Library Update: BABIP
One of the most commonly cited sabermetric statistics for hitters and pitchers is Batting Average on Balls in Play (BABIP) and our Library entry on BABIP is newly renovated. Not only does the entry now contain the BABIP equation, which was previously absent (oops!), but it walks you through a more nuanced explanation of why we care about BABIP and how to use it responsibly.
We have a BABIP entry in the Offense and Pitching sections of the Library but the entries are the same, so clicking to either will do.
Click over to the Library for a full breakdown, but feel free to post questions about BABIP here. We also have a post going up in the Library blog today that illustrates the importance of BABIP and the concept of luck, so feel free to post questions there as well. You can also reach out to me on Twitter @NeilWeinberg44 or stop by our weekly Q&A chat Wednesdays at 3pm eastern to ask questions about advanced stats or the site’s features.
Neil Weinberg is the Site Educator at FanGraphs and can be found writing enthusiastically about the Detroit Tigers at New English D. Follow and interact with him on Twitter @NeilWeinberg44.
I never knew bunts weren’t counting in BABIP calculations. Is there a specific reason for that? It does seem odd that bunts would not be counted, but bunt hits (and bunt popouts) would be.
It’s a little sticky because you never know for sure, but if you’re sacrificing a runner you are intentionally trying to make an out and that out has likely been ordered from the bench. There’s no reason to factor this into either the batter or pitcher’s BABIP. Basically, if you successfully sacrifice, you’re just doing what you were told and it doesn’t show up in BABIP. If you screw up and don’t advance the runner or you find a way to get yourself on base, you either get docked or credited for that.
If you think about it in terms of the three possible bunt outcomes, that sort of makes sense.
1) SAC = both sides get what they want
2) OUT = pitcher wins
3) Reach = batter wins
That write up is fantastic. It boiled down what I’ve read in a hundred different threads and articles into one place.
I’ve always wondered about fouling out. I take it that it counts as BIP, though it’s not “fully” in play, in the sense that if the fielder misses or drops the ball, it does not mean the batter gets a hit. I assume it’s one of those between-the-cracks events that has to be classified one way or the other, and it seems better to consider it a BIP than a K.
The other gray area I’ve wondered about is balls that are hit high off the wall, e.g., the Green Monster. If the ball is well over the outfielder’s head, it’s not in play in basically the same way that a HR isn’t. So in that sense, it seems to me that it shouldn’t count as BIP. OTOH, once it bounces back to the field it is in play, and how many bases the batter gets is partly a matter of luck, literally, the way the ball bounces.
Also, what about reached on error? My understanding is that FG values that basically about the same as a hit, so shouldn’t the numerator be H – HR + ROE? Or is ROE considered a BIP that did not “fall” for a hit?
From your remarks about the maximum BABIP, I take it you don’t believe that Trout’s BABIP, which was just about .380 for his first two full years, is not sustainable? It is a little lower this year.
By the way, in the first paragraph of the BABIP entry, I think you mean “catcher interference”, not “catcher inference”, though that is certainly an interesting typo, subject to some relevant interpretations!
Thanks for catching the typo!
Fouling out is a ball in play. Your point is well-taken, but the idea behind BABIP is the influence of luck and defense, so while a foul ball dropping doesn’t equal a hit, it extends the at bat. In a relatively simple metric, we can’t do too much to adjust for the true value of a foul fly ball.
Hitting it off a high wall is interesting, because it is a guaranteed hit independent of the fielders, but it’s hard to really do much to control for when a ball is impossible to field and when it isn’t. This is again, something more suitable for a really advanced metric beyond BABIP.
ROE is a tough one, because the idea is that you’re already defining it as something that shouldn’t have happened. So you don’t really need to throw it into BABIP. Really, in a perfect world, we wouldn’t have errors at all and things would be hits or outs, but until we get there, we’re kind of left with this limbo.
Great site though there doesn’t seem to be a user forum to ask questions so I’m posting here: I noticed that (GB + FB + LD) is <= (AB + SF + K). The greatest discrepancy is with players that bunt a lot such Elvis Andrus or Michael Bourn. Are the missing batted balls failed bunts – either for sacrifice or for a hit? I'm using your data for building statistical models and the missing data points are troublesome. Thanks!
The right side of your equation doesn’t include sac bunts, so that would be one difference. The other thing is that you have strikeouts in the right side as well, which are not going to be counted in batted balls. You might nean to subtract Ks, not sure if that’s just a typo. Try adding sac bunts to the right side and see what happens
Yes that was a typo. I meant to subtract Ks. I’m compiling multi-season data from 2010-present. Ex. Bourn has 1282 GB, 588 FB, 545 LD or 2415 batted balls. His AB (3270) + SF (18) – K (774) or 2514 batted balls. There is a discrepancy of 99 missing batted balls. He has 30 SH, so that can’t account for the difference. I’m thinking the missing number is the bunts that weren’t sacrifices. Basically BU-SH Is that possible? In which case a BU is not a GB?
OK solved. The missing batted balls are bunts that were not sacrifices – either resulted in a hit or force out. I’m trying to create “true” balls in play batting averages which would mean eliminating all bunts from the equation (they should be tracked with their own BABIP since they are a special case). I don’t suppose the site tracks failed sacrifice bunts? I would imagine these drive down the ground ball BABIP very little for power hitters who never bunt and a lot for hitters with lots of SH – and presumably a fair number for failed sacrifices that got categorized as ground ball outs…
Sorry we don’t have that on the site. B-Ref might.