Quarterly Report: Masahiro Tanaka’s Dominance

Roughly a quarter of the 2014 season is in the books, and the sample sizes are creeping toward a representative level. Over the next couple of weeks, let’s take a somewhat deeper look at some of this season’s more noteworthy players and performances to date. “Noteworthy” doesn’t always mean “best”, though it does in most cases. Let’s kick it off today with a look at Masahiro Tanaka‘s first seven starts as a Yankee. (This article was written prior to his eighth start, on Wednesday night.) His 58/7 K/BB ratio obviously speaks volumes about his ability. Is his future success based almost exclusively on this solid foundation, or is there even more to him?

The raw traditional numbers are pretty impressive – 5-0, 2.57, with that superb K/BB ratio. Poke down just a little further beneath the surface and there’s even more good stuff. The best swinging-strike percentage (14.5%) in the American League. A much higher than league average ground ball rate. He’s the only ERA qualifier in the AL to average seven innings per start to date. It isn’t too early to state that Brian Cashman was in “under-promise but over-deliver” mode when he dubbed Tanaka a “number three starter” upon his signing. Only aces do what Tanaka has done over any seven-start stretch, let alone the first seven starts of their MLB career.

Let’s also consider the context of his performance to date – four of his starts have come at home, in a hitter-friendly yard with a 2013 park factor of 110.0, based on my own calculations utilizing granular batted ball data. According to this method, Yankee Stadium was the fourth most hitter-friendly park in the majors in 2013, behind Colorado, Boston and Milwaukee. It was the sixth most hitter-friendly park for fly balls, and second most for line drives. Two of his three road starts were at even more hitter-friendly parks – Fenway and Miller Park – and the other was at Toronto, versus their power-laden lineup. His home starts have been against the well-regarded offenses of the Orioles, Rays and, well, the Cubs. Overall, he has faced stronger than average clubs in more hitter-friendly than average venues.

It doesn’t take too long to identify a potential Achilles’ heel – vulnerability to the longball. Tanaka has given up as many homers – seven – as walks in the early going, and the one David Ortiz hit off of him is still in orbit.

So what do we have in Masahiro Tanaka? Is he “just” a strike-throwing bat-misser with a home run problem? That wouldn’t be a crime – Fergie Jenkins and Robin Roberts, to name two, rode such a package all the way to the Hall of Fame. Or might he be something even better – a superior K/BB guy who also manages contact very well, like the 2009-13 version of Clayton Kershaw? Let’s take a look at Tanaka’s 2014 plate appearance outcome frequency and production by BIP type data for some hints. Keep in mind that the sample sizes remain small, so most of the contextual information incorporated below is from the 2013 season. No matter – we’re not searching for exactitude here, just looking for some indicators.


FREQ
Tanaka % REL PCT
K 29.9% 149 96
BB 3.6% 42 1
POP 8.2% 105 57
FLY 23.0% 82 8
LD 18.8% 91 12
GB 50.0% 115 90


PROD
Tanaka AVG OBP SLG REL PRD ADJ PRD ACT ERA CALC ERA TRU ERA
FLY 0.357 1.250 237 114
LD 0.696 1.000 117 109
GB 0.262 0.262 113 125
ALL BIP 0.333 0.587 121 97
ALL PA 0.228 0.257 0.402 79 64 2.57 3.18 2.61

First, let’s look at the frequency table. Tanaka’s superior K and BB rates are the headliners here – his K rate is 149 percent of the 2014 MLB average, and in the 96th percentile of MLB pitchers, using 2013 data for context. His BB rate is as low as it gets, in the 1st percentile. His line drive rate is also very low, in the 8th percentile, though that is the most fluid of the frequency figures going forward, a clear regression candidate. His ground ball rate is very high, in the 90th percentile, and is likely real. He manages to pull off the odd combo of a very strong ground ball tendency and a higher than average popup rate, something accomplished by Tim Hudson in 2013. Frequency-wise, it’s a slightly less extremely grounder-focused version of Justin Masterson‘s profile.

Frequency is only one part of the story, however. The second table lists the production from and hints at the authority of the batted balls yielded by Tanaka. The actual production allowed for each BIP type is listed in the “AVG” and “SLG” columns, and is converted into run values, compared to MLB average and scaled to 100 in the “REL PRD” column. Estimates of context, i.e., ballpark, team defense, simple regression and luck are applied in the “ADJ PRD” column in an attempt to isolate Tanaka’s true talent. In the three right-most columns, his actual ERA, his calculated ERA (based on his ADJ PRD) and his “tru” ERA (adjusted for context) are listed. For the purposes of this exercise, HBP are excluded from the OBP calculation, and SH and SF are counted as outs. Again – this is relatively small sample, with much subjectivity in the contextual adjustments, so let’s not get caught up in absolute precision here.

