I was fiddling with The Hardball Times (THT) stats page, and I noticed that quite a few Mets players had low batting averages on balls in play (BABIP), despite relatively high line drive rates (LD%).

Conventional wisdom says that BABIP and LD% are generally linearly correlated.  An oft-repeated rule of thumb is that BABIP = LD%+0.12, and MLB’s mean BABIP is usually around .300.  Keeping in mind that that’s a very rough estimation, it seemed that numerous Mets, both starters and role players had BABIPs lower than would be expected.

 Taking THT’s team batting stats and plugging them into a spreadsheet lead me to a few interesting nuggets. 

 1) The AL and NL had BABIPs of .302 and .298, respectively.  Assuming approximately equivalent AB’s, that puts MLB’s mean BABIP right on the rule of thumb expectation.

2) Despite the above BABIPs, the NL as a whole actually had a slightly higher LD% than the AL, 21% to 20%.  Thus, the AL BABIP was only 0.102 above its LD% and the NL was a surprisingly low 0.088 above.  So, it wasn’t just the Mets that were coming in low on the BABIP/LD% rule of thumb, it was baseball in general!  But wait, that’s not the end of the story.

 3) While baseball was much closer to having a BABIP that is 0.095 above it’s LD%, than than the 0.12 rule of thumb, the Mets still threw under that bus.  The Mets, as a team actually has the *smallest* differential between BABIP and LD% in all of baseball.   They managed this by having the highest LD% rate in all of baseball, while being in the bottom third of BABIPs in baseball.  I’m not sure what to make of that.  On the one hand, I curse the Mets rotten luck.  I mean, they were killing the ball, but it didn’t matter, the ball just didn’t find holes (they always seem to find a way into Willie Harris glove–I irrationally hate that guy).  On the other hand that leads to some optimism that if some regression to the mean happens, the Mets could be better offensively than last year. 

4) Unfortunately, the team that had the *second* lowest BABIP/LD% differential were the Philladelphia Phillies.  So even if some regression to the mean occurs, it’s just about as likely to help the Phils as well.

 On a side note, I do find it weird that two of the highest scoring teams in the NL both at the rock bottom of the BABIP/LD% differential. 

5) The two NL teams with the best BABIP/LD% differential couldn’t be more different.  The Cubs lead the league in scoring and lead the league in wins.  The San Francisco Giants were a moribund offensive team, that would have been last in the league in scoring if not for offensively challenged San Diego.

Bottom line, what does all this mean?  I suspect, not much of anything other than that baseball is a fickle game.  Which is to say, it’s interesting to look at the data, but in the end BABIP has a lot of random noise thrown into it, and because of the weakness of the correlation with LD% even a full season of data won’t necessarily overcome the noise.