Hoskins, BABIP, and Sustainability
The topic of Rhys Hoksins and sustainability is a big one, because no one believes he will hit 80 home runs a year, but they do want to know how real he is. Today The Athletic Philly wrote about Hoskins and sustainability and this set of paragraphs caught me.
The rest of Hoskins’ success at the plate is no fluke.
The average major league player has a BABIP (batting average on balls in play) of around .300. Through his first 143 plate appearances, Hoskins’ BABIP is .257, suggesting he’s actually getting somewhat unlucky on the balls he doesn’t smash over the fence.The Athletic
BABIP might be one of the most misunderstood stats in baseball. One of the reasons for this is that it was at the center of a large pitching theory (DIPs). DIPs is the basis of FIP and it essentially says that a pitcher has no control over batted balls off of them. The major league average on balls in play is .300 and the assumption was that pitching BABIP trended towards .300 for pitchers over a large enough sample size. This also lead to a belief that batter BABIP actually trended to .300 over time as well. The reason for the trending on pitchers was that the large sample size of batters would average out. It turns out that pitchers have an influence on the type of contact off of them, and that different types of contact are hits at different rates.
Here is what we do know about BABIP:
- Infield fly balls are almost never hits
- Home runs don’t count as balls in play
- Line drives are hits at a high rate (.682), then grounders (.242), and then fly balls where IFFB are included (.130)
- These categories might still be too broad
- Minor league BABIP is higher than major league BABIP
- Other factors like speed contribute to BABIP
- BABIP is hitter dependent
Now there is an argument to be made that Hoskins has been unlucky on fly balls as he has a .000 BABIP on them. But this is where we get into the HR/FB rate for Hoskins. According to FG’s calculation it is 38.6% this year. If you account for infield fly balls he has actually hit a home run on 17 of 39 outfield fly balls this year. Right now the current leader in HR/FB rate is Giancarlo Stanto at 33.8% followed by Aaron Judge at 32.8%. We are still in small enough sample size territory that this is the difference of 3 home runs of Hoskins’ 17, so he would still be at a historical rate if it were not for some “luck”. This is due to the high number of fly balls Hoskins is hitting. Right now if he qualified he would be hitting fly balls at the 7th highest rate in baseball. What we have seen from other high fly ball rate hitters is a lower BABIP because the ball for the most part either doesn’t count in our calculation or is an out.
The other major predictor of future BABIP is past BABIP. So here is Hoskins the past two seasons as his home run rate has gone up:
Now part of this is extreme infield fly ball rates at those levels that have not manifest completely in the majors yet (we have not reached a stabilization point for all batted ball data).
The last part of all of this is how we classify luck. We use luck to describe the parts of the system that aren’t explain or that are noise. In theory Hoskins should have a high BABIP based on what he has done so far, but that doesn’t mean what he has done will stabilize or continue. We are still too early to say he will hit line drives at 26.4% pace, and while those are falling for hits at an above average rate now, we don’t know if that is prone for regression. We also know that he is a pull hitter and that various metrics are not going to trend to league average (like ground ball BABIP). One last thing since I mentioned it earlier briefly, stabilization is not the same predictability. We can say that say FB rate has stabilized, but that just means that we can reliably say that measures what he was during that time, it is not the point where we can say that is what he is going forward.