Shohei Ohtani and MLB’s Japanese Pitcher Fragility Myth

There’s a distinction between the skill that makes offseason centerpiece Shohei Ohtani most attractive to teams and the factors that have turned him into one of the most fascinating free agents in the history of professional sports.

Ohtani’s offensive abilities, CBA-suppressed initial salary, and resulting “almost every team enters, one team leaves” soap opera have made him the majors’ most interesting man during an offseason that Ohtani himself has slowed to a crawl. Dominant Nippon Professional Baseball closer Dennis Sarfate, an MLB export who succeeded Ohtani this season as MVP of Japan’s Pacific League (and who’s held Ohtani to a career 2-for-11 line), told me Monday that Ohtani possesses “off the charts” foot speed, “unreal” bat speed, and “some of the best power I have ever witnessed,” matching or surpassing the pop of sluggers such as Josh Hamilton, Nelson Cruz, Ryan Braun, and Prince Fielder. Despite those skills, Ohtani, who’s bound by the CBA’s severe restrictions on amateur spending, will make the minimum major league salary (or yearly totals close to it) until he hits arbitration or signs a team-friendly extension. No matter what the Marlins might say, Ohtani’s potential for massive surplus value makes his salary, plus a $20 million posting fee and smaller signing bonus, a preposterously small price to pay. Both his bat and his salary are as special as advertised.

It’s Ohtani’s arm, though—a prosaic selling point for a pitcher—that forms the foundation of his appeal. Ohtani touches triple digits as a starter, pairing his heater with a filthy forkball and slider. The arm that throws those pitches would make him the most coveted free agent on this winter’s market even if he hit like Bartolo Colón and his suitors had to pay full price for his services.

That arm is also the aspect of Ohtani that makes him most susceptible to a sudden decline. In Ohtani’s case, of course, durability isn’t a consideration at this stage; for the money he’ll be making, teams would more than recoup their costs after one successful season. But the impact he’ll have on whichever franchise signs him does depend on how he’ll age—and one school of thought holds that his NPB past puts him at an increased risk of flaming out early.

In the most recent edition of his long-running, highly regarded newsletter, Baseball Prospectus cofounder Joe Sheehan argued that some skepticism concerning Ohtani is warranted, not only about his ability to punish major league pitching but also about his long-term future on the mound. Sheehan observed, correctly, that the 11 Japanese-born starting pitchers who’ve made at least 40 major league starts after spending significant time in NPB—excluding Tomo Ohka and Mac Suzuki, who, as Sheehan noted, “came to the U.S. as pitching prospects”—have produced most of their value early on, averaging 2.4 Baseball-Reference WAR in their first through third MLB seasons and only 1.1 WAR in their fourth through sixth. (Among the 11, only Masahiro Tanaka and Kenta Maeda signed for more than six years.) “The track records grow more stark with each one who comes over: they have initial success and then they fade away or they get hurt,” Sheehan wrote. Later, he concluded, “All of the value in signing Japanese pitchers is up front.”

Sheehan is far from the first figure in baseball to cast doubt on the durability of ex-NPB pitchers. In late 2011, as teams prepared to bid for Yu Darvish, Tom Verducci made a similar argument in Sports Illustrated. Verducci wrote that when Boston had signed Daisuke Matsuzaka in December 2006, Red Sox analysts were worried about “the Third-Year Wall”—as Verducci described it, a tendency for ex-NPB pitchers to suffer “a sharp decline in performance about three years into their transition to Major League Baseball.” Verducci quoted multiple team executives who acknowledged this alleged limitation, among them then-Rangers assistant GM (and current Twins GM) Thad Levine, who told him, “The anecdotal assessment suggests starting pitchers have a two-year window of success followed by a rapid decline, followed thereafter by disappearance.”

Verducci attributed this pattern to “the strain of pitching more often (every fifth day, rather than every sixth or seventh day) [as is standard in Japan] over a longer season against deeper lineups with more power, to say nothing of different baseballs, ballparks, training regimens, cities, travel demands and cultures.” Sheehan advanced the same explanation, writing, “The differences between the two leagues, from the ball to the style of play to the season length to the pitcher-usage patterns may combine to raise a single barrier to long-term success.” In the past few years, other reporters have echoed the same sentiments, including Joel Sherman in the New York Post and Rangers writers Evan Grant and Gerry Fraley.

It’s true that the highest-profile pitchers to come from Japan have been at their best soon after emigrating. Mid-1990s trailblazer Hideo Nomo debuted with back-to-back fourth-place finishes in NL Cy Young Award voting but managed only a 91 ERA+ over the rest of his career. Matsuzaka suffered a steep decline amid many injuries after his first two seasons. Darvish and Tanaka, who’ve had their own elbow issues, are coming off campaigns in which their work was diminished. There are obvious exceptions to the apparent trend—most notably Hiroki Kuroda, a model of consistency whenever and wherever he pitched—but given those easily recalled examples of Japanese pitchers who faded after an initial splash, it’s easy to see why one would be concerned about Ohtani’s long-term performance.

However, what’s missing from most of these analyses is any kind of control group. It shouldn’t be noteworthy that Japanese imports tend to decline over time, because all pitchers tend to decline over time. Some research suggests that pitchers, as a group, head downhill from the moment they make the majors—and most of them make it to Major League Baseball much earlier than the typical player posted from Japan. The 11 NPB veterans Sheehan lumped together were 28.8 years old, on average, in their rookie years stateside. At that age, it would be odd if they didn’t decline within their first few seasons.

