Determining the true talent level of hitters is no simple task. While some metrics tend to gain “statistical significance” over the course of a full season, many do not, and even for the ones that do, many hitters do not reach the requisite “statical significance threshold” from which we can start drawing conclusions. One such metric, BABIP, requires more than 650 plate appearances to reach a predictive or indicative level. Given the volatility of BABIP luck, AVG/OBP/SLG/OPS determination, a function of BABIP, becomes a fickle task.
Thanks to a few good men, however, talent spotting is a less exacting task than it once was.
A couple of years ago, Chris Dutton and Peter Bendix did some research on batted-ball data and created a metric called xBABIP (“expected BABIP”). xBABIP dispelled the myth that BABIP was primarily a function of “LD%+ .120.” Rather, as Dutton and Bendix found, BABIP was better explained as a function all batted-ball types and ratios with speed/power/strikeout considerations.
Last year, Derek Carty and Chris Dutton debuted the simple xBABIP calculator on THT. This tool has empowered users to determine a player’s xBABIP and compare it to their actual BABIP. Therefrom, one could forecast a hitter’s expected batting line, assuming all the input ratios were to remain constant. Over the course of 500+ PA, these ratios tend to be significant, though conclusions can still be drawn at the 300 PA threshold (we’d really only be waiting on IFFB% stabilization).
For all 270 hitters who accrued 300 or more plate appearances this season, I applied the xBABIP formula (by park) to determine each hitter’s expected batting lines. In short, what I have created is a spreadsheet of “what you can expect as a baseline for production in 2011, assuming all else remains constant.” In other words, this is how these hitters should have hit in 2010.
My methodology was tedious but simple: once you calculate xBABIP, you reverse engineer expected hits (xHits). Here, a judgment call is necessary—do you keep a hitter’s power ratios and assume they represent his true talent line or do you assume all hits are of a single type? I use the latter strategy, as I do not have a means by which to accurately calculate the proportion of singles/doubles/triples/home runs saved/gained by bad/good luck. Conceding that no matter how I apportion the hit types that it would be by guessing, I simply assume that all hits gained/subtracted would have been exclusively of the singles variety. It should be noted that as a result of this, the expected ISO forecast is not to be relied upon. This methodology is most effective at finding a hitter’s expected batting average and on-base percentage. Nonetheless, BABIP considerations have an impact on hitter’s slugging percentage, so I have accordingly included an xSLG forecast in my chart.
At this point, before I reveal the file and discuss my xBABIP adjustments further, it is essential that I note that the park factor constants utilized in calculating expected BABIP are from 2009. Since there was less than three months worth of park data available at the time the xBABIP tool was released (assuming the new parks were used at all in determining xBABIP park factors), the numbers on Yankees and Mets hitters may not be completely accurate. Take their numbers with a grain of salt. (I’ve italicized Yankees/Mets hitters in the spreadsheet for reference.)
It should also be noted that hitters with high HR/AB rates should be treated with caution. Relatively low Ball-In-Play (BIP) hitters are subject to more underlying numbers volatility.
Now that you understand how I calculated the numbers, let’s take a look at them. You can access the full xBABIP spreadsheet by clicking here. The password to manipulate the spreadsheet is “soto18.” Again, please note that these numbers assume persistence. Adjust mentally as you expected a hitter’s batted ball distro to change (more line drives and groundballs, more AVG, more flyballs (especially infield flies), less AVG). If a player’s numbers and xBABIP do not settle well with you, look at his three-year or career BABIP numbers. If there’s a large disparity, his 2010 xBABIP may be the byproduct of fluke rather than true talent.
Here you find my eternal nemesis of irrational hatred, Justin Morneau, topping the BABIP-luck chart. He’s really a mid-high .270′s AVG hitter in my mind, but xBABIP confirms that there is no reason to reasonably expect even a .300 AVG in 2011. Obviously Josh Hamilton will likely come down to earth some next season (though still hit for a plus-AVG, as he did in 2007 and 2008), along with Austin Jackson (unless he decides to taunt my calls for regression with more virginal sacrifices to the BABIP gods), Colby Rasmus, Nick Swisher and Jayson Werth (buyer beware?).
Amongst guys who should see their AVGs grow in 2011, we find Carlos Peña, who hit below .200 this season, topping the chart of “guys you can expect more from next year.” Also on this list are Aaron Hill (.273 AVG), my boy Matt LaPorta (.262 AVG), Jose Bautista (though that’s likely a byproduct of his super home run binge…I’d expect a .270 average tops), Zorilla (.278 AVG) and Mike Napoli (.272 AVG, though lots of squatting may depress this figure). Juan Rivera‘s on this list too, but that, to me and my gut, feels like a flukey outlier at his age.
Post your love/hate in the comments.