August 4, 2011
The Asian Equation
Finding Relief from NPB
The last group in my analysis of the player’s who have migrated to MLB from Nippon Professional Baseball (NPB) are the relievers, the least appreciated members of a successful baseball team. Yet, of all NPB imports, they have been the most numerous (explaining the length of this article, for which I apologize in advance) and the cheapest. Diminished quality is the most obvious reason for these extremes, since starters who don’t meet MLB standards get shifted to the bullpen, and lesser talents also keep salaries down. Additionally, the typical NPB pitcher’s arsenal matches well with an MLB reliever’s skillset.
As I discussed in my last Asian Equation article, NPB is a breaking ball league, which translates better to relief than starting. A good breaking ball might fool major league hitters the first or second time they see it in a game, but it probably won’t the third or fourth time. As an illustration, here’s how batter OPS rises against two of the biggest NPB starting-pitcher busts as compared with three current MLB pitchers: the best, the most mediocre, and an old junkballer. While MLB batters’ performance improves against each pitcher the more times they see him in a game, the change is far more dramatic with Matsuzaka and Kawakami.
NPB pitching expert Patrick Newman of NPB Tracker offers two further reasons for NPB relievers’ success. First, relievers’ workloads in NPB is similar to MLB, unlike the workload of starters. NPB starters go deeper into games, have longer rests in between starts, but pitch fewer innings per season. NPB relievers, much like MLB relievers, throw between 55 and 75 innings per season in 1-2 inning chunks. Second, NPB relievers who have found success in MLB typically bring their “out” pitch with them. Patrick points to the success of the sliders of Takashi Saito and Akinori Otsuka and the changeup of Hideki Okajima; I’d add Kaz Sasaki’s forkball and Koji Uehara’s splitter. A good MLB reliever from any country needs an out pitch, of course, but needing just one plus pitch makes it easier to succeed.
NPB relievers also share other general requirements with their MLB counterparts: the need to stay healthy and eliminate walks. The most successful relievers on this list brought their “out” pitch with them, avoided injury, and kept their walk rates low. When they succeeded, however, the return—both on and off the field—was far more moderate than starters or position players. This, too, has several causes. Nearly all signed as free agents, effectively halving their costs by eliminating the posting fee, and when they did sign, it was typically for much less. Finally, relievers have lower WARPs, giving us no incredible bargains; except for some outliers, gains and losses were less extreme.
For all these reasons, NPB relievers have proven to be the best values, at least according to the rough metric of dollar-per-WARP, with the average return on free-agent investments being $2.3M per WARP in 2007-8. As with the rest of the series, I’m indebted to Patrick Newman at NPB Tracker for helping formulate the ideas behind this analysis. BP intern Alex Chiang also helped with research into many of the statistics and salaries here. In the final column of this series next month, I’ll see what conclusions we can draw from this import analysis and what we can look forward to in future transpacific exchanges.
Masanori Murakami was the first Japanese-born player to jump to MLB, succeeding with great control of a middling fastball and an outstanding curve. He put up impressive numbers in his brief stint with the Giants, especially compared to his NPB career thereafter. In Japan, he played in a much smaller park, and his coaches forbade him from throwing brushback pitches—then considered unsportsmanlike. Converted back to starting, Murakami had a few good years, eventually compiling a 103-82 record with 30 saves, registering a 3.64 ERA and a 1.24 WHIP including a 4.1 K/9 and a 2.5 BB/9. Not bad, but hardly Hall-of-Fame-worthy.
Murakami’s value is obviously boosted by 1960s-era reserve-clause salaries. In 1964, the minimum salary was $6,000, and the average salary was $14,863. Murakami made $15,000 in 1964, but his 1963 salary is hard to determine, so I’m assigning him the major-league minimum. In today’s dollars, his salary would have put him at $0.08M per WARP, still atop the list.
