To kick off part two of this series, we are gong to start with a mini lesson in small sample sizes. The stats in the first article were through last Tuesday’s games and the stats in this part of today’s article will be through Saturday’s. That’s four games more in the second set.
Below is a table showing the batting averages and BABIPs of all the players named in the first article.
Average BABIP Player Tues. Sat. [Diff] Tues. Sat. [Diff] Albert Pujols 0.320 0.356 0.036 0.279 0.314 0.035 Torii Hunter 0.319 0.314 0.005 0.300 0.292 0.008 Andre Ethier 0.312 0.308 0.004 0.317 0.329 0.012 Adam Dunn 0.328 0.299 0.029 0.333 0.320 0.013 Nelson Cruz 0.278 0.262 0.016 0.280 0.276 0.004 David Wright 0.282 0.284 0.002 0.404 0.400 0.004 Carl Crawford 0.274 0.290 0.016 0.354 0.367 0.013 Kevin Kouzmanoff 0.269 0.237 0.032 0.339 0.309 0.030 Jhonny Peralta 0.225 0.209 0.016 0.327 0.298 0.029 Russell Martin 0.225 0.224 0.001 0.308 0.297 0.011 J.J. Hardy 0.174 0.156 0.018 0.188 0.164 0.024 Average Difference = 0.016 Average Difference = 0.017
The thing to focus on is the average difference in change in batting average and BABIP over the four days. Both are about .015 or in layman’s terms, 15 points. Considering the small period of time over which this change happened, that is surprisingly large. Well, only surprising to those who beforehand did not have a good concept of how much rate stats fluctuate in the early going.
In case you are worried that the average difference in batting average and BABIP of these eleven batters does not reflect all batters, for those interested, the average change in batting average from Tuesday to Saturday for every batter was 0.017 and for BABIP it was .015. Close enough.
Despite the volatility of these rate stats, there is still validity in a higher or lower than expected batting average, or higher or lower BABIP. In today’s article I will detail several players whose BABIPs lead me to believe that soon enough they should start playing closer to what preseason expectations were.
A note before we begin: Players performing better than expected should not start playing worse than expected in the future; they should be expected to play at the level we expect them to play at. It sounds obvious enough when I say it like that, but it remains a concept many fail to grasp. For example, let’s say we expected Nick Swisher to hit .250 this year. Because he is batting over .300 now does not mean that the baseball gods are going to exact retribution on him and make him hit below .250 for the remainder of the season so he finishes with a .250 average.
Karma does not exist in baseball in terms of luck. A streak of good luck does not necessarily follow a streak of bad luck. For another (and probably clearer) example of this, read Dave Cameron’s explanation over at USS Mariner.
With that fallacy out of the way, here is the first set of players:
Low production, lower BABIP (greater chance level of production increases)
Note: Stats from here to the end of the article are through Sunday’s games
1) Lance Berkman — Owners of Berkman this year have to be disappointed with his current .172 batting average. So far he has gotten unlucky with balls in play as his BABIP sits at .153, so unless something is physically wrong with Berkman, he should start performing like himself sooner rather than later.
2) Jimmy Rollins — Similar to Berkman, Rollins is someone who has just been down on his luck so far this year with a .221 BABIP. Nothing you can do but wait.
3) Cody Ross — Ross is a player you know I like this year, and what I am about to say will only make you like him more. Like Berkman and Rollins, Ross has been unlucky; he is currently sporting a .254 BABIP. But unlike the other two, (whose Line Drive rates (LD%) are both 14 percent) Ross’ LD percentage is at league-average 19 percent.
This means that while Rollins’ and Berkman’s lack of luck on balls in play is somewhat deserved because of their low LD rates, (and we only expect them to do better because we expect their LD percentages to return to their career norms) Ross still has a low BABIP despite hitting a normal amount of line drives. Therefore there is even a greater chance that Ross rebounds.
Instead of repeating the same explanation for more players, I am going to list a few more batters who fit the same mold as Ross, meaning they are especially likely to rebound. One thing to kind in mind, however, is that LD%—like other rate stats—is volatile right now, so any high LD percentage today could become low in less than a week. Keep an eye out for dramatic shifts.
Player OPS BABIP LD% Jason Kendall 0.510 0.206 23 Kelly Johnson 0.724 0.224 17 Chris Young 0.594 0.237 20 Mike Aviles 0.539 0.262 24 Randy Winn 0.673 0.262 22 Grady Sizemore 0.804 0.268 21 Emmanuel Burriss 0.544 0.271 25
A wide range of players in this list, all of whom I expect to pick up the pace in the near future.
Now it is time to talk about the overachievers:
High production, high BABIP (greater chance level of production is reduced)
The hitters that fall into this category tend to be obvious (of course Youkilis won’t bat over .400 the whole season) and over-discussed, so I scoured the player universe for less obvious and lesser-discussed players.
1) Christian Guzman — Guzman’s sole fantasy value this year has come from his .373 batting average, the result of a .442 BABIP. The most impressive part: He’s doing it all with a 10 percent LD rate. Eventually something’s gotta give—either the BABIP will fall or that line drive rate will catch up. Most likely they both will make the shift towards each other, resulting in a lowered batting average for Guzman.
2) Jordan Schafer — This Braves outfielder has caught the eye of some in deep mixed and NL-only leagues, and I’ll warn those people to be careful. Schafer is hanging onto a .260 average, which is precariously supported by a .400 BABIP. He is not going to hit leadoff again for a while, so it may be time to cut or bench the young outfielder.
As a final disclaimer, I know I used line drive percentage many times in this article as a predictor of BABIP. While LD percentage is a significant determiner of BABIP, we know from the work of Bendix and Dutton that many other factors are in play. For my purposes, however, a simple check of the line drive percentage sufficed.