Buster Posey and Rookie of the Year, Part II
A little over a week ago, I posted my current Rookie of the Year ratings based on my rendition of Wins Above Replacement- accounting for the player’s batting runs, defense and baserunning runs above or below the league average, while throwing in a positional adjustment to account for positional difficulty and a replacement level to account for playing time. This is one of the reasons why Jason Heyward rated ahead of Posey in my rankings, as he had a bonus of +15 runs and Posey a +10. The difference of five runs narrows the gap between the two from 20 runs above replacement to 15, but we’re still not accounting for the additional offensive and defensive performance from Posey that would undoubtedly narrow the gap even more.
Extrapolating Posey’s batting line- which I have as +18 runs in 360 plate appearances- would give us +27 runs in 550 plate appearances (135 games; considered to be right around a full season for an everyday catcher) is inadvisable. This is due to random variation driven by being “lucky” or “unlucky,” or simply hitting a hot or cold streak that could sway the player’s perceived performance one way or the other in a small sample. In order to get a better estimate of the player’s true talent production, one must regress to the mean. Ideally, we’d regress Posey not to the Major League average, but to his specific population. So if we project Posey’s performance over 550 PA, and regress it to all catchers, gives us the following batting line:
I know this can be a little hard to read, so just click on the image to see the enlarged line. Posey’s projected overall line has him at .312/.362/.483 with 29 doubles and 17 home runs; +19 runs above an average hitter. For any hitter, that’s a darn good performance- for a catcher, it’s outstanding. I have some qualms with regressing to the overall population mean, however, as it’s riddled with backup catchers and the like. Catchers as a whole are hitting .254/.327/.390, and we all know that Posey’s talent level is better than that- and the drop from +.05 runs above average per plate appearance to +.005 seems a bit harsh. We can re-do the regression by excluding backup and “part-time” players, but…personally, I think it would be a little bit more fun (and perhaps more accurate) if we regressed Posey’s line a bit differently- weighting half of the regression on all Major League catchers and the other half on Posey’s closest comparable (based on scouts’ comparisons, at least), Joe Mauer. I feel a tad bit uncomfortable regressing Posey completely with Mauer, but it’s blatantly obvious that using the population mean isn’t sufficient. This way we give Posey more credit for having better tools than most catchers, while still remaining somewhat conservative. This gives us a final full season projection of…
That feels about right to me. A .319/.370/.493 line is a step down from his current .325/.369/.506 line, obviously, but we’re assuming there will be a bit of regression. +22 runs, in combination with a replacement level of +18 runs, gives us +40 runs above replacement. Posey played 248 innings at first base for -2.2 runs, but his estimated 952 innings at catcher gives him a bonus of +8.3 runs for a net positional adjustment of +6.1 runs. Add in another +4 runs for defense, -3 for baserunning, +0 for situational hitting, and we have a grand total of 48 RAR, which translates into approximately 5.1 Wins Above Replacement. If you’ll remember the aforementioned article, Heyward was also listed at 5.1 WAR. That would make them roughly equals, assuming that Heyward ended his season on that day. Then again, Heyward’s hit .375/.464/.375 over the past week, which would undoubtedly increase his batting runs by a few, and his replacement level by about 3 runs. In other words, I can’t imagine a scenario in which Posey would beat Heyward by my ratings, unless Heyward’s production were to fall off a cliff or Posey goes on an absolute tear. But, their rate of production is more similar than their straight WAR would suggest.
Of course, the standard caveats apply- a projection is nothing more than that- a projection; an approximation of true talent. There’s also a margin of error in our estimations of their components (i.e. hitting, defense, baserunning, etc.) and the like. But I think this makes for an interesting discussion point- would Posey make or break the projection? With the knowledge that there’s a high probability that he’ll regress to the mean given more plate appearances, does a .319/.370/.493 line seem reasonable? Or does the .312/.362/.483 line seem more likely? And should we pay less attention to playing time?