A Series of Unrelated Items!
It’s too bad the exclamation point is completely played out. Thanks to the unstoppable forces of the internet and sarcasm, you can’t use one without implying irony or misplaced enthusiasm. I’m not attempting to do that, I’m just trying to gussy up a links post, because some of this stuff is really interesting. The rest of it is Mariners-related.
1: With all the pitch fx data out there, it was only a matter of time before we started to get solid data on pitch types, and how each pitcher’s results (by pitch) differ from league averages. Fangraphs started showing league averages in the pitch fx sections of player pages, which was cool, but if you want to know what the league average whiff rate was on a slider, well that was tougher. Trip Somers’ site added those data, but what about HR rate? What about batted balls? Harry Pavlidis’ post at The Hardball Times has you covered. I keep coming back to this after months, and it still raises new questions.
The slider has the best raw results, while the change looks great from a FIP standpoint. Of course, the slider’s always been a pitch with a serious platoon split – but why is that?
2: Well, it may be because off-handed hitters see it early. I find the last two graphics in this Dave Allen post fascinating. Having platoon-split data by pitch is wonderful, and having a plausible explanation for why it exists is a perfect accompaniment.
3: Some of you may have seen Colin Wyers’ article at Baseball Prospectus asking a simple question: how do we know how good defensive metrics are? Are they better than they were, and if so, how much?
This started over a year ago with some investigations into UZR’s year to year reliability, and then Colin started to investigate bias in the play-by-play data itself. The article today sums it up: if we have issues with the data (that persist over time), what does reliability really mean? What are we measuring? The question spurs great responses from the usual suspects (Tango, Mike Fast, et al.).
While this seems pretty technical – and it is – I think Fast’s article at THT does a great job of laying out why it matters and why it’s important that these questions get a fair hearing. In comments, Tango and others point out that Wyers’ question isn’t the same as a conclusion. The take-away here isn’t: defensive metrics are worthless. It’s, how do we know how good they are? We can look at how UZR’s constructed, and we note the more granular data we collect these days, and both seem like they’d add a lot. But how do we know? Is this equivalent to moving from using ERA to tRA (a massive, huge improvement)? Or is it marginal?
One of those statements that I’ve never understood (think “Ichiro hurts this team by…”) is something to the effect that the ‘stats guys’ care only about confirming each others results; that it’s a club of people telling each other they’re smart. The on-going debate over defensive metrics shows how ridiculous that claim is, and it does so as quietly and effectively as Ichiro doing whatever Ichiro’s doing renders the vague, inchoate complaints about him meaningless.
4: In last night’s 3-2 loss, the M’s left Casey Kotchman in to face LOOGY-masquerading-as-closer Brian Fuentes. Apathy can affect players/staff as much as it affects M’s bloggers, but it’s still a rather remarkable decision. I know the team’s got injuries (Russell Branyan’s run-in with a rogue hotel coffee table being the most recent and the most bizarre), and the go-to guy for a situation like this – Mike Sweeney- is on rehab with the Rainiers. But how about Milton Bradley? How about Josh Wilson? Dave mentioned on twitter that Fuentes’ FIP v lefties is 0.53 and against righties it’s 6.39. If you’re trying to win the game, you pinch hit. If you’re trying to ‘showcase’ Kotchman, you pinch hit.
Incidentally, reason this season has been awful and surreal #459: I’m writing about how Josh Wilson should’ve been given a high leverage AB.
5: It may be nothing, but divisive prospect (“5 tools!” “Makes Wlad Balentien look like Kevin Youkilis!”) Greg Halman has cut his K rate markedly in July, and hasn’t struck out in 13 PAs, which would seem like a vanishingly small number for anyone but Greg Halman. Halman’s flourished in the PCL, with 20 HRs and a wOBA of .400 and solid CF defense. He’s clearly not ready for the majors at this point, but in a season of disappointments on the big club, Halman heads the list of surprises on the farm. We’ll see how long this ‘improvement’ keeps up, but it’s nice to finally see a tolerable K-rate to go along with the prodigious power, even over a short sample. You can quibble about a decline in walk rate, but he’s been hitting so well that it seems more like nit-picking. Matt Mangini and Halman’s seasons in AAA are among the biggest surprises of 2010. Too bad the biggest is that Chone Figgins has been a replacement level player thus far.