Wednesday, September 20, 2006

Comparing League Averages

Before I get into the main feature here, a few short points:

1. Peter Gammons wrote a nice piece about his illness and subsequent recovery. It is a heartfelt thank you to those who worried about him and helped him fight his way back. He takes a moment to comment on baseball near the end, and seems to think that Matsuzaka to the Yankees is a forgone conclusion. I hardly think that's true, but he must have his reasons for saying so, beyond the obvious. I'm sure we'll hear more as the season closes and the posting period begins.

2. A lot of Red Sox bloggers and bulletin boards have started to link to me. The same can be said for Cubs fans. Thanks for the support folks. You know my allegiance is with the Yankees, and I would give my right arm to see Daisuke in Bombers' pinstripes next year, but I'm also very happy as a Matsuzaka fan that you have all entered the arena. Welcome.

Now for the main attraction.

In an attempt to normalize the numbers between the Japanese Pacific League and the American League, I compiled the numbers to date for the 2006 season of all pitchers with 100+ innings pitched to their names. I use these leagues as my sample because Matsuzaka is most frequently associated with the AL, and the Pacific League is the half of NPB with a designated hitter, making the samples a bit more equivalent. That sample included 54 American Leaguers and 23 Pacific Leaguers. It's not a perfect 1:1 ratio, but it is large enough to draw some basic conclusions about what we might expect from a player in the Pacific League should he enter the AL as a starter.

The following table represents the league average 100+ inning pitcher in both leagues. It also included a "per 9 inning" set of ratios to compare the numbers on a more even footing, and a conversion rate for some Equivalent Averages I'll use later. I'll take the next step after you've had a chance to digest the numbers below, and I've made a few comments.

These numbers can be understood by referring to a series of pieces written by Clay Davenport of Baseball Prospectus on the Japan to MLB conversions. Read the piece to get yourself up to speed on the educated reasoning on the high level of Japanese play as compared to the Majors, but I can sum it up for you quickly here. Davenport, and many other SABRmetricians, have concluded that Japanese baseball is generally closer to the Major League level than it is to AAA. Many believe Japan is a kind of AAAA league, but the BP guys have placed NPB even closer to the Bigs than that.

By my personal estimation, the difficulty in scouting Japan is that there are scores of Major League caliber players mixed in with guys that would struggle at the AAA level in the US. It's not that the league as a whole is AAAA, but rather that each team has some below replacement level players eating up space. On a whole, the stars in Japan will fare very well according to the metrics. The numbers you see above show that a league average pitcher in the AL is roughly the equivalent of a league average pitcher in Japan's Pacific League. There is a slightly favorable set of ratios in the Pacific, mainly due to the playing style and the presence of replacement level players on each team. I've used the "per 9 inning" ratios to build conversion rates that I will apply to a handful of pitchers in Japan, including Daisuke Matsuzaka. I contend that these pitchers would approach the EQAs I am producing once inserted into a Major League rotation. All statistics are projected out to a 200 IP standard.

I see a couple of things that need a bit of tweaking right away. The ERAs seem to be a bit low, even given the conversions. I think there is a park factor element that has kept the Japanese numbers a bit on the low side, as well as the amount of sacrifice bunting that goes on. That having been said, I think it's fair to say that the viability of most of these pitchers as American League prospects is great at the top, but tapers off to average and below fairly quickly. Let's examine the sample against the AL average pitcher.

1. ERA

There are 11 pitchers on this list who fare better than the AL average. That amounts to about half the frontline pitchers in the Pacific League. If we add a quarter to a third of a run to all the ERAs, we only eliminate two pitchers from the + side of things. Seibu rookie Hideaki Wakui, and Nagisa Arakaki of SoftBank get booted. Even with a margin of error that would account for some extra run production, the Japanese #1s and 2s fare well in the AL.


Again, 10 players find themselves above the AL average in WHIP when we do our math. A slight bump in these numbers cuts the sample to around 7 players, which isn't great, but still looks good for the aces.

3. Hits per 9

11 guys are again on the plus side. Add another hit per 9 to everyone and you whittle the number to 6. I would guess that this number, more than any of the others would become inflated and help to raise the ERAs a bit. Major Leaguers swing away and Al batters especially never bunt.

4. HRs per 9

9 players are above AL average in this category, but I think the 3rd starters may see more of a power surge in their balls put in play in the AL and I think most of these guys will be hurt by the huge sluggers and live ball in the Majors. The DHs in the Pacific League are often good hitters, but there aren't so many Mark McGwires running around Japan.

5. BBs per 9

There are 8 players that qualify as + control pitchers here. I think the general trend among Japanese starters is that they possess good control and make hitters swing the bat. The problem is, players generally oblige and you see a lot less working the count in Japanese ball. The style of play that has been made popular by high OBP clubs that have had success may inflate these walk totals, although I think it's a fairly accurate sample.

6. K-Rate (Ks per 9)

11 pitchers feature better than average K-rates in the American League. I think there's a big fall-off from the top guys to the league average pitchers that are near the bottom of the list. I have a hard time believing that Ryan Glynn would have a better K-Rate than C.C. Sabathia, Randy Johnson, or Roy Halladay. Dropping these numbers doesn't hurt the two aces too much, but evens the playing field to MLB standards just a bit.

7. K/BB Ratio

I love this stat and I think this is where we see the value of Daisuke Matsuzaka against his peers. A 5.31 K/BB ratio would put him hundredths of a point behind Johan Santana, and the rest of the field would find themselves gradually tumbling down the list with the Paul Byrds and Carlos Silvas of the world.

When we consider Daisuke Matsuzaka in this equation, and Kazumi Saito for that matter, we have to recognize that there is top shelf talent in Japan that can dominate AL batters. I would argue that, including the Central League, there are probably 10-15 pitchers that would fare better than the MLB average. Remember, that means simply that there are a dozen or so Japanese starters that could come to the US and avoid embarrassing themselves every 5 days. If you really examine the sample of players that could come in and dominate at the front of a rotation, you're talking about 6-8 guys that would make an MLB top 3, and 4 or 5 that could front a rotation. That sounds reasonable, and helps the case that Matsuzaka is a #1 in the American League.

The final word on this is that Kazumi Saito turns 29 this year. Hiroyuki Kobayashi is 28. Shingo Ono is 31. The rest of these players are very young, including Matsuzaka and you might see the next generation of Japanese stars emerging from this list as some of the old guard faces a decline period. Provided the arms stay healthy, a handful of these guys may actually get a chance to pitch in the Majors before their 30 birthdays, and maybe sooner.


Post a Comment

<< Home