Validating the New xBABIP Equation With the Decliners

Let’s now follow up yesterday’s 2017 BABIP decliners list by looking back at who the new xBABIP would have convinced us to avoid heading into the 2016 season. Like I did when validating xBABIP using the surgers, I’ll compare how the would-have-been 2016 list performed versus their 2015 xBABIP and 2016 Steamer projections.

I included a list of all hitters that outperformed their 2015 xBABIP by at least 0.30 points (.330 BABIP vs .300 xBABIP), which totaled 55 hitters.

Once again, xBABIP proved clairvoyant, as its .298 2015 group mark was quite close to the .302 2016 actual mark. Of the 55 names on the above list, 45 of them experienced some level of BABIP regression. Of course, since the group average was .341, a monkey could have forecasted that these hitters would decline, and Steamer projections agreed that regression should be projected, even without knowing the fancy components of my xBABIP equation. But xBABIP was closer, albeit by just a point, and, more importantly, xBABIP isn’t even meant to be predictive, but strictly backward-looking. So it’s not entirely fair to pit 2015 xBABIP against 2016 Steamer, since at the very least, xBABIP is missing an aging component, which certainly plays a role.

Kris Bryant appears second on the list and the prospect of serious BABIP regression was one reason I thought he was overvalued heading into 2016. He has now outperformed his xBABIP two years running, as his 2016 BABIP of .332 was also well above his .305 xBABIP mark. He’s even facing a shift now, which makes it all the more baffling that he continues to be the exception. I wonder what he may be doing that’s not captured in xBABIP, or if two years is simply not a large enough sample yet. But BABIP doesn’t stabilize until 820 balls in play, and Bryant is at 752, so he’s not quite there yet. We still have to consider the possibility of regression closer to league average with Bryant, and his price certainly doesn’t account for such risk.

Most of you figured that Miguel Sano wasn’t going to sustain a .396 BABIP, but perhaps you could have convinced yourself that he hits the ball so hard, maybe he could keep it above .350 or something. Well, even a drop to an above average .329 mark was enough to fully illustrate the downside of a guy who strikes out 36% of the time. Love the power, but gotta hope he starts making more contact, or becomes a little more aggressive and swings more often.

Ahhh Dee Gordon. When you’re value is primarily tied to steals, and you’re getting by with an inflated BABIP, there’s a huge amount of risk. Because when that BABIP drops, your opportunities to steal declines along with it. And that’s precisely what happened to him, as he went from attempting a steal once every 8.4 plate appearances in 2015 to once every 9.3 plate appearances in 2016. That’s not a huge difference, but his BABIP falling back to Earth also had the effect of causing his batting average to tumble from .333 to .268.

Though Addison Russell was a solid fantasy contributor, one of the reasons I was down on him in 2016 was the expectation that his BABIP would collapse. It did, but he cut his strikeout rate, enjoyed a power spike, and hit with tons of runners on base en route to 95 RBI. That was enough to offset the BABIP regression.