Before I continue our series on ways to improve the 2011 Mets, I’m going to set out how I like to evaluate hitters. Note that I said hitters–not fielders. Because, while I’m a “saber guy” I don’t have any confidence in any advanced metric used to measure fielding. The reason why is that depending on the stat you use, be it UZR or range factor, or even the dreaded fielding percentage, you get wildly different results. For whatever reason, in spite of the variability, many of the advanced stats that evaluate players now incorporate fielding in their formulas. I think that this results in errors. So, I rely for overall evaluation on True Average, or TAv. This stat, developed by Baseball Prospectus, measures the value of a player’s offensive contributions, both at the plate and on the bases, with adjustments for the player’s league, home park and the quality of pitching he’s faced. Conveniently, it’s scaled to batting average, so the league average TAv is always going to be .260 and the best hitters in the league will have TAv’s significantly above .300 (Josh Hamilton was at .350). What’s great about TAv is that it takes everything we care about in a hitter–power, speed, ability to get on base, and gives you one number to evaluate a player. Because it adjusts for ballpark and league, we can make a real apples to apples comparison between say, Jose Reyes and Derek Jeter. TAv, however, reflects actual performance. It doesn’t completely tell us who under- or over-performed. Or, to put it another way, TAv is not a perfect tool for telling us who might be that proverbial diamond-in-the rough for next season. To do that, as we did for pitchers, we need to look at BABIP, or a player’s batting average on balls in play. We can also look at their mix of batted balls–on average, line drives are more likely to be hits than pop-ups and so on. So, to give an example, take Aaron Hill of the Blue Jays’ numbers for 2009 and 2010.
Aaron Hill had a great 2009, but was not very good in 2010. That’s true whether you like sabermetrics or batting average. And it’s important to understand that nothing I’m about to say changes that fact. But (thank goodness) 2010 is in the past. What we care about is figuring out what Aaron Hill is more likely to do in 2011. Here, the BABIP strongly suggests that Hill is the victim of bad luck. His .196 BABIP was the lowest in the majors in 2010, and even his .288 BABIP in 2009 was slightly below average. But we can dig even deeper. Here’s a breakdown between the types of balls Hill was hitting in 2009/2010 (fly balls, ground balls, line drives & pop ups):
That’s a big drop-off of line drives and a lot more fly balls and pop ups, the latter two of which, on average, are less likely to be hits than liners and grounders respectively. In fact, for 2010, Hill had the 4th lowest line drive percentage in the big leagues. The three guys “ahead” of him, Julio Borbon, Pedro Feliz and Mark Reynolds, didn’t exactly tear it up last season either. While Borbon and Feliz both hit the ball on the ground most of the time, leading to higher BABIP’s, Reynolds hit even more balls in the air than Hill, with a fly ball rate of 40% and a pop up rate of 14.7% and had a BABIP of .257. So, what this tells us is that Hill was unlucky, but not that unlucky. Let’s say his BABIP was the same .257 Reynolds had (a reasonable supposition given their batted ball profiles). That would give Hill 6 more hits on the season (he put 447 balls in play and 108 hits for a BABIP of .198, a BABIP of .257 would mean 114 hits). His batting average would be only .217. Note that even though Hill had a batting average of .198, his .TAv was a higher (but still not high) .237. Why? Because, again, TAv properly weights the various components that go into a player’s performance. Clearly, a .198 hitter with 26 home runs is more valuable than one with the same average who hit 5 homers. Hill also supplemented his batting average with 41 walks–not a high number for someone with 580 plate appearances–but still enough for an On Base Percentage nearly 90 points above his batting average. So, that’s TAv. I’ll be working in various other stats from time to time, but TAv will be at the core of how I evaluate hitters.