Info on Blog

Friday, December 22, 2006

Why I Don't Like Matsuzaka's Deal

I posted the below on another site (Baseball Analysts) and it used some of my stats from here before so I thought I would put it here too.

If you compare the stats of all the Japan-born starters in baseball-reference, pre-US Japan stats, then U.S. stats, you will find that all the pitchers tend to see their HR/9-rate fly up 50% or more, K/9 fall a good chuck, 20-30%, BB/9 tends to go up and ERA increase by at least 1 point. I realize that it is small samples, but I find the fact that all experienced a big jump in ERA to be significant.

Comparative Stats for Japan Vs. MLB

Stats: ERA - WHIP - K/9 - BB/9 - HR/9 - Ages

Ishii
Japan: 3.38 - 1.29 - 9.7 - 4.7 - 0.8 - (18-27)
U.S.A.:4.44 - 1.53 - 6.9 - 5.7 - 1.1 - (28-31)
After: 3.44 - 1.33 - 8.6 - 3.0 - 0.8 - (32)

Irabu
Japan: 3.37 - 1.30 - 9.1 - 4.1 - 0.6 - (19-27)
U.S.A.:5.15 - 1.41 - 7.1 - 3.1 - 1.6 - (28-33)
After: 4.44 - 1.42 - 8.3 - 2.5 - 1.4 - (34-35)
2003: 3.85 - 1.35 - 8.5 - 2.4 - 1.2 - (34)

Nomo
Japan: 3.15 - 1.32 - 10.3 - 5.0 - 0.7 - (21-25)
U.S.A.:4.13 - 1.35 - 8.7 - 4.1 - 1.1 - (26-36)

Yoshii
Japan: 3.43 - 1.29 - 5.6 - 2.8 - 0.8 - (20-32)
U.S.A.:4.68 - 1.36 - 5.3 - 2.6 - 1.3 - (33-37)
After: 4.87 - 1.40 - 3.9 - 1.9 - 1.3 - (38-41)

Matsuzaka
Japan: 2.95 - 1.14 - 8.7 - 3.2 - 0.7 - (19-26)

Igawa
Japan: 3.15 - 1.24 - 8.5 - 2.9 - 0.8 - (19-26)

Findings and Thoughts

I have seen some projections go from low 3's to mid-3's, but only one near 4 (4.01 by Baseball Prospectus, saw this on Roto Authority's take on D-Mat). I would go with a high 3 ERA since that is 1 point above his career stats, with a likelihood of mid-4 if he struggles like most of them did their first year, probably over the new culture, new everything.

Why? A number of reasons. HR/9 increased from 0.3 to 0.5 for these pitchers - any of that would put him at or above the 1.0 max you want from a good pitcher. It will help that he will pitch in Boston for half his games (roughly) as HR is depressed there by around 14% according to Bill James corrected park factor stats. But you have to realize that lousy hitters have gone to Japan and become HR hitters to see that HR must be depressed relative to the U.S., no matter how you slice the data. It's going up, even in Boston.

In addition, his K/9 should drop as well, and all his brethren saw drops, and for the high K/9, drops of 1.6 to 2.8, or about 15-30%. Dropping him 2.0 like Irabu would put him at 6.7 - still good, but hardly dominating - but a drop of 30% would push him to 6.1 and near the danger area where the pitcher has to be a really good control pitcher to survive. His dominating K/9 rate will not survive to the U.S.

Also, WHIPs have gone up as well, though with a low 1.14, it can go up .11 like Irabu and still be a very good 1.25.

But that is dependent on him reducing his BB/9. Irabu and Nomo were able to drop it a 1 point, but Ishii saw his rise 1 point. The drops counteracted a rise in H/9 (not in chart but can see via WHIP and BB/9 that they rose to counterbalance the drop in BB/9). Boston has been relatively neutral for walks (Bill James) so he won't get any help or hurt there. But Yankee Stadium helped Irabu probably (Bill James stats for last three seasons, not Irabu's seasons) and Dodger Stadium hurt Ishii and Nomo probably (same problem again). With a big drop in K/9 though, he needs to get his BB/9 down to at least 2.5 to have a superior K/BB still.

