Responding to "Nico's Guide to Optimal Roster Performance"

By Jason Wojciechowski on February 12, 2011 at 9:45 PM

About a week and a half ago, Nico at Athletics Nation, the A's site I most love to shit all over, presented four points about roster, lineup, and bullpen management. I decided to write a response even before I saw what Nico wrote, though in this case, the expected shitting-all-over didn't occur -- these are mostly reasonable points (if somewhat obvious). The one I take exception to, mainly because of the strident tone, is, lucky for us, the first one.

Here, then, is Nico's first point: rest Kurt Suzuki once a week (either via off-days or sitting). Now, I'm not morally oppoesd to this. Suzuki's 2010 wOBA graph, from Fangraphs, displays an alarming downward tendency. The problem is that in his 2008 and 2009 seasons, when he batted more and caught more innings at catcher, the same trend doesn't appear. In other words, it's not clear that Suzuki catching so much that he gets tired is actually a problem we can diagnose from the available data.

That said, the projected offensive difference between Suzuki and Landon Powell is nil -- Marcel says they'll both post a 93 wRC+ and PECOTA pegs them to put up TAvs of .253 and .249. Suzuki looks nimble and quick, Jack-like even, blocking balls behind the plate, while "Tiny" Powell appears slow and lumbering, but Matt Klaassen had Powell doing quite well at preventing passed balls and wild pitches last year. In fact, accounting for playing time, Suzuki and Powell had almost exactly the same defensive value in Klaassen's system (-0.8 runs in about 1100 defensive PAs for Powell; -3.2 runs in about 4400 for Suzuki -- funny how that works out).

Mayhap the answer we're looking for isn't "Get Suzuki more rest!" It might well be "Platoon these guys!" Powell's a switch-hitter after all, albeit one with an extremely limited set of available platoon data. Unfortunately for this theory, Suzuki actually has a reverse split of .033 wOBA points in his career (that is, he has hit righties significantly better than lefties -- for what it's worth, it's been almost entirely BABIP-based, with a .231 mark against lefties vs. .295 against righties; his walk and K-rates are actually better against lefties and his isolated power is nearly the same). The thing is that he's not nearly deep enough into his career to have any confidence in that split (doing some math I found on the internet, I believe you'd estimate his true split to be about 0.007 points of wOBA -- that's positive points, i.e. even with his observed negative split so far, it's still safest to assume that he's like almost all other right-handed hitters in the history of the game, hitting lefties better), so sitting him against right-handed pitchers, especially tougher ones, is probably still the right move from the outside.

Of course, "from the outside" creates a huge amount of fudge-space. What do the A's know about the relative defensive abilities of the two players? About their psychological profiles relative to the pitching staff? About Suzuki's swing mechanics and the likelihood that his weird BABIP split is real? About the odds that Powell really is a .310 wOBA guy instead of a guy who'd be exposed with more playing time? This isn't a situation where we're screaming at the Braves for putting Alex Gonzalez in Yunel Escobar's place, a situation, in other words, that is so obvious that the outside data overwhelms the size of the fudge-space. All I can reasonably conclude from the evidence laid out here is that I can't conclude anything.

Nico's post, I will note cattily, contains nary an Arabic numeral in the three paragraphs about Suzuki. I report, you decide!

Because I wrote way too much about Kurt Suzuki, let me briefly blow through Nico's other three points by saying: "Sure" (with zero sarcasm, however we tend to use this particular s-word). He wants to rest Hideki Matsui occasionally against lefties, putting Conor Jackson in left and Josh Willingham at DH, giving both of the old guys with creaky bodies a chance to recuperate while also letting Jackson, who has hit well against lefties in his career, have a chance to succeed and contribute. To this I repeat: sure, sounds good. Nico lays out a batting order that isn't egregiously bad (i.e. Kouzmanoff, Ellis, and Pennington are his bottom three), and therefore: sure. Finally, he says that the bullpen is so stacked that nobody should be pitching four-out-of-fives or three-in-a-rows. Sure.