Sunday, March 14, 2010
The Echo Of Justin Speier: Royals 12, Angels 3
- The steady erosion of PECOTA for pitching, and the explosion of new metrics. I imagine many people are happy to see this state of affairs, because the flaws of that system have been hammered on for years. Unfortunately for people like me who aren't genuine seamheads and miss the finer points of the arguments, or fall asleep reading them, what I glean is that there have been no genuinely straightforward revolutions in sabermetrics since Voros McCracken's discovery of DIPS. It's probably a gross oversimplification to say that all the other systems are filigrees (and this will undoubtedly get my Empirical Knowledge Club card revoked), but I have yet to read an article anywhere that doesn't become an unbelievably long and intimidating slog.
- Opaque writing. It's not that I mind things like WXRL — introduced by Keith Woolner in the 2005 dead tree edition of the Prospectus — and yes, I largely nodded off reading that, too — but that same year, Nate Silver came up with his QERA metric, and now I read it's being eaten alive by SIERA? If you think metric obsolescence is bad, check out this graf from the introduction (emphasis mine):
With that in mind, we have invented a new statistic, Skill-Interactive Earned Run Average (SIERA), which corrects the problems with old estimators while adding a few more realistic assumptions. This was done first by un-foiling all of the individual components in QERA while making an adjustment for the issue with the ground-ball denominator issue, and testing to see which interactions and squared terms were relevant by using multiple linear regression analysis. Essentially, we changed the GB/BIP to (GB-FB-PU)/PA and evaluated all of the terms in the exponential regression, removing those with insignificant p-values; while the QERA formula only shows three variables, un-foiling the formula reveals several more. We identified two terms that were not useful: the squared term of walks, and the interaction between walk and strikeout rate. The squared terms on strikeout and ground-ball rates were both significant, and we also found important interactions between walks and grounders and between whiffs and grounders that have strong effects on run scoring.Double. You. Tee. Eff. Over. "Un-foiling"? I'm guessing that they mean to reveal by that. (Is this some sort of trade lingo in the statistical world?) It's the kind of leaden prose, the itch to be nifty rather than clear, that makes me want to snuggle up to the know-nothings who scurried under the stove after the Moneyball light came on.
- Where's the byline? BPro has a policy of only identifying the editors (this year's was Steven Goldman and Christina Kahrl) of its annual. Several years ago, Jay Jaffe did the honors for the Angels, and his trademark, lively writing is a joy to read. Not that I'm complaining about the quality of writing in this year's publication, mind you, but it seems to me that if Baseball America can identify the authors of each of its team sections, so should BPro.
What's obvious is that the Angels need pitching depth, now more than ever. Unfortunately, if yesterday's performance is any indication, the guys on the B team behind the likely 25-man-roster are not ready for the Show, and may never be. Rafael Rodriguez and Ryan Brasier both got lit up by the Royals' B+ squad, Rodriguez in particular giving up a couple longballs that were no doubters to right center. The recent model Angels have managed to win with depth, and for the first time since 2002, they don't appear to have it.
Far better to base projections on one spring training game, or perhaps reading chicken entrails.
As for the bylines, BP has stated multiple times that they don't identify authors because multiple writers and editors contribute to each team section.