I think those are all big question mark areas that are reasonable doubt that Matsuzaka will deliver on $103M in money paid over 6 seasons. I think something like Millwood's 2003 season in Philly would be a good comparison: 4.01 ERA, 0.8 HR/9, 2.8 BB/9, 6.9 K/9, 1.25 WHIP, though a bit low on HR/9. Or Bartolo Colon's 2003 season: 3.87 ERA, 1.1 HR/9, 2.5 BB/9, 6.4 K/9, 1.20 WHIP. Nice, but is it really worth $17M per season?

3 comments:

  1. "is it really worth $17M per season?"

    Ab-so-lute-ly.
    They have the money to spend, and they needed to replace Shilling after 07.

    Zito will get the same money for declining numbers as Dice-K's improve.

    With salaries increasing 10% a year (and more for premier starters), this is a 'looking better every year' coup for the Sox.

    btw, you are too astute an analyst to use Fat Toad's numbers for comparison. In fact, there probably is no comp for DK - he's that good.
    Think Barry.

    ReplyDelete
  2. nomo is the most comparable guy, yet you're taking his numbers from when he was 26-36 for US data. those numbers would show decline for almost anyone who's not named roger clemens.

    Then i saw you looking at HR/9. I don't think you can say that if he gives up more homeruns than 1.0 than he won't be a good pitcher. chien-ming wang doens't strike out anyone and is still a good pitcher. in the end it's their results that count. basically i just don't see a direct causal link from homeruns/9 to success.

    Also there is another way to slice it. Parks in Japan are smaller, thus more conducive to hitting homeruns. that's why guys over here go over there and become prolific homerun hitters. Even if japanese hitters are weaker overall, that could be balanced out by the stadium size factor and who knows if it'll be depressed or elevated.

    ReplyDelete
  3. eugene, the reason I showed Nomo's numbers for 26-36 is double-fold: Matsuzaka's contract will cover his 27-32 years and assuming he does well enough, Boston will re-up to cover years 33-36, at the least, plus, that's the only data that's available for comparison.

    Here's Nomo's numbers from 27-32, since you want comparable numbers:

    Nomo: 4.30 - 1.36 - 9.3 - 4.2 - 1.2 (27-32)

    Those are actually worse numbers than for the entire age period I provided. I guess his name is really roger clemens? :^)

    The HR/9 is just another factor, it makes it harder to be a good pitcher if you are giving up so many homers. It is not a be all or end all, but it adds to the overall puzzle solving exercise.

    However, I find it puzzling that you don't see the direct causal link from HR/9 to success since whenever there is a homer, there's at least one run scoring, and on average more than one run.

    Yes, I am totally aware of the smaller parks, Terry Whitfield of the Giants went from hitting 10 homers if he was lucky here in the U.S., then goes over to Japan and suddenly he's a homerun hitting god. My main point is that Matsuzaka's HR/9 is probably depressed because all the other Japanese pitchers came over and experienced a steep increase in their homerun rate.

    I agree that Chien-Ming Wang doesn't strike out anyone and has been a good pitcher. The point is that historically, most pitchers do not de well with bad peripherals. Yes, there are exceptions to the rule, and Chien Ming Wang appears to be one, as does Roger Clemens, in terms of excelling after 40, but do you want to bet $103M that Matsuzaka is one of the exceptions?

    Yes, results count, but Matsuzaka hasn't pitched yet in the majors, so there are no results for us to look at for him except for the results of other pitchers coming over to the U.S. from Japan. And the results is that they have pitched pretty well in Japan but then came here and suffered increases that made their overall pitching performance worth less than the $103M that Boston is paying him for the next 6 seasons.

    ReplyDelete