Darvish A Cut Above Other Japanese Imports

This past week, the Texas Rangers landed Japanese ace Yu Darvish for a cool $112 million. The right-hander’s virtues are apparent: he has what’s considered an ideal pitcher’s build (6-foot-5, 215 pounds), he throws 95 mph, and he thoroughly dominated the competition for the Pacific League’s Nippon Ham Fighters in his early-to-mid-twenties. The projection systems either like the 25-year-old (a 3.62 ERA and a 169/46 K/BB ratio from Dan Szymborki’s ZiPS) or want to propose cyber marriage (a 2.40 ERA and a 223/44 K/BB from Brian Cartwright’s Oliver).

But Darvish’s signing has also been met with some skepticism. Some starting pitchers coming from Japan to the U.S. have found success (Colby Lewis after initially bombing in the majors, Hiroki Kuroda), but many others have disappointed.

Daisuke Matsuzaka, Kei Igawa, Kaz Ishii, Hideki Irabu and Hideo Nomo all got lots of cash and press, but Dice-K is the only pitcher among that group to post a career adjusted ERA better than the league average in Major League Baseball (and you’d be hard-pressed to find anyone who would say he has lived up to Boston’s $103 million investment). Critics say that for whatever reason — cultural adjustments, four days’ rest between starts instead of six, pitching backwards in a more fastball-heavy league, exhaustive workloads at a young age that eventually take a toll — Japanese pitchers haven’t lived up to the hype. What makes Darvish any different?

With that in mind, I want to take a closer look at Darvish’s body of work in Japan compared to other starting pitchers who have come stateside. I found each pitcher’s ERA, K/9, BB/9 and HR/9 over their last three seasons prior to arriving in the majors and compared them to the league average for those seasons. This is crucial because offensive levels in Japan have changed drastically over the years, and they plummeted in 2011 due to the introduction of a smaller, smoother baseball (the average ERA was 2.95, a full run lower than in 2010). I then put pitchers’ ERA, K/9, BB/9 and HR/9 on a scale where 100 is average, over 100 means the pitcher was better than average in that category and under 100 means he was below-average. A pitcher’s most recent work is more instructive of what he’ll do in the future, so finally I weighed the last season prior to coming to the majors more heavily (50 percent) than the second (30 percent) and third (20 percent). These numbers aren’t park-adjusted, but they’ll give us a quick-and-dirty comparison of Darvish to other starters.

So, how does Darvish stack up with other Japanese starters who have made the jump over the past two decades? Going by adjusted ERA, Yu is in a class all by himself:

Player	Years	Age	ERA+
Yu Darvish	2009-2011	22-24	215
Daisuke Matsuzaka	2004-2006	23-25	170
Hideki Irabu	1994-1996	25-27	146
Kenshin Kawakami	2006-2008	31-33	144
Colby Lewis	2008-2009	28-29	140
Hiroki Kuroda	2005-2007	30-32	134
Kaz Ishii	1999-2001	25-27	131
Kei Igawa	2004-2006	24-26	118
Hideo Nomo	1992-1994	23-25	116

That’s a huge gap. Darvish’s 215 ERA+ easily trumps Matsuzaka’s (170 ERA+), and none of the other starters on the list had an ERA that was even 50 percent above the league average. How did Yu lap the competition? Let’s take a closer look at his component stats and how they rank compared to the other Japanese starters:

Missing Bats

Player	Years	Age	K/9+
Kaz Ishii	1999-2001	25-27	156
Hideo Nomo	1992-1994	23-25	155
Hideki Irabu	1994-1996	25-27	153
Yu Darvish	2009-2011	22-24	147
Daisuke Matsuzaka	2004-2006	23-25	140
Colby Lewis	2008-2009	28-29	136
Kei Igawa	2004-2006	24-26	129
Kenshin Kawakami	2006-2008	31-33	123
Hiroki Kuroda	2005-2007	30-32	100

By strikeouts per nine innings, Darvish ranks fourth. But this measure seriously underestimates Darvish’s ability to fool batters compared to the three guys ahead of him. Why? Because, as you’ll see in a minute, Darvish did a far better job of limiting walks than Ishii, Nomo and Irabu. Pitchers with poor control face more hitters per inning and thus have more chances to strike out hitters each frame. Ideally, I would have used strikeout percentage for all of these pitchers (Ks out of total batters faced), but that sort of data isn’t available on Baseball-Referece before 2007. Darvish’s weighted strikeout percentage was actually 58 percent better than the league average from 2009-2011 (Lewis’ was 45 percent above average). Considering Darvish’s control was also solid, you’d have to say that he was the best at missing bats.