To determine whether ex-NPB pitchers are really riskier than non-NPB pitchers, we have to compare the trajectories of the two groups. For each of the 11 ex-NPB pitchers, I generated a list of non-NPB pitchers who met four criteria. To qualify as comps, the non-NPB pitchers had to have a season that fell within one year (both in time and in seasonal age), one WAR, and 20 innings pitched of the NPB pitcher’s first year in the majors. Yu Darvish, for example, was 25 in his 2012 rookie season, when he threw 191 1/3 innings and amassed 3.9 Baseball-Reference WAR. Therefore, his comp class includes 24-to-26-year-old pitchers from 2011-13 who threw between 171 1/3 and 211 1/3 innings and were worth between 2.9 and 4.9 WAR. Naturally, some of the pitchers in each import’s comp class will have been slightly younger or better than the pitcher they’re being compared with, and others will have been slightly older or worse. In theory, one could get even more granular in selecting the comps by classifying pitchers based on stuff or experience, but this method gives us a decent-sized sample and, as we’ll see, yields a group of pitchers with stats very similar to the NPB pitchers’.

The search for a control group yielded a combined 105 comps across the 11 ex-NPB arms. With those names in hand, I calculated the average WAR and IP produced by the 11 ex-NPB pitchers in their rookie MLB campaigns and the following five seasons. Then I did the same for the non-NPB comps. To avoid overweighting the non-NPB sample with any one pitcher’s comps, I calculated composite year-by-year averages for each of the comp cohorts—Nomo comps, Matsuzaka comps, Darvish comps, and so on—and averaged those averages to arrive at WAR and IP figures for the non-NPB pitchers’ “matched” rookie seasons and the following five. The most recent NPB products, Tanaka and Maeda, have pitched only four and two years in the majors, respectively, so I excluded them and their comps from the samples for subsequent years.

Overall, both the NPB pitchers and their non-NPB comps averaged 2.4 WAR and 159 innings in their first years in the sample, so their starting points are the same. After that, though, they diverge—and, in light of all the fretting about NPB pitchers, perhaps not in the way one would expect. The solid blue and red lines in the graph below show the year-by-year WARs produced by NPB and non-NPB pitchers, respectively. The dotted blue and red lines display the respective groups’ average innings pitched.

There’s only one possible interpretation here: The ex-NPB pitchers have been better bets, long term, than their non-NPB peers. The ex-NPB pitchers produced higher average innings-pitched totals every season from Year 2 through Year 6, and higher WAR totals in four of the five. Although both groups declined dramatically over the six years in the sample, the ex-NPB pitchers retained much more value on the back end, averaging a combined 1.3 WAR in years 5 and 6, relative to 0.4 WAR for their non-NPB comps.

Here’s one more way to express the NPB pitchers’ aging advantage. If we start with a hypothetical four-WAR, 200-inning pitcher and run him through the typical ex-NPB pitcher’s trajectory, he’d throw 954.7 innings and generate 17.6 WAR over six years (including that first, four-WAR year). Run the same starter through the typical non-NPB pitcher’s trajectory, and he’d throw 823 innings and record only 11.9 WAR. That’s a 48 percent WAR edge and a 16 percent innings edge in favor of the NPB pitchers.

Given the differences between leagues and the possibility that foreign pitchers might benefit at first from exploiting hitters who haven’t faced them before, the belief that Japanese pitchers decline particularly quickly sounds somewhat reasonable—right up until you really study the stats, at which point it falls apart. The myth was probably born from the fact “Japanese pitchers” is still a small enough sample that the individual declines are burned into our minds; when we think of that group, we immediately remember Matsuzaka, Nomo, and others who failed to sustain the hype that they generated right before and/or after their big league debuts. “Non-NPB pitchers,” meanwhile, is such an amorphous mass of humanity that it’s hard to hold on to specific stories and names. And for every comparable pitcher who held on to his value for several seasons—Dan Haren and CC Sabathia stand out in the “Matsuzaka comp” class—there are three who sank just as swiftly: Erik Bedard, Jeff Francis, Chien-Ming Wang, Carlos Zambrano, Jon Garland, Rich Hill (who bobbed back up years later). Kuroda’s comps include Derek Lowe, Jarrod Washburn, and Ryan Dempster, all of whom faded while Kuroda kept cranking out solid seasons. Darvish’s comp group has highlights like Stephen Strasburg and Gio González, but it also lists tarnished starters such as Jordan Zimmermann, Mat Latos, Matt Harrison, and Chris Tillman.

It doesn’t help the perception of ex-NPB pitchers that they tend to be old by big league rookie standards; although they’re new to MLB, they’re not new to pro baseball, and some observers seem to struggle to adjust their expectations accordingly for their post-NPB staying power. Ohtani is the exception; at 23, he’s significantly younger than any of the established NPB starters who’ve been posted previously. With his age, his build (a broad-backed 6-foot-4), and his even broader skill set, there’s scant reason to think that he can’t contribute beyond his first few MLB seasons—and his NPB background shouldn’t detract from that forecast at all. All pitchers present risk. But it’s not just Japanese pitchers who front-load their best seasons—it’s pitchers, period.