Weekend Report: Beyond the WHIP

on 08/30/2002

        K   BB	HR  IP    H   GB/FB NL rank
Rusch   109 56  24  166   185 1.65  9
Sheets  136 64  16  174.1 201 1.60  11
Ruben   68  93  28  139   159 0.60  50/last

I'm doing something a bit different this week. First - opening the article with a table is unconventional. Second - I'm going to use all of this statistical know-how to actually do some analysis. One of the more perplexing things on the Brewers this year has been how badly Ben Sheets has been hit. As the above table shows, Ben has been hit quite a bit more often than Rusch despite a superior strike out rate and a vastly better HR rate. The second most perplexing thing is where the roughly 40 strikes from Rusch went, but I'll focus on Ben today. Now, some might comment about how it's all luck since I'm the one touting the notion that the pitcher has little affect on ball in play. This, of course, fails to do proper justice to the issue. Still, until late last week I was lacking in an explanation. Then Mike Emeigh at Baseball Primer wrote this article http://www.baseballprimer.com/articles/emeigh_2002-08-21_0.shtml on whether or not pitchers can influence where a batted ball goes. The point of the article is that some pitchers may have some influence on what portion of the field a ball is hit to, but most of the effects on location reside with the hitter.

The table below shows how often pitches are pulled by hitters vs. pitchers broken down by Fly Balls (FB) and Ground Balls (GB), as well as the handedness of the pitcher and hitter. The trends are fairly clear: GBs are pulled a lot more than FBs, and RHBs pull a lot more than lefties.

GB              FB
Pull% LHP  RHP  Pull% LHP  RHP
LHB   67.3 68.4 LHB   41.1 42.4
RHB   71.8 71.9 RHB   47   46.4

The next table shows a breakdown of ABs against the batter types that the pitchers have faced. Notable here is that opposing managers have been able to alter their line-ups enough so that Rusch has faced right handed hitters in a better than 3-1 ratio, while Ben and Ruben have faced a nearly equal number of righties and lefties. If I had access to play-by-play data the following would be unnecessary, but using all of the available numbers I can calculate roughly the distribution of where the ball in play from these pitchers have gone. First, based on the GB/FB ratio we know that Ben and Rusch (a brief digression - Rusch is an excellent ground ball pitcher?) have both allowed about 62% of balls in play on the ground. Ruben, on the other hand, has been the most extreme fly ball pitcher in the NL by allowing over 62% of his balls in the air. For the purposes of this calculation the GB/FB ratio is assumed constant over batter handedness.

          ABs vs LH  RH	 1st/2nd  SS/3rd  LF/CF	  RF/CF
Rusch     154        493 126      207     103     101
Sheets    329        358 166      176     119     100
Ruben     247        305 80	      95      145     139

So what do all of those numbers say? Well, Ruben should probably beg to get a 5th OF to play CF who is awesome defensively. As I suspected, Glendon does a better job of utilizing the best defender on the club by getting a lot of balls hit to Jose. By contrast, Sheets's ground balls are distributed more evenly between the right and left sides. Stopping at this point, I'd be pretty comfortable saying that most of the difference in hit rate between Glendon and Ben is due to Ben's larger dependence on EY. The natural conclusion is that Ben would benefit substantially from a better defender at second when he's on the mound.

Carrying this observation forward requires making some predictions. Above and beyond a normal platoon difference we would expect that Ben would be more vulnerable to lefties because they account for the bulk of the ground balls hit to 2nd when he's pitching, whereas righties strongly favor Jose. Predictions based on the OF defense are less clear because everyone is pretty close in their distribution of fly balls and we don't know the precise distribution among the various OFs. The charts below provide some support for these ideas. Ben's batting average behaves as expected, while Rusch provides no clear data because his platoon advantage against LH works to cancel the effects of LH hitting more balls to EY. Based on the batting average differences, it would appear that Rusch does a slightly better job in preventing HRs against righties than he does lefties, but this is probably a result of only facing the cream of the lefty crop. Ben shows a huge OPS gap, most of which is explained by the batting average difference. Ruben shows a sizeable BA gap, but a much smaller than expected OPS gap. I interpret this to be a result of the fact that the primary benefit of being a GB pitcher is not batting average, but power related. By just giving up tons of fly balls, everyone does the same thing to Ruben - take him deep, though he does manage a small platoon benefit.

vs LH    BA   OPS
Rusch   .247 .755
Sheets  .325 .906
Ruben   .271 .907
vs RH    BA	  OPS
Rusch   .298 .796
Sheets  .263 .692
Ruben   .302 .872

It's time for some conclusions. I haven't investigated this, but based on the ranks of the pitchers I'm pretty sure that having such a wide difference in GB/FB ratio between starters is unusual. In another recent Baseball Primer article http://www.baseballprimer.com/articles/gerrard_thock_2002-08-26_0.shtml the author argues that for Derrick Lowe to be a continued success he needs to have a great set of IF behind him. He bases this on the fact that Lowe is the most extreme ground ball pitcher in baseball, with 80% of his balls being on the ground. We aren't quite to that point, but there is a large gap between Ben and Glendon and Ruben. And both sets are towards the extreme of both categories. It would seem that tailoring players to fit the starter may be in order. When Ruben takes the mound it's probably best to throw the most offensive IF out there you can put together, even if it means having a second base type player at SS (simply because it won't matter that much). On the other hand, using OF corners of Hammonds and Jenkins plus a defensive whiz like Shinjo would probably help Ruben a great deal. Similarly, Ben would be well served by spot starting Ronnie at second during his starts.

Player of the Week

Taking it all in, this week seemed like a good week to honor a pitcher. Both Jamey Wright and Ben Sheets had excellent starts. By making himself tradeable, Jamey gets bonus points. In the end, Ben was dominant in a "big game" environment - matched up head to head with Mark Prior - so he gets the nod as player of the week.