The Rangers didn’t release Tateyama’s contract numbers, so I used the major-league minimum, possibly inflating his value. Still, he’s delivered for Texas thus far in 2011, pouring in strikes with a repertoire that includes a two-seamer, curve, and a Vulcan-grip screwball. Tateyama’s sidearm motion imparts tons of movement to his pitches while maintaining great control. It’s easy to overanalyze Tateyama’s small sample, and he’s also at the cheap end of his career, so whether he regresses or gets paid, he’s likely to lose his spot atop this list.
The Yomiuri Giants’ first-round draft pick, Kida began his career as a starter before becoming the primary closer for the Orix Blue Wave. Kida signed with Detroit, where he had a 6.42 ERA and 1.61 WHIP between 1999 and 2000. The Tigers cut him loose, he returned to Orix for two seasons, and then Los Angeles and Seattle gave him another crack, but he saw little MLB action with either club. Christina Kahrl liked to sharpen her verbal cutlery on Kida, calling him “one more crash test dummy with a Japanese name on his uniform,” “squid scraps,” “a body for junk innings,” and “yet another ronin moundsman from the Empire of the Sun to draw and entertain.” All of it is hilariously true, though his fungibility and low price made Kida a bit of a bargain.
The most valuable of the NPB success stories, Otsuka is the only reliever to come via the posting system, though San Diego paid just $300K; without this posting fee, his per-WARP cost drops to $1.2M. This low fee came in part from Otsuka’s unassuming NPB career. Though his overall numbers were strong (2.89 ERA, 1.01 WHIP, 10.7 K/9, 3.0 BB/9, and 137 saves), he’d battled shoulder problems in 1999 and only spent four of his seven NPB seasons as his team’s sole closer.
With San Diego, Otsuka enjoyed early success with his fastball, but his best pitch was a slider that tumbles like a splitter. Combined with a hesitation delivery, this arsenal served him well in his first season; his step backward in 2005 came more from control problems than stuff. Batters increased their contact rate against him, particularly on strikes, even as they swung more against balls. Otsuka did well after changing leagues and teams in 2006, and his control returned before injuries ended his 2007 season, eventually requiring three different Tommy John surgeries. If he’d stayed healthy, hitters might have caught up with him, but it’s appropriate that one of NPB’s biggest relief successes also succumbed to the most familiar problem among NPB imports.
Saito’s durability (second only to Shigetoshi Hasegawa) is unsurprising from a pitcher who logged over 1,500 innings starting and relieving for the Yokohama BayStars, where he also averaged 8.6 K/9, never dipping below 6.1. He brought those strikeouts with him to Los Angeles: his rookie-season 12.3 K/9 was second only to Brad Lidge among pitchers with at least 75 innings pitched. Saito’s whiffs slipped to 10.9 K/9 in 2007, but he honed his control from 2.6 to 1.8 BB/9. These combined to create a 1.40 ERA, 0.72 WHIP, and an All-Star berth. Saito’s easy motion, combined with the hitch often employed by NPB pitchers, made his mid-nineties heater seem even faster, and he kept batters off-balance with a slider, shuuto (similar to a two-seamer) and occasional curve. He was also unafraid of pitching inside: he led NPB three times in hit batters.
After elbow problems, Saito avoided surgery, resurfacing with the Red Sox in 2009, followed by successful stints with the Braves and Brewers. Strand rates skew the Boston and Milwaukee numbers in his favor, but Saito remains an outstanding NPB bargain reliever and is one of the few to rely on his fastball.
After ten years with the Hanshin Tigers, Yabu signed with the 2005 Oakland A’s, bringing with him the broad repertoire of a typical NPB pitcher. Yabu touched the low 90s with his fastball, but his offspeed and breaking stuff proved more effective. They only worked when he kept it in the zone, however, which he didn’t do enough to compensate for his low strikeout rates. Oakland cut him loose, he missed making the Rockies’ roster in 2006, but then the Giants gave him another shot in 2008, when he again exhibited the same control problems. Fungible enough for a marginal WARP return at a relatively low price, Yabu shows that roster filler can come from either side of the Pacific.
