I was recently asked about BABIP (batting average on balls in play) and how to correctly use it as a predictive tool in fantasy baseball. The particular player who came up in this conversation was J.D. Martinez, who is currently sitting with a league-high BABIP of .443.
Since the league average BABIP for hitters tends to settle at around .300, isn't it safe to presume that Martinez will very likely see a ton of regression over the rest of the season?
In short, that's not how this works.
It is true that the overall league average BABIP tends to settle within a few points of .300 year in and year out. It's also true that there are typically around 25 players each season to have a .350 BABIP or higher (minimum 350 PA), so it's not out of the realm of possibility that some player could finish a season in the neighborhood of .400.
However, the truth is that while BABIP can be used as a predictive tool, it has to be done on a personalized level. For the most part, if a player's strikeout and walk rates remain relatively constant, then the difference between that player's BABIP and batting average should also remain somewhat constant. That's when you can use BABIP to predict future movement in a player's batting average, by way of an expected adjustment due to "luck."
For example, Joey Votto's career rates are 17.6 percent strikeouts and 16.2 percent walks. His career BABIP is .353 and his career BA is .312, a difference of 41 points. If you go season by season, though his BABIP fluctuated year to year, this BABIP-BA gap remained relatively consistent. It grew a bit in seasons when his walk rate was higher, and shrunk significantly in 2017, when his strikeout rate was a career-low 11.7 percent, but overall most players will see the gap between those two stats remain relatively constant over time.
So, let's look at Martinez. For his career he has a 25.3 percent strikeout rate and a 7.9 percent walk rate. Those are pretty much in line with his current 2018 rates of 27.0 percent and 7.4 percent. For his career, his BABIP is .345 and his BA is .288, a gap of 57 points. As such, we should expect a similar gap for 2018, which would put his expected BA, given a .443 BABIP, to be around .386 rather than his current .351.
Therefore, I would argue that Martinez might actually have some staying power in his numbers, rather than regression in his future. Even once his BABIP drops to .400, which it almost certainly will due to personal regression, if he keeps his rates constant, I'd expect the batting average to remain in the .340 neighborhood.
BABIP is a tool to be used on a case-by-case basis. In the case of J.D. Martinez, I have full confidence that it points to continued success going forward.