Monday, May 3, 2010

I tread where AJ Mass fears to tread

AJ Mass, professional fantasy sports analyst for and amateurish stat-head, today published an article on the law of averages and the impossibility of accurate fantasy prognostication.  Well thanks AJ.  His major premise is that subsets of statistical data based on random events are not predictive of subsets of future random data.  In terms of contribution to the mathematical community this isn't exactly a revelation.  He even simplifies the game of baseball to a sequence of roulette wheel spins, and then uses that argument to justify (improperly) that end-of-year aggregate data projection for baseball is a fool's errand (or a Gambler's).  I find his conclusion rather humorous, considering AJ Mass is the author of the Daily Notes feature on ESPN's fantasy baseball page, which uses 2009 and 2010 pitching statistics to do a favorability ranking for every starting pitcher that day versus specific opponents.  So while on the one hand crapping on the accuracy of data projection re: full season outcomes, here he is writing a daily article purporting to know whether Livan Hernandez is a better start this week than Brett Cecil.

Now maybe his article is a reaction to the undoubtedly thousands of complains he's received since the season began whenever he's gotten a pick egregiously wrong in Daily Notes, but I found it amusing that he's gone on at length to describe his own personal worthlessness in the fantasy sphere.

During the article, Mass took a pass on predicting the future value of three different players under different career circumstances:  Colby Rasmus (the Hot Start), Mark Texeira (the Slow Start), and Roy Oswalt (a category he treated as Personal History versus a particular team, when the more interesting treatment would have been the Can This Old Veteran Keep It Together).  In accordance to my earlier writing regarding the primary focus of majority fantasy writing, I'd like to talk about the sleeper, Colby Rasmus.

Unlike most rookies, he was afforded a full season of work, starting the 2009 season on the Cardinals roster.  His line of .251/.307/.407 72/52 was awful by most standards (certainly fantasy ones), especially for players with no speed (3 stolen bases all year).  His 20% strikeout rate was, well, not very good either (though not quite as bad as some of the other Cardinals outfielders), and his .282 BABIP suggests he wasn't unlucky enough to write off the season as a cosmic fluke.

So then what can we make of his current .288/.422/.630?  Here are the relevant supporting figures:

18.7 BB% is almost triple his 2009 rate and 50% more than his historical minor league walk rates.
.375 BABIP is higher than average but within realistic boundaries.
37.0 K% means Austin Jackson is going to have a run for his money.
.342 ISO means there's some serious power here.

The walk rate improvement is, I think, the most notable piece of information to examine.  Mike Silver of The Hardball Times wrote the following in October:
His 50.1 Swing% is possibly the most disturbing of all, as any hitter who swings this often will never garner many free passes. In addition, his 78.6 percent Contact% shows that he is missing on pitches too often. The Contact% is fixable. However, if he can’t refine his swinging tendencies, his OBP outlook will be severely capped due to lack of walks and strikeouts. Assuming pitchers continue to throw him in the zone at approximately the same rate as in '09 (50.9%), Rasmus would have to drop his swing percentage by a few points to see any real gains in his walk rate—and this kind of approach overhaul is not easy to do.
If Contact% is fixible, Colby must not have heard.  His contact rate is down 6.3% from last year.  However, he's dropped his swing rate 10%.  The combination of seeing more balls and swinging at fewer of them has resulted in the dramatic increase in his walk rate.  Is the power legit?  Rasmus slugged .512 and .551 in 2006 and 2007.  His .396 in AAA in 2008 could be worrying, but can be attributed to a knee injury if one were so inclined as to make excuses.

The problem with his sterling minor-minor league performance (and his 2010) is that we have a full year in the majors to use as a comparison, when he was healthy and had a starting job, and could only put together a .407 slugging%.  If I had to guess, I'd say a fair component of that poor power output is a direct result of swinging at bad pitches, and that his improved discipline at the plate portends good things for Rasmus's 2010 season.  He'll come down from .670-land, but I don't think a .275/.375/.520 is out of the question, if we can map his earlier power and later success to a projection.

Not bad for the 18th round.

No comments:

Post a Comment