Delivering small-sample value, Takahashi wasn’t a great bargain for the Mets. He was no great shakes in Japan, either: in fourteen years of starting and relieving for the Hiroshima Toyo Carp, his ERA was always higher than 3.50 with consistently high walk rates and a career 6.6 K/9. Toronto signed him before the 2009 season, released him after a spring training calf injury, and the pitching-hungry Mets signed him immediately. Takahashi’s primary stats look good, but his .280 BABIP, 82.9 percent strand rate, and (hence) FIP all show his ERA could have been worse. The Mets also only used him in low-leverage situations; he was given a lead in just one of his twenty-eight appearances, and all four times he entered a tie game, he exited with the Mets trailing. Instead of trying his luck in 2010, Takahashi opted to return to Japan; otherwise, he would likely have been far lower on this list.
The success of “Okey Doke” and his eponymous changeup created a rush in NPB reliever signings, just as teams sought the next Ichiro. Okajima (one of Patrick Newman’s favorite pitchers) spent twelve years in NPB, mostly in a setup role for the mighty Yomiuri Giants, where he typically whiffed more than a batter per inning and registered BB/9 rates around 4.
Okajima’s 2006 Red Sox signing slipped under the radar due to the big contracts given to posting-fee stars Daisuke Matsuzaka, Akinori Iwamura, and Kei Igawa, but Okajima turned heads with his rookie season—an All-Star year when his 8.2 K/9 and 2.2 BB/9 led to a 2.22 ERA and 0.97 WHIP. Batters were baffled by his changeup, which allowed him to get away with a fastball that only reached the high 80s. These two pitches and a looping curve induced a 44.6 groundball rate and a .238 BABIP, the latter combining with his 85.6 percent strand rate to produce a FIP more than a point higher than his ERA. The gap between ERA and FIP steadily narrowed after his career year as his BABIP rose and his strand rate fell; he also started giving up more fly balls and home runs. Boston put him in successively lower-pressure situations, even demoting him to Triple-A this season. His peak and cost combine to make him a very good import bargain, despite the rapid erosion of his skills.
The Mets gamble in signing Takahashi—who never found consistency as a starter or a reliever for the Yomiuri Giants—has paid off. Patrick Newman has a great explanation of Takahashi’s NPB career on NPB Tracker, noting his typically wide variety of pitches— particularly a screwball with sinking action that brings strikeouts and ground balls. That package served Takahashi well with New York, where he began as a middle reliever, shifted to a starting role, and then returned to the bullpen, going 8-for-8 in save opportunities as the Mets’ closer after K-Rod’s injury. As BP 2011 notes, Takahashi enjoyed more success as a reliever than as a starter, displaying the fool-me-once tendency noted in the introduction: the first time they saw him, hitters had a .598 OBP, which leapt to .918 in their second plate appearance and .990 in their third.
Takahashi’s success with both New York and the Angels offers a model NPB import outcome as a breaking-ball reliever who translates his out pitch success to MLB. He’s pricier than comparable pitchers, but that’s appropriate for a player who’s succeeded in high-leverage situations. If he continues this season as PECOTA projects, his per-WARP price drops to $1.9M, making him an even better bargain.
The longest-tenured NPB reliever, Shiggy’s per-WARP cost comes in just below average. A former Rookie of the Year in Japan, Hasegawa started for six seasons with the Orix Blue Wave, compiling a 57-45 record, 3.33 ERA, and 1.33 WHIP. His 5.1 K/9 and 2.9 BB/9 show his lack of dominance; he succeeded by controlling his breaking pitches and dominating changeup, a rare pitch in breaking-ball-happy Japan.
Hasegawa lacked the tenure to become a free agent, and the posting system hadn't been created yet, but he somehow persuaded Orix to sell him to Anaheim, where he found success in the bullpen. Troy Percival convinced him to begin strength training, boosting Shiggy’s fastball into the low 90s, and Hasegawa studied the tendencies of opposing batters. As he improved, Anaheim shifted him into higher leverage roles, and he was soon setting up (and sometimes filling in) for Percival. He continue in a setup role with Seattle, putting together his best season ever in 2003 while filling in at closer for an injured Kaz Sasaki, registering 16 saves, a 1.48 ERA, and a 1.10 WHIP while earning his only All-Star berth. That season came from luck more than skill, as evidenced by career lows in K/9 (3.9), BABIP (.250), and a 93.2 percent strand rate that helped create a 3.78 FIP.