What you see is a guy who has been pummeled in the air to date, to the tune of a .357 AVG-1.250 SLG. We shouldn’t overreact to this, as six of his seven HR allowed have been at Yankee Stadium and Fenway Park, both homer-friendly stadiums, and two of the homers in particular were fairly “soft” homers. The other five were legit and then some, in the case of one of them. It’s fair to say that he has allowed harder than average fly ball contact, though the adjusted for context 114 ADJ PRD feels much more realistic. Actual production allowed on both liners and grounders suggests harder than average batted-ball authority as well. So while batted-ball type frequency, thanks to his ground ball tendency, is a strength, the level of authority within each of those groups has been harder than average. Overall, the grounder tendency prevails, as his overall ADJ PRD – you might call it his adjusted contact score – on all BIP is a league average-ish 97. Add all those K’s and those few BB’s back, however, and you have a “tru” ERA of 2.61, almost exactly the same as his actual ERA.

How does this early estimate of Tanaka’s contact management ability measure up that of the game’s current ruling class of pitchers? Let’s take a look at this same data for the 2013 performances of Yu Darvish, Felix Hernandez and Clayton Kershaw.


FREQ
Darvish % REL PCT
K 34.2% 172 99
BB 9.9% 128 87
POP 10.4% 137 85
FLY 31.2% 111 79
LD 20.1% 94 24
GB 38.3% 89 22
Hernandez % REL PCT
K 27.8% 140 95
BB 5.9% 77 18
POP 5.6% 75 24
FLY 25.1% 89 26
LD 21.6% 101 52
GB 47.7% 111 82
Kershaw % REL PCT
K 26.2% 132 93
BB 5.9% 76 16
POP 8.5% 112 69
FLY 24.0% 86 16
LD 23.2% 108 84
GB 44.3% 103 60


PROD
Darvish AVG OBP SLG REL PRD ADJ PRD ACT ERA CALC ERA TRU ERA
FLY 0.279 0.857 118 116
LD 0.648 0.898 102 109
GB 0.243 0.249 101 120
ALL BIP 0.309 0.536 101 107
ALL PA 0.193 0.272 0.335 72 76 2.83 2.81 2.93
Hernandez AVG OBP SLG REL PRD ADJ PRD ACT ERA CALC ERA TRU ERA
FLY 0.264 0.682 85 101
LD 0.739 0.991 129 105
GB 0.227 0.251 93 101
ALL BIP 0.330 0.488 99 97
ALL PA 0.238 0.281 0.352 78 77 3.04 3.04 2.98
Kershaw AVG OBP SLG REL PRD ADJ PRD ACT ERA CALC ERA TRU ERA
FLY 0.190 0.469 42 57
LD 0.571 0.671 69 87
GB 0.203 0.244 80 97
ALL BIP 0.265 0.376 61 79
ALL PA 0.192 0.239 0.273 53 65 1.83 2.04 2.52

Frequency-wise, two of the three, Hernandez and Kershaw, displayed 2013 K and BB rates somewhat comparable to Tanaka’s 2014 marks. On both the K and especially the BB side, however, Tanaka’s numbers are better, though they were accumulated over seven starts rather than over a full season. Hernandez and Kershaw also showed groundball tendencies in 2013, though again not as strong as Tanaka’s 2014 marks. Overall, Tanaka bests all three frequency-wise.

Authority-wise, it’s a different story. Two of the three, Darvish and Hernandez in this case, were in the same league average-ish neighborhood in 2013 where Tanaka resides in 2014. Darvish’s adjusted contact score of 107 was worse than MLB average, but when you strike out almost 35% of the batters you face, it’s not a big deal. Hernandez’ 2013 adjusted contact score of 97 exactly matches Tanaka’s 2014 mark.

Kershaw is a different animal, however. On top of his exceptional 2013 K and BB rates, he held hitters to a puny .190 AVG-.469 SLG on fly balls – after adjustment for context, that’s an amazing fly ball contact score of 57. He also yielded below average production and authority on liners and grounders, for an overall adjusted contact score of 79, best in the NL last season.

I feel comfortable conclusively stating that Tanaka is not Kershaw when it comes to managing contact. The peak version of Kershaw might never allow a homer as loud as the one Tanaka yielded to Big Papi earlier this season. Based on the limited data available to this point, Tanaka’s contact management ability appears to most closely match that of Felix Hernandez among this small group of elites, and that’s plenty good enough. With such incredible K and BB rates, it’s enough to make Tanaka a legitimate ace and Cy Young Award candidate.

Not to beat a dead horse here, but again – it’