I spent the week following basketball, and my attentions were rewarded. Michigan had its best week since 1998 and it's now back in the Top 25 for the first time since then. I caught a Pacers game on Friday, my first viewing of LeBron James in person. He's a remarkable ballplayer, able to drive to the basket, to shoot jumpers without squaring-up (something he probably shouldn't do anyway), and to hit the open man even if he's 30 feet away. Even if you're a lukewarm NBA fan, I'd recommend seeing him if you get the chance, the same way I'd recommend listening to Willie Nelson even though I can't stand country music. I also saw Butler beat Wright State on Saturday. Butler isn't a bad team, around 80th in the latest Sagarins and they have a shot at making the NCAA Tournament if they can get around Wisconsin-Milwaukee in the conference tourney. But what struck me is that those Butler kids are the same age as LeBron and no individual could match LeBron in one single skill except free throw shooting.

Anyway, it's a slow time of year for Cardinals fans. Mark Mulder evidently won't play in the WBC; the Dutch team seems to have named its roster without consulting the players. Isn't that why we fought the War of 1812? Tony LaRussa is saying his usual silly stuff heading into spring training, and for now I just want to ignore it. Assistant GM John Mozeliak may be headed to Cincinnati, in which case the Reds' attempt to clone the Cardinals' organization will have begun in earnest. I don't know what exactly Mozeliak does, so that's not exactly pulling my eyes away from Michigan's upcoming schedule.

With my focus elsewhere and my lack of interest in Baseball Prospectus already present, I didn't notice that PECOTAs were released last week to subscribers. No real surprises there, as you'd expect from computer projections. PECOTA's not as bullish about Rule 5 pick Juan Mateo (5.14 ERA) as ZiPS, the only noteworthy number I could dig from the spreadsheet. I don't play fantasy baseball, so from my standpoint PECOTA's main advantage over its competitors is its pretty presentation, specifically the comparables. It feels like pseudo-science, but it fleshes out the projection to some extent. There are some odd ones. For example, in spite of a mediocre forecast, Chris Duncan's top comps include Derrek Lee and Nick Johnson. In spite of a good forecast (3.39 ERA), Chris Carpenter's top comps are Erik Hanson, Shane Reynolds and Jon Lieber. Hmm, Reynolds was a better pitcher than I remembered.

Pujols's top comp is Frank Robinson -- learn from your father's mistake, Mr. DeWitt. My favorite list is for John Rodriguez: Michael Tucker, Irv Noren and George Altman. I'll admit I didn't know much about Noren and Altman, but it's cool seeing J-Rod matched up with a couple of former All-Stars.

I did a little googling on PECOTA and found this year-old post at USS Mariner. It's a doubly cautionary tale about these kinds of projections. On one hand, as Zumsteg in essence says, it's just a model. It can't encompass all variables and even if it could, there's always the problem that reality won't be right (see Tony Womack's 2004 season). While many have understood this for a while, I'm just starting to appreciate fully the limits to modeling. On the other hand though, well, there's a subtext to that USSM post. Mariners fans love Ichiro and they liked the Beltre signing a year ago -- he was 26 and coming off an MVP-caliber season -- but they didn't love PECOTA's projections so much. In spite of its limitations, PECOTA pretty much nailed Ichiro (311 / 355 / 415 v. 303 / 350 / 436) and you know how things turned out with Adrian Beltre. Unless you do a boatload of research, you won't outsmart the model, especially with hitters. And if you're not careful, playing with the numbers could lead you to some awful conclusions.