Hasegawa succeeded in MLB because of his fastball, change, slider, and (later in his career) a splitter. He defies conventions for NPB success since he lacked an out pitch and survived despite inflated walk rates. Perhaps because he came to the States early in his career, he didn’t suffer the usage injuries that have ended the careers of other relievers. Longevity and good-but-not-great performances created the best of the NPB relief corps, showing that sometimes good is good enough.
Tadano was the second player to bypass the NPB draft for MLB, following Mac Suzuki, as described in my last Asian Equation article. A top college pitcher, Tadano had appeared in a gay porn video, so conservative Japanese teams simply didn’t draft him. Cleveland had no such compunctions, and Tadano did well in their lower minors before reaching the big club in 2004, a year highlighted by a July 2 game against Cincinnati when he whiffed ten in seven innings, earning a win in the 15-2 blowout. It would be his only major-league win, however, thanks to an underwhelming fastball, slurve, and changeup; only the last fooled batters but not enough to bring consistent success. He wallowed in minor-league mediocrity with Oakland for two seasons before returning to Japan, where he’s still refining his eephus pitch. Relatively low cost led to mild success, but he’s far from an import bargain.
Uehara began his career with the Yomiuri Giants, earning both Rookie of the Year and Sawamura (Japan’s Cy Young) awards after a 20-4 season that included a 15-game win streak, 2.09 ERA, and 0.89 WHIP. Uehara started and relieved in NPB, maintaining BB/9 rates between 1.0 and 2.0, often leading to WHIPs near or below 1.00. As closer in 2007, he racked up 32 saves with a 1.74 ERA, striking out more than a hitter per inning for the first time. With the Orioles, Uehara resisted the bullpen, but injuries eventually pushed him there, a good thing considering he’s nearly doubled his strikeout rate in relief while batters have hit 85 points lower against him. He’s added velocity in all his pitches, lifting his strikeout rate to 11.6 K/9 while maintaining the control that was his hallmark in Japan.
His weakness is his fly-ball profile, leading to a home run rate of 1.1 HR/9, a tendency that comes from his repertoire. BP 2011 describes his arsenal thus: “He cuts the fastball in two different ways to produce different movement, and also has a splitter that dives into the ground, an occasional slider, plus a very slow, wide curve, and all of them are thrown with impeccable control.” The home runs come when he leaves his four-seamer up in the zone to righties, which could hurt him in his new Texas digs. On the more expensive end of the scale, Uehara’s done quite well and has the rest of the season to drive his price down—if he meets his PECOTA WARP projection for the rest of the year, his price comes down to $3.6M per WARP. As someone who has been able to translate his dominating pitches and control from Japan, Uehara remains one of the best NPB reliever success stories.
Takatsu was the NPB career saves leader, earning the nickname “Mr. Zero,” not because of the obscure Batman villain, but for his clean slates in 11 Japan Series appearances. Takatsu’s 260 saves were supported by a 3.13 ERA, 1.30 WHIP, and a mildly unimpressive 7.1 strikeouts and 3.0 walks per nine. Along with a Japanese commentator who evidently raided Craig Sager's wardrobe, this NPB All-Star Game footage shows Takatsu inducing weak contact with a two-seamer and a Bugs Bunny changeup. Those pitches also produced good results for the White Sox in 2004, when Takatsu assumed closing duties, finishing with 19 saves, a 2.39 ERA, and a microscopic 0.98 WHIP, fueling “Shingo Time!”
