Batted Ball Location and BABIP
In recent months, Eno Sarris has talked a lot about going to the opposite field and how a hitter who sprays the ball to all fields may have a higher BABIP than those who primarily pull the ball. Furthermore, last summer Jeff Zimmerman updated an older BABIP formula and shared with us an xBABIP spreadsheet. Simply copy and paste a couple of lines of numbers from a player’s page and the calculator spits out an expected BABIP. The formula incorporates a hitter’s power, speed and batted ball distribution and does a darn good job of it. But with Joey Votto crediting his ability to go to all fields as a major factor behind his always near league-leading BABIP marks, I felt that it was time to start doing the research to determine if he was indeed on to something.
I looked at the cumulative totals for all hitters from 2002-2013 that recorded at least 100 plate appearances with all three of the following batted ball types: pull, center and opposite. That of course means pulled balls, those hit up the middle and then balls hit the opposite way. I was left with 841 players in my sample and the correlations with BABIP are as follows:
Pull: -0.28
Center: 0.26
Opposite: 0.17
Now in the more visually attractive graphical format:
Though the correlations aren’t very strong, it is clear that the location of the batted ball plays a role. From the charts, we find that batters with higher pull percentages generally have lower BABIP marks. Hitting the ball up the middle appears to be the best, while going the opposite way also positively impacts a hitter’s BABIP.
These correlations make it very clear that there is a lot more to determining a batter’s true talent BABIP than just looking at where on the field balls are hit to. This point is rather obvious given the current incarnation of the xBABIP formula. But the correlations are still meaningful and suggest that perhaps the location of a hitter’s batted balls should be incorporated into the formula as well. As such, we are going to be working on improving the xBABIP formula and spreadsheet, as this could be one of the few remaining missing links.
Mike Podhorzer is the founder of ProjectingX IQ, an advanced fantasy baseball analytics platform that transforms projection data and in-season performance signals into actionable intelligence. He is the 2015 Fantasy Sports Writers Association Baseball Writer of the Year and three-time Tout Wars champion. He is the author of the eBook Projecting X 2.0: How to Forecast Baseball Player Performance, which teaches you how to project players yourself. Follow Mike on X@MikePodhorzer and contact him via email.
Cool stuff Mike. I’m thinking that a good portion of the correlation would be driven by teams that aggressively shift against dead pull hitters, shrinking the zones where they can deposit a hit on balls in play.
Would the correlation show up strongly if you regressed BABIP to a control variable that was set to +1 for pull, 0 to center, and -1 to opposite?
This is exactly why Votto posited the theory. He said extreme pull hitters are easy to defend. You know where they are going to hit the ball.
It seems as if you are using the batter tendencies instead of the actual hits vs outs for the different locations, so you are doing more of a test for what kind of hitters hit the ball where than the success rates of different batted ball locations. Since pull hitters are much more likely to be slower runners, it appears you are in effect just confirming that bias, along with the tendency for teams to play more aggressive shifts for pull hitters.
There is an interesting possibility for this to have an impact with your formula, but you may be approaching it from the wrong angle. Another area for study could be to use the players’ raw numbers for babip for each of the field sectors and regress them for each player while accounting for a regression of the player’s running speed and batted ball speed, which could give a different picture, and a possibly more detailed view of the players.
Pretty sure the true cause of the variation in BABIPs by direction hit is caused by batted ball types. Contacts that are pulled have a GB/FB ratio of about 2.5, while central have a GB/FB ratio of about 1.1 and opposite field a GB/FB ratio of about 0.4. Since there is a negative correlation between BABIP and GB/FB ratio, it would make sense therefore that Pull hitters have a lower (or negative) correlation with BABIP. Really what I’m saying is the lower babip is a symptom of the high ground ball rate that is an effect of pulling the ball more often.
Agreed: http://www.fangraphs.com/blogs/an-unsolicited-follow-up-study-of-pull/
I think it may be more helpful to think of variation in BABIP as having multiple inputs rather than one “true” cause. No doubt batted ball type is an input and perhaps the most important one but that doesn’t mean other factors aren’t also at play.
As evidence of that point, if BABIP were simply a matter of batted ball type than we’d expect balls hit the opposite way to have a higher BABIP than those hit up the middle, when in fact, if we are to believe Mike’s research, the opposite is true.
One explanation might be that there is a substantially higher BABIP for groundballs hit up the middle than those hit the opposite way.
In any case, thanks to Mike for the research.
Yup, this is exactly what I stumbled upon and really made me scratch my head. According to the correlations, a higher oppo % leads to a higher BABIP, yet BABIP is lowest on opposite field balls! Weird stuff.
I also think opp way % is descriptive of a complete hitter more so than a power hitting heavy pull hitter which could partially be the reason the correlations trend as they do. Still pretty cool work
What’s the best place to get batted ball location data? BaseballHeatMaps?
In fact these days I have an idea for extensions to xBABIP in de-lucking hitter stats. I might try to post it to Community Research, if I have time.
It’s on the splits pages of each hitter, though Baseball Heat Maps I think has pretty maps also. Make the time and post to the Community!
Mike, I did! 😀
Does it matter what side of the plate they are hitting from? It seems left handed hitters are more pull hitters which commands a shift of the defense.
Not sure if that’s true and with no handedness filter on the leader, I can’t check.