Fangraphs had an interesting article http://www.fangraphs.com/blogs/index.php/dice-bb which illustrates, very well, the difference between the varied points of view in baseball cyber-space.  In this paradigm, we simply adjust Matsuzaka's "luck stats" (strand rate, BABIP, HR/F*, etc) so that they are equal to the league average, and boom.  We realize that you can't walk 5 men a game and be a good pitcher, ergo, Matsuzaka isn't a good pitcher. The paradigm here is purely mathematical:  if a pitcher gave up few HR/F and a low BABIP, that was chance, and so we need to adjust it back when predicting the future. It's possible that's correct.  And it's certainly true that Matsuzaka doesn't project to a 2.80 ERA this year. IF the question is, "Does Matsuzaka figure to give up more than 2.80 runs a game in 2009?" then the answer is, of course he does.  He's pitching in Fenway Park, for one thing.   His BABIP and HR/F* may even rise some, and his S% will definitely rise. And IF the question is, "Do you want to be AWARE of what it looks like when you normalize 'luck stats'?,", then the answer is, yes, of course you want to be aware of those things. But let's also think about the paradigm that factors in Matsuzaka's free moral agency.  :- )  The, um, dice don't control everything. . === The Exception to the Rule, Dept. === Notice that in 2007, Matsuzaka gave up 25 homers while walking 3.6 men, and that in 2008, Matsuzaka gave up only 12 homers while walking 5 men a game.  This suggests immediately to the Jamesian (classic) sabermetrician that Matsuzaka simply chose to never give in to the hitters. Daisuke Matsuzaka hates it when anybody hits the ball hard against him.  When he is walking 5 men a game, it is not because he can't throw strikes! It is because he is choosing to "nibble." In 2008 -- if you watched the games -- you saw that Matsuzaka was teasing and nibbling and frustrating and baffling the hitters all year long.   He might fan a guy and then walk a guy, and then maybe a hit would fall in -- and then Matsuzaka would simply choose to stop walking hitters.  Here comes the overhand curve, and here's the fastball located on the hands, and here's the forkball, and see you later. When he had the luxury of nibbling, that's what he did.  When he couldn't afford to walk anybody, he did not. And that's why it looked so weird to see a 5 BB per game pitcher run a low ERA. Here's a rule for ya:

Pitchers who walk 5 men a game have lousy control.

It's a rule, and it's right 99% of the time.  But I got your exception right here, my friend.  Because every other pitcher who walks 5 men a game has lousy control.   DiceK is most definitely an exception to that rule.  That's why you see every other guy with 5 BB's getting his head beat in, whereas you see DiceK go 18-3.  Because DiceK was a 5BB pitcher with very good control. . === A Picture's Worth 1,000 Words Dept. === This comment, from a guy who has SEEN Matsuzaka pitch, was spot on:

As a huge Sox fan who has watched most of Dice-K’s starts, your statistical analysis misses something you see with the naked eye: they guy is really hard to hit. He’s only given up 124 hits in 163.7 innings (.76 H/Inn) this year, which is Petco-esque. He takes all kinds of chances and puts guys on base because he has tremendous confidence that no one can hit him. It’s maddening to watch, but it works, and has worked for two years (especially when you can keep the ball in the park).

Don't forget that this brainy pitcher is performing in Fenway Park.  If DiceK were in Yankee Stadium, you'd see a different pitch selection. The best comp to Matsuzaka remains our original suggestion of David Cone:  a 91-94 mph fastball, a slider, an overhand curve, a fork ... five or six strikeout pitches and great pitchability.  http://www.baseball-reference.com/c/coneda01.shtml  Cone also had some years, early on, when he walked some guys -- and for exactly the same reason.  He liked swings on pitches outside the strike zone. . === W-L, ERA Dept. === Matsuzaka's 15-12, 4.40 followed by 18-3, 2.80 -- that's the kind of thing that leads people tell you to ignore W's and L's.  But you'll notice that the career W's leaderboards are not populated by Cha Seung Baek and Brian Bannister.  Take W's, L's, and ERA over a period of years, and you get a valuable read on what was happening in the games. DiceK, over the course of two years, has now averaged 18-8, 3.72 per 162 games. Don't plan on 18-3, 2.80.  But do plan on seeing weird, wonderful, and winning ways from Daisuke Matsuzaka in 2009.   18-8, 3.72 isn't a bad guess. If you're guessin' that DiceK isn't a real good pitcher, guess again.  :- ) Cheers, Dr D .