Limiting Walks

Player	Years	Age	BB/9+
Colby Lewis	2008-2009	28-29	266
Hiroki Kuroda	2005-2007	30-32	178
Kenshin Kawakami	2006-2008	31-33	174
Daisuke Matsuzaka	2004-2006	23-25	157
Yu Darvish	2009-2011	22-24	153
Kei Igawa	2004-2006	24-26	125
Kaz Ishii	1999-2001	25-27	98
Hideki Irabu	1994-1996	25-27	96
Hideo Nomo	1992-1994	23-25	58

See? Those high K/9 totals for Ishii, Irabu and Nomo came in part through innings extended by bushels of base on balls. Lewis was Maddux-esque in limiting walks. Yu comes in at the middle of the pack, behind crafty vets Kawakami and Kuroda and in the same area as Matsuzaka. Dice-K, of course, has gone on to drive pitching coaches, managers and fans bats$!t crazy by nibbling and walking 4.35 batters per nine innings in the majors, the 11th-highest rate among qualified starters since 2007.

We often hear of how Japanese baseball is much more breaking ball and off-speed heavy than baseball in the states. Pitchers, some with 6-7 offerings, must pare down their repertoire in the majors and rely more on the fastball to get ahead in the count. One theory for Matsuzaka’s huge walk total is that he has never learned to trust his fastball, going to his breaking and off-speed stuff a maddening amount in even and hitters’ counts. Darvish also throws a gaggle of different pitches. Take a look at his pitch selection in 2011, from Patrick Newman’s NPB Tracker website:

Darvish’s 2011 pitch mix

Pitch	Pct. Thrown
Fastball	45.8
Slider	25.5
Shuuto	9
Cutter	8.6
Curveball	5.3
Changeup	3.2
Forkball	2.6

There is one potentially big difference between Matsuzaka and Darvish that might make Yu less prone to pitching backwards in the states: Darvish brings the heat. Rangers CEO Nolan Ryan noted that Darvish’s fastball has picked up speed in recent seasons, and we see that on Darvish’s NPB Tracker page. He sat 90-92 in his early twenties, but he was up to around 95 mph with his fastball for most of the 2011 season. Here’s what ESPN analyst Keith Law had to say about Darvish’s fastball in his top 50 Free Agents preview:

Darvish generates good arm speed through hip rotation, and despite slightly late pronation, his arm works reasonably cleanly and he repeats the delivery well. Darvish will show the usual assortment of pitches, led by a 91-95 mph fastball that’s been reported up to 97 this year

.
Scouting reports on Matsuzaka had him sitting 92-95 in Japan, but he has been toward the lower end of that spectrum in the majors (just under 92 mph). That fastball hasn’t been effective, either, with a run value of -0.3 per 100 thrown.

It’s hard to say exactly how Darvish’s fastball will translate — will he keep the 95 mph gas, or will he too lose a couple ticks while getting two fewer days’ rest in the big leagues? We won’t know for a while, but Darvish is starting from a higher baseline than Matsuzaka velocity-wise. Maybe that extra zip and instruction from Ryan and pitching coach Mike Maddux will make Darvish less bashful than Matsuzaka has been about pounding the zone with his fastball. That, in turn, would lead to fewer free passes.

Keeping the ball in the park

Player	Years	Age	HR/9+
Yu Darvish	2009-2011	22-24	290
Hideki Irabu	1994-1996	25-27	198
Daisuke Matsuzaka	2004-2006	23-25	187
Colby Lewis	2008-2009	28-29	137
Hideo Nomo	1992-1994	23-25	135
Kaz Ishii	1999-2001	25-27	122
Hiroki Kuroda	2005-2007	30-32	114
Kei Igawa	2004-2006	24-26	100
Kenshin Kawakami	2006-2008	31-33	98

This is where Darvish has truly shined. He has served up just five dingers in each of the past two years, and Patrick Newman tells me that one of them in 2011 was a “running home run” — an inside-the-park job. Park factors could explain part of Darvish’s dominance in this category. Newman says that the Sapporo Dome, home of the Ham Fighters, has high outfield walls (around 19 feet tall), lots of foul territory and the following dimensions: 328 feet down the lines, 381 feet in the power alleys and 400 feet to center field. Oliver creator Cartwright estimates that the Sapporo Dome suppresses homers by about eight percent compared to a neutral Japanese park, close to the 94 HR park factor that Matsuzaka faced at the Seibu Dome (Lewis had a tougher go of it with Hiroshima’s new park in 2009: Mazda Stadium has a 126 HR park factor).

Still, Darvish scarcely gave up any shots on the road either, and he was so far above everyone else that park factors wouldn’t fully explain his low HR totals even if the Sapporo Dome played like Petco Far East. That stinginess in surrendering homers will serve him well in Arlington, given that Rangers Ballpark increases round-trippers by 19 percent for left-handed hitters and 14 percent for right-handers.

Conclusion

Whether you’re going by scouting reports or stats, Yu Darvish stands out as the best Japanese starter to try his hand at retiring major league hitters. With top-notch strikeout rates, decent control and precious few homers served up, Darvish’s adjusted ERA was leaps and bounds better than that of other Japanese imports. That doesn’t guarantee him success, as he’ll have to adjust to a different pitching schedule, a new ball and a more fastball-centric league. It’s OK to be skeptical. But claiming Darvish is the next Matsuzaka, Igawa, Irabu or Nomo ignores the fact that he has a much better track record than those guys did upon coming to the show.