Fun w/ Pythagoras #5: Fielding | Disciples of Uecker

Disciples of Uecker

We'd like to go to the Playoffs, that would be cool.

Of all the proprietary stats I would love to view, the Milwaukee Brewers’ shifting plans and fielding research tops the list. The Brewers have had one of the most aggressive shifting clubs in the MLB, showcasing the fact that for all the praise given to other organizations’ use of stats and new baseball models, the Brewers still have a radical, forward-thinking front office of their own. Anyway, one of the reasons I’d love to see the Brewers’ shifting research is due to an extreme issue that their club faces: Miller Park.

Miller Park is a lovely place to watch a game, affording great sightlines to fans and an exceptional chance to see the longball. Even when the park played closer to an average run environment, it had the propensity to encourage all comers to hit more homers in Milwaukee than they would elsewhere. Aside from the Park’s “factor,” which determines the park’s influence on the run environment by neutralizing club characteristics and contrasting the park with the National League as a whole, “splits” between home/road performances for the Brewers and their opponents showcase a seismic change in home runs —and, surprisingly, infield errors. Not only are clubs more likely to hit home runs in Miller Park than other parks, they’re now more likely to commit infield errors in the Menominee Valley.

One of the best sources for multi-year park splits is Bill James’s Handbook series, which pretty much takes every possible stat one could want and kindly publishes it in a paperbound format to save one’s eyes from the glare of a computer screen. The following splits come from the 2010 and 2013 Handbook editions, respectively, published by Acta Sports in Chicago (yes, you should buy them on their own merits, and not simply because Bill James is one of the greatest American skeptics, rationalists, and populists ever to have lived).

YEARS Miller Park HR split Miller Park E-infield split Miller Park R split Park Factor
2007-2009 104 87 95 ~98
2010-2012 129 101 107 ~103

Anyway, comparing 2010-2012 to 2007-2009 at Miller Park, the run environment increased by approximately 10%. This means that between 2007-2009, clubs (Brewers and their opponents) scored slightly fewer runs at Miller Park than they would elsewhere, but between 2010-2012 clubs scored notably more runs at Miller Park than they would elsewhere. If your answer is, “well, this can be explained by the Brewers’ awful pitching staff in 2010,” this split stands for both the Brewers’ arms and opposing arms (and, the Brewers’ hurlers were above average in 2011 and 2012, so poor pitching wouldn’t explain the split in those years). Interestingly enough, the infield error split increased by 16% between 2007-2009 and 2010-2012, while the home run split increased 24%! While Miller Park’s overall factor was closer to average than some other parks (such as Coors, PetCo, Citi, or New Comiskey, among others), the elements underneath the surface, the events actually occurring in games, mercilessly swirled toward more home runs and more infield errors.

I first discovered the Brewers’ absurd fielding split toward the end of 2012, and the issue haunted my mind to no end: how could the Brewers address their extreme fielding circumstances at Miller Park? It’s worth noting that the extreme park factors at Miller Park mean that fewer balls are batted into play at Miller Park than at other parks the Brewers frequent, which means that their Miller Park runs allowed total is skewed more by the balance between HR, K, and BB than by other factors. This holds true for Brewers pitchers and fielders in 2013:

2013 Batted Balls in Play FIPRatio FIPConstant Runs Average Reached on Error BABIP
NL Home .691 0.51 3.37 3.88 .009 .295
MIL Home .696 1.18 3.24 4.42 .010 .300
NL Road .698 0.77 3.54 4.31 .009 .296
MIL Road .712 0.84 3.72 4.56 .014 .283

There are a lot of strange dynamics in these splits:

(1) Batted Balls in Play (BIP): As noted, the Brewers’ pitchers allow notably fewer batted balls in play at home than on the road, which skews their defensive stats, too.

(2) Errors vs. Average on BIP: One might expect a lower batting average on balls in play if the club faces fewer batted balls in play, but in fact, that lower rate of balls in play corresponds to a lower rate in errors, rather than fewer hits on balls in play.

On the other hand, the Brewers’ pitching and fielding is terrible on the road, where more batted balls in play are allowed, more errors are allowed, and therefore, the BABIP is lower. In fact, the Brewers’ rate of errors on the road is almost absurdly higher than at home: neutralizing plate appearances, the Brewers’ pitchers have allowed approximately 27 more batted balls in play on the road than at Miller Park, and approximately 5 more errors.

Overall, a running thread at Disciples of Uecker is that the Brewers’ fielders have not helped the Brewers’ pitchers, and I think that this is true in one significant way: when we analyze how the Brewers’ fielders are not helping their pitchers, it becomes remarkably clear that the road games are much more menacing than home games.

First, the general picture, in terms of FIP. As of this morning, here is the 2013 NL FIP:

NL runs average NL FIPRatio NL FIPConstant Defensive Efficiency
4.09 0.64 3.45 .694

You might be pleasantly surprised to know that, overall, the Brewers’ FIP picture blames the pitchers much more than the fielders:

MIL Runs Average MIL FIPRatio MIL FIPConstant MIL Defensive Efficiency
4.48 1.03 3.45 .695

In this case, the Brewers are approximately 26 runs below average (against the park-adjusted NL runs average), and their FIP suggests that the pitchers can mostly be blamed for that. However, split between home and road, a curious case emerges: when we compare the Brewers’ home FIPRatio (their ratio of K, BB, and HR) to the NL FIPRatio, we might suggest that the Brewers are at least 37 runs worse than expected (their defense, on the other hand, is at least a handful of runs better than average at home); on the road, however, the Brewers pitchers’ ratio of K, BB, and HR is only around 3 runs worse than the league norm for road teams; their fielders, however, are approximately 8 runs worse than the league on the road.

When we consider the weight of the home performances compared to the road performances (the Brewers have pitched approximately 56% of their innings at home), these runs above and below average almost perfectly account for the Brewers’ total of 26 park-adjusted runs below average. But that overall total masks a unique fact hidden by the club’s overall FIP: while the fielders are the major reason the Brewers are below average on the road, the pitchers’ balance between K, BB, and HR is the major reason the Brewers are below average at home.

There are a lot of solid, pressing questions about infield defense raised here at DofU, and I think that fan perception that the Brewers are not a great fielding club is generally accurate. Even though the club’s overall FIP and defensive efficiency suggests that the defense is fine and the pitchers are to blame for the Brewers’ problems, I’d like to ask:

(1) Do the Brewers have a good defense at home, or are elements of the park (such as the lack of Batted Balls in Play) unduly influencing their performance?
(2) Do the Brewers have a bad defense on the road, or are elements of road parks (such as the excess Batted Balls in Play) unduly influencing their performance?
(3) To what extent are defensive shifts improving or causing these problems?
(4) Is there a way to dampen the Brewers pitchers’ extreme FIP at home?

Baseball Reference. Sports Reference, LLC., 2000-2013.
Disciples of Uecker where cited.

Share Our Posts

Share this post through social bookmarks.

  • Delicious
  • Digg
  • Newsvine
  • RSS
  • StumbleUpon
  • Technorati