His 7.2 strikeouts and 3.0 walks per nine were almost identical to his NPB career numbers, but his .207 BABIP, 84.3 percent strand rate, and 3.91 FIP all pointed towards regression, which came with a vengeance in 2005. His BABIP returned to .296, and while he started off the season by saving 7 of his first 8 chances, his 6.28 ERA, 1.71 WHIP, and 1.065 OPS against (despite a .263 BABIP) showed that the contact he induced was no longer soft. He also walked too many batters, which Ozzie Guillen cited when naming Dustin Hermanson as his new closer. The White Sox released Takatsu in mid-July (later awarding him a World Series ring), and he signed with Bobby Valentine’s Mets, where he pitched fairly well but not enough for any team to give him a shot in 2006. Since leaving MLB, Takatsu became the first pitcher to earn a save in MLB, NPB, Korea, and the Chinese Professional Baseball leagues. One good season isn’t enough to call Takatsu a bargain, and his per-WARP fee reflects that, but he was far from a failure, at least in 2004.
As the dominating closer for the Yokohama BayStars, Sasaki set the NPB’s all-time (229) and single-season (45) save records, with a career 2.31 ERA and 1.00 WHIP. His success came from an excellent 12.4 K/9, a manageable 3.4 BB/9, and a low hit rate (5.6 career) that gave him six sub-1.00 WHIP seasons. No NPB reliever since Murakami had thrived in MLB, so Seattle took a chance by signing Sasaki to a two-year, $9M deal before the 2000 season. Sasaki didn’t disappoint, winning Rookie of the Year honors for a club-record 37-save season, 3.16 ERA, and 1.16 WHIP. Though he mixed in the occasional curve, Sasaki’s main pitches were his low-90’s four-seamer and devastating forkball, which combined to hold opposing hitters to a .181 average. He reined in his walks and home runs in the next season, setting a new club record for saves with 45 and becoming an All-Star for the first time.
After another strong 37-save season in 2002, Sasaki signed a two-year, $16M deal, but his personal life finally caught up with him. A heavy drinker, Sasaki hit the DL twice in ’03, the second time for a cracked rib allegedly sustained by falling on his suitcase, a story that few believed. He made just 35 appearances, notching 10 saves with a 4.05 ERA and 1.38 WHIP, all MLB career worsts. Sasaki asked out of the last year of his contract to return to Japan, citing problems with his marriage. Had he stayed, his punishing pitch regimen might have caught up with him; he threw 50 pitches before an appearance and 100 pitches on his off-days. As it stands, his per-WARP price is higher than average, appropriate for a player with an excellent (albeit brief) peak, though his time in MLB is indubitably tainted by his abrupt departure.
After ten years relieving for the Yakult Swallows (just one as their full-time closer), Igarashi inked a two-year deal with the New York Mets. In his first six NPB seasons, he consistently exceeded 10 K/9, though he also topped 3.9 BB/9 in four of those seasons. He tied the NPB record for fastest pitch (98 MPH) before undergoing Tommy John surgery; he could only hit the mid-nineties thereafter. The Mets hoped he’d bring his dominating heater-forkball combo as their setup man, but he’s struggled with his control and batters haven’t been fooled by the forkball, swinging at just 25 percent of pitches outside the zone. He’s bounced on and off both the DL and the bus to Triple-A Buffalo, recently returning from the latter after a dominant 31-inning stint (0.87 ERA, 0.77 WHIP, 2.6 BB/9, and 9.9 K/9). There’s still a sliver left of the season for Igarashi to drive his per-WARP cost down (or up) significantly, so the jury’s still out, but the prognosis doesn’t look good.
Toxic Assets (Negative WARP Return)
When considering negative outcomes, the per-WARP metric doesn’t work well, so I’ve arranged these in descending order of negative WARP, with more expensive players appearing later.
The Mets brought Kashiwada to their training camp from the Yomuiri Giants before the 1997 season, and Bobby Valentine thought that he could become a valuable LOOGY. Kashiwada faced 89 righties and 56 lefties, and fellow lefties hit .289/.411/.400 off him while northpaws hit .289/.375/.500. The Mets demoted him in August, recalled him in September, and then released him after the season, but not before he made history by becoming the first Japanese pitcher in New York to pitch and win a game.
