Society for American Baseball Research, Rogers Hornsby Chapter

At the May 2012 Hornsby Chapter meeting, the subject of BABIP came up and that led to a spirited discussion.  Dan Walsh then took it upon himself to do some research on the subject.  The article below is a consolidation of the emails that Dan sent to the group with his findings.

Jan Larson

—

PART I

Last night, the conversation went something like this: Monte: “So, you’re telling me that the top 5% pitchers have the same BABIP as the bottom 5% pitchers?”  Jim responded: “Yes”.  Monte: “I don’t believe it.”

Using the 2011 season, I looked at the numbers for all major league pitchers who qualified for the ERA title (96 pitchers), and based the top 5% and bottom 5% solely on ERA.  Here they are, with BABIP and ERA. The MLB averages in 2011 were: BABIP = .295    ERA= 3.94

Top 5%                   BABIP             ERA

Clayton Kershaw          .274              2.28

Roy Halladay             .305              2.35

Justin Verlander         .240              2.37

Cliff Lee                .293              2.40

Jered Weaver             .252              2.41

Bottom 5%

Brad Penny               .312              5.30

Fausto Carmona           .296              5.25

Brian Duensing           .334              5.23

A.J. Burnett             .296              5.15

Bronson Arroyo           .281              5.07

A quick look at the numbers appears to favor Monte. (Monte, you can stop reading at this point). However, the stats raise some interesting questions: How does Halladay record the 2nd lowest ERA with a BABIP above the league average? And Arroyo the 5th highest ERA with a below-average BABIP.
As I recall, Monte and Ira were sure that the quality of pitching does affect whether or not a batter gets a hit.  I certainly agree with them on this point.  All pitchers are not created equal, and their pitching skills can impact the placement of a batted ball.  Jim was equally certain that all pitchers have an identical BABIP.  Well, that just can’t be…perhaps, all pitchers have a historical norm; however, they didn’t all have a BABIP of .295 last year, or in any other year.  Regardless, that’s missing the point of what Jim was saying (not trying to put words in your mount here, Jim).  BABIP is a pitcher’s average on batted balls ending a plate appearance, excluding home runs.  It measures the hit rate after a ball is put in play.  After the ball is in the field of play, whether or not it falls for a hit is dependent on the defense and random chance (i.e., luck).  Even when Mariano Rivera saws a bat in two, thae batted ball still has a chance to be a hit.  If A-Rod, Jeter or Cano can’t make a play, then it isn’t Rivera’s fault.  Thus, Rivera has little ability to prevent a hit once the ball is in play.  That’s how I understand BABIP as it applies to pitchers.

How about Halladay and Arroyo? If you accept the premise that the pitcher does not have complete control over his hit rate, and you want do know how effective a pitcher is; then, you need to look at those things that the pitcher does control – specifically, walks, strikeouts, and home runs allowed. Among our 96 pitchers, Halladay had the 3rd lowest BB/9, 3rd lowest HR/9, and was 17th in K/9. Whereas Arroyo, was 17th in BB/9, 89th in SO/9, and #96! in HR/9. Conicidenatlly, Cincy had the lowest team BABIP in the NL last year; perhaps Arroyo relies on his defense to prevent runs because his pitching skills certainly are not doing are not doing the job.

Thanks to Jan for sending all the BABIP links.  For the original research go to Baseball Prospectus.  Attached is a spreadsheet file with all Pitchers and team data for 2011.   Enjoy!

PART II

Here’s the comparisons – the attached spreadsheet shows those 10 pitchers for the last 5 years, with their career BABIP.  Is there any consistency?  A couple of these guys show some consistency from year-to-year.  Weaver is interesting – BABIP has decreased each of the past 4 years.  However, overall this group of pitchers is up and down each year – which is one of the points that Voros McCracken noted in his original research.  You can’t predict BABIP based on the pitcher’s previous performance in the stat  – it doesn’t correlate from year to year.  McCracken also noted that the majority of pitchers who have pitched significant innings have career rates between .280 and .290.  That pretty much holds true for this group, as shown in their career BABIP. Maybe that’s what Jim was  getting at (putting words in his mouth agin..).   Based on his career BABIP, one would expect Verlander to eventually settle down into a 28-29% hit rate, rather than the ridiculous 24% rate – or maybe not – let’s see what it looks like at the end of this season.

 

Here’s an article from the Technology column of The Atlantic, courtesy of Norman, about a new type of ball glove: