When Acronyms Disagree: PECOTA's Favorite Hitters

I've spent plenty of time talking about the accuracy of PECOTA and CHONE, two of the most reliable publicly available projection systems, on this site. But while they generally come to very similar conclusions, they don't get there the same way. PECOTA relies heavily on a player's comps while CHONE (I believe) mostly just adjusting their previously compiled stats. So I thought it might be interesting to look at some guys who project very differently in the two systems. Today I'll examine four hitters that PECOTA is bullish on.

Cristian Guzman

PECOTA: .323/.361/.455
CHONE: .287/.331/.407

Well, yeah, I can see how a guy with the following numbers over the last four seasons would be hard to project:

2005: 492 PAs, .219/.260/.314

2006: 0 PAs

2007: 174 PAs, .328/.380/.466

2008: 579 PAs, .316/.345/.440

I don't think it's a stretch to say that PECOTA is giving essentially no weight to Guzman's miserable 2005 season. And, even without knowing that he had LASIK surgery between '05 and '07, that seems pretty reasonable. First of all, the .260 OBP is now four years behind him. But more importantly, he has pretty clearly established a new level of performance over his last 753 PAs. PECOTA may not know about the eye surgery, but his comps seem to suggest that for players with this progression, something has changed.

Not surprisingly, CHONE doesn't see it that way. If you assign any relevance at all to his '05 campaign, it's going to be tough to project him at better than league average. And really, while I think I see where PECOTA's coming from, projecting the soon-to-be 31-year old to improve on his excellent '08 does indeed seem pretty optimistic.

Jayson Werth
PECOTA: .283/.380/.511
CHONE: .264/.365/.445

The thing that really jumps out here is the SLG, as PECOTA's .380 OBP is essentially his average over the last two seasons. But having a guy whose career SLG is .451 projected at 60 points higher than that is definitely a little strange. His top comp is Jason Bay, who saw his SLG increase by 104 points from his age 28 to 29 seasons; of course, that can pretty clearly be explained by his '07 knee injury, but maybe that's factoring in.

Even BP isn't so sure about Werth, as his writeup in the annual notes that he's "really not an effective everday player". I think you have to side with CHONE on this one, althought I suppose the fewer PAs he gets against righties, the better chance has of approaching that .891 OPS.

The Justin Upton
PECOTA: .278/.361/.506
CHONE: .257/.349/.444

Here's a situation where I think you have to prefer PECOTA's methodology. CHONE actually has him declining from the .250/.353/.463 line he put up as a 20-year old. Sure, that's partially because he hit just .221/.283/.364 in the bigs in '07, but: he was 19!

Upton's BP writeup includes a reference to Griffey's first year, when he hit .264/.329/.420 as a teenager. And Griffey, who does show up at 10th on Upton's comps list, was up to .327/.399/.527 by the time he could legally drink.

You do have to account for the possibility that Upton really figures things out all at once. PECOTA "only" has his 50th percentile line at .269/.352/.484. His weighted mean projection -- which is the one listed above -- is given a boost by his extremely high breakout rate (59%), and ridiculous 90th percentile projection of .312/.396/.597.

Matt Wieters
PECOTA: .311/.395/.544
CHONE: .274/.352/.439

I used a requirement of 100 '08 ABs in the majors to filter out some of the 931 hitters projected by PECOTA, but that didn't strike me as a particularly good reason to exclude the best prospect in the majors.

To say Nate Silver's system is a fan of Wieters would be a significant understatement, as it has him leading the entire American League in VORP (at 59.6) if he gets a full year of PAs. His .365/.460/.625 line at AA Bowie last year is impressive enough, but until "researching" for this post I hadn't realized that the Bay Sox play in a significant pitchers park. So much so that BP has his '08 Bowie performance as the equivalent of a .627 SLG (!) in an average big league park.

True, his SLG in single-A Frederick only comes out to a .513 EqSLG, but PECOTA is still projecting his EqA to slide significaintly from last year's .351, to .319. At first it seems crazy to expect such a lofty line from a rookie catcher, but few 22-year olds -- at any position -- have ever had a year like Wieters' 2008.

Photo: ESPN.