---

Wednesday, October 12, 2011

Should the Yankees Make a Play for Yu Darvish?

The reason for the last Pitcher A vs. Pitcher B thread was to segue into the question in the title.  Whether it’s fair or not, Darvish is going to be compared to other Japanese pitchers that came to the majors.  While it’s not certain he’ll be posted, it’s probably worth exploring in case he is.

Statistics only tell us so much about how good a pitcher is, particularly when trying to assess him in a new league against more difficult competion.  So although I’m going to try and crunch some numbers here to take an educated guess at what Darvish might do in the majors, it’s going to be important that the scouting reports on Darvish feel he’s a viable candidate to provide value as a starting pitcher in MLB.

The track record of Japanese pitchers in MLB is mixed.  Going through the list of Japanese league pitchers who’ve come over to pitch in MLB gives me the following names.

Fukumori, Kazuo
Hasegawa, Shigetoshi
Igarashi, Ryota
Igawa, Kei
Irabu, Hideki
Ishii, Kazuhisa
Kashiwada, Takashi
Kawakami, Kenshin
Kida, Masao
Kobayashi, Masahide
Komiyama, Satoru
Kuroda, Hiroki
Kuwata, Masumi
Matsuzaka, Daisuke
Nakamura, Micheal
Nomo, Hideo
Nomura, Takahito
Ohka, Tomokazu
Okajima, Hideki
Otsuka, Akinori
Saito, Takashi
Sasaki, Kazuhiro
Suzuki, Mac
Tadano, Kazuhito
Takahashi, Hisanori
Takahashi, Ken
Takatsu, Shingo
Tateyama, Yoshinori
Uehara, Koji
Yabu, Keiichi
Yabuta, Yasuhiko
Yoshii, Masato

There are differences in Japanese leagues and parks which I haven’t bothered to try and account for.  You also have starters and relievers and a wide array of ages in this group, which is also something I’m not worrying about right now.  You also have generally older pitchers coming over in terms of the typical player peak, and I ignoring that as well.

I figured the most basic question to ask is how did these pitchers’ performance in MLB compare to their performance in Japan.  So I used matching innings for each pitcher’s Japan time and MLB time to see if I could figure it out.

The idea behind matching innings is to ensure we have the same weighing of Japan and MLB data.  Basically, I compared the pitchers’ career MLB innings to their career Japanese innings and used the smaller of the two as the matched innings.  So in the case of Kazuo Fukumori, I pro-rated his performance over 677.3 Japanese innings to his 4 MLB innings.  I did that for everyone, and then added up the cumulative Japan and MLB performances.  So every pitcher contributed in the same percentage to both data sets.

League	IP	H	R	ER	HR	BB	SO
Japan	8682	6400	2719	3327	485	3398	7858
MLB	8682	8232	4281	4004	1067	3438	7192
Japan->MLB		1.29	1.57	1.20	2.20	1.01	0.92

FIP: Fielding-independent pitching

This is a look at how the pitchers’ cumulative matched inning performances compared between Japan and MLB.  The Japan->MLB line is just the ratio of MLB divided by Japan for each stat.  So on average, we saw an increase in hit rate of about 29%, an increase in runs and earned runs, a more than doubling of HR rate, a fairly consistent walk rate, and about an 8% decrease in strikeout rate.  One thing I’ll mention on the R/ER thing is that the gap between runs and earned runs doesn’t look right to me.  I don’t know why the data (which I pulled from The Baseball Cube) shows such a discrepancy.  It could be that official scoring in Japan is more liberal with assigning errors, or there may be a data error in the earned runs column.  Because of that, I’d probably focus on RA/FIP instead of ERA.

Then, I just took a 3-2-1 weighted average of Darvish’s 2008-2010 performance and applied those changes in the various component stats.

Player	Type	IP	H	R	ER	HR	BB	SO	WP	BK	RA	ERA	FIP
Darvish, Yu	Japan	195	141	43	39	7	46	201	5	0	2.00	1.78	2.33
Darvish, Yu	Translated	195	181	68	47	16	46	184	22	1	3.15	2.15	3.10

That’s a #1 starter’s line, but it’s also probably optimistic.  So I figured it’d be interesting to do the same exercise for Daisuke Matsuzaka for the three years prior to his arrival in MLB.

Player	Type	IP	H	R	ER	HR	BB	SO	WP	BK	RA	ERA	FIP
Matsuzaka, Daisuke	Japan	189	148	54	48	12	40	197	6	0	2.59	2.29	2.59
Matsuzaka, Daisuke	Translated	189	190	86	58	26	41	180	23	0	4.07	2.76	3.76
Matsuzaka, Daisuke	Actual MLB	189	173	92	89	19	91	172	7	0	4.39	4.25	4.16

I’ve added a third line here, which is just Matsuzaka’s career MLB performance pro-rated to the 189 innings of his 3-2-1 weighted pre-MLB line.  As you can see, Matsuzaka was obviously not as good as the straight translation would have pegged him.  He’s been somewhere between 0.3 - 0.4 runs per nine worse according to RA/FIP.

So what if I just used the difference between Matsuzaka’s translated line and actual line applied to Darvish’s translated line?

Player	Type	IP	H	R	ER	HR	BB	SO	WP	BK	RA	ERA	FIP
Darvish, Yu	Translated*	195	165	74	72	12	104	177	6	0	3.40	3.31	3.78

That still looks pretty solid to me.  It’d almost certainly be the best projection on the Yankees aside from CC Sabathia if he doesn’t leave. 

As far as I know, Darvish has solid MLB stuff.  He’d also be two years younger than Matsuzaka was when he came over.  The concern I’d have would be if his style of pitching is conducive to MLB success.  The big jump in Matsuzaka’s walk rate is an illustration of why that’s a potential problem.  If Darvish’s style is based more on getting batters to swing at bad pitches and less on having good command, he may suffer more in MLB than the raw numbers indicate.  That’s where the scouts come in.

But at least from what this tells me, I’m all aboard the sign Yu Darvish bandwagon.  Since the posting fee won’t count against the luxury tax, and there’s no draft pick compensation to lose if he’s signed, it’s a gamble worth taking in my opinion.