Kobayashi faced high expectations with the Cleveland Indians thanks to an impressive NPB resume. As the Marines’ closer and key cog of the YFK trio, Kobayashi racked up 227 saves over eight seasons, including his best season, 2002, when he saved 37 games (including a NPB record 33 in a row) while posting a 0.83 ERA and a 0.74 WHIP. Kobayashi stepped into the closing role after Joe Borowski came back to earth and closer-in-waiting Rafael Betancourt flopped. Kobayashi converted his first two chances, but inconsistency—particularly in surrendering the longball—diluted Eric Wedge’s confidence, and Jensen Lewis eventually swooped in and seized the reins. Kobayashi’s fastball-slider mix neither overwhelmed batters nor pleased umpires, though his splitter enjoyed occasional success. In 2009, when he was demoted to Triple-A but exhibited the same problems with strikeouts, walks, and home runs and was released at his own request. Kobayashi never lived up to those high expectations, though he nearly achieved replacement value on the mound, perhaps damning him with faint praise.
When Bobby Valentine was manager of the 1995 Chiba Lotte Marines, his pitching coach Tom House turned Komiyama around, giving him his first winning season. His 11-4 year also featured a career-best 2.60 ERA and was the first season his K/9 was above 8 and his BB/9 was below 3. With Valentine and the 2002 Mets, Komiyama showed the same control, but his mid-eighties fastball didn’t complement a merely adequate curve well enough. As BP 2003 put it:
The “Greg Maddux of Japan” turned out to be the Greg Maddux of the greater Norfolk metropolitan area, although to be fair, the Mets didn’t really give him an extended opportunity. But when you don’t throw hard, you’re old, and you don’t fool anybody, even your manager’s idiosyncratic affection for Japanese imports peters out.
This was Valentine’s last year with the Mets, and his love of Japanese retreads like Komiyama certainly helped him find the clubhouse door.
Pitching for three different NPB teams, Fukumori drifted from middle relief to closing for the Tohoku Rakuten Golden Eagles, where he recorded 17 saves despite a 4.75 ERA and 1.69 WHIP. His NPB career numbers were similarly unimpressive: a 3.99 ERA, 1.38 WHIP, 6.1 K/9, and 3.3 BB/9. Texas tried Fukumori anyway, but his mix of pitches neither baffled batters nor found the strike zone. He claimed his problems were due to a herniated disc, but no MLB team believed him—a good idea, given his returns.
Milwaukee signed Nomura to be a LOOGY in 2002, ignoring the warning signs of his inconsistency in Japan since 1997. He held MLB lefties to a .185 batting average, but their OPS against him was .823 thanks to his generous walk rate and a .444 SLG. Nomura was cheap, but Milwaukee got what they paid for.
Yabuta had some good years with the Chiba Lotte Marines, but they were late in his career, from 2005 to 2007, when he acted as the team’s setup man as part of the “YFK” relief trio. However, this came after starting, getting bounced to the minors, and generally impressing little. With Kansas City, where he signed a two-year deal, Yabuta had decent strikeout numbers in the minors (his 9.0 K/9 eclipsed his 6.6 NPB average), but his 3.5 BB/9 showed a control problem that followed him to the majors. His changeup and splitter helped him induce ground balls at a 48.3 percent clip, but that didn’t carry over into 2009 when his offspeed stuff no longer fooled batters. Kansas City cut the cord on Yabuta after 2009—unsurprising given his weak performance and elevated price.
Once a very good pitcher in NPB, Kuwata wasn’t the same after an elbow injury that kept him out for nearly two years. After three straight seasons above 8 K/9, he never again posted a season above 6.6. In his final seven NPB seasons, he had an ERA over 4 in six of them and a WHIP above 1.50 in four. The Pirates gave the 39-year-old righty a shot anyway, and he fell flat on his face, walking batters more often than whiffing them. He was cheap, but that doesn’t forgive these negative results.