statement you would expect for a team that will reach it's 100th loss in the near future. The Astros' team ERA (4.61) is 4th worst in the majors and 2d worst in the NL (the Rockies racked up a 5.61 ERA).
But if you have been following the advanced pitching stats, you know that the Astros' pitching isn't quite as bad as the ERA result. The difference between the Astros' FIP and ERA is one indication. The Astros pitchers' ERA exceeds the team's FIP by the fourth highest margin in the major leagues. My previous articles in this series have looked closer at the difference between the Astros' FIP and ERA. In addition, the Astros pitchers are ranked 7th worst in SIERA and 10th worst in tERA. Although the Astros' team pitching doesn't fare well on these two advanced metrics, the rankings are higher than the ERA.
Here is a comparison of rankings which suggests to me that the poor results by the Astros' pitchers cannot be totally blamed on the pitchers.
Astros' ML Team Ranking
Pitcher BABIP 3d Worst
Pitcher Groundball % 3d Highest
Line Drive % 30th (Lowest)
Team UZR 28th (3d Worst)
Team DRS 30th (Worst)
The Astros' pitching BABIP (Batting Average on Balls In Play) is among the highest in the majors. To varying extents, pitcher BABIP is viewed as random variation. Put another way, part of the differences in BABIP among pitchers is usually attributed to luck. However, team defense is a factor which can also affect a pitcher's BABIP. Run prevention is a joint function of defense and pitching, with the impact of both elements showing up in the pitcher's allowed hits and runs. And the Astros' defense was very bad, according to the advanced metrics, DRS (defensive runs saved) and UZR (ultimate zone rating). According to those systems, the Astros' defense has cost the Astros 3 to 7 wins compared to an average fielding team.
The Astros pitchers were among the best at inducing groundballs. Ordinarily, this is a good quality, because groundballs are less likely to produce extra base hits and HRs. Moreover, a groundball is likely to produce a much better result than a line drive. Simply put, a groundball usually is easier to field than a line drive. And the Astros' pitchers were good at avoiding too high a percentage of line drives. I suspect that the Astros pitchers' good line drive rate is due to their ability to keep so many batted balls on the ground. (To illustrate this point: the three lowest LD rates--Rays, Padres, and Astros--are among the four highest groundball rate teams.)
However, groundballs probably are susceptible to more BABIP random variation and defensive impact. Pitchers can induce a groundball, but the defense is responsible for converting it to an out. The advanced defensive metrics, which I showed above, are a better reflection of defensive impact than simply looking at errors and fielding percent, because the results consider the fielders' range. The Astros have pitchers who induce groundballs; that's a good thing. But it isn't an optimal combination with a poor defense. And the result of this combination likely is reflected in the high BABIP.
The implications are relevant to team construction as the Astros put their roster together in the future. First, this should affect the evaluation of pitchers who have taken the mound for the Astros this season. Some of them may not be as bad as it appears on the surface, because of weak defense behind them. Second, if the Astros plan to rely on groundball pitchers in the future, the defense must be upgraded dramatically. And, really, the defense needs to be upgraded regardless of the pitching type, because a bottom three defense won't win games.
Although it's true that groundballs produce more hits than flyballs, the avoidance of HRs, extra base hits, and line drives that accompany balls hit in the air make groundball pitchers an attractive option.
The comparison below (Baseball-Reference.com) for team pitching shows the NL average and Astros batting averages for various batted ball types.
NL Avg. Pitcher BA / Astros Pitcher BA
GB .238 / .253
FB .224 / .256
LD .719 / .739
Bunts .343 / .242
The Astros generally allowed a higher batting average on all types of batted balls except for bunts. It is possible that the Astros defense contributed to above average hits in both the outfield and infield. Undoubtedly, bad luck affected the results too. Maybe some of the pitchers allowed harder than average contact on some of the ball types, who knows. There is evidence among hitters that higher than average line drive batting averages tend to regress to league average. For that reason, it's possible that the Astros' pitchers suffered more than their share of bad luck on line drives. Fly ball were particularly costly to the Astros' pitchers, who had the 6th highest rate of HRs per fly ball (12.6%). (To some extent, a regression in the Astros' HR rate should be expected, since fly/HR rates tend to normalize.) Combined with the relatively high batting average on fly balls, groundballs appear to be the least damaging option for Astros pitchers this year.
Although the combination of poor defense and groundballs contributed to the Astros' high pitching BABIP, the best strategy is improving the defense rather than changing the groundball tendency. Even with the above average hit rate on groundballs, the Astros pitchers' batting average on the non-groundball part of BABIP was .371. The very high batting average on line drives is the major reason that line drive rate is highly correlated with BABIP--groundball and flyball rates have a low correlation with BABIP. The significance of line drives is shown by the non-groundball component of the Astros BABIP. Although the Astros pitchers did a good job of reducing the line drive percentage, they were still hurt by weak defense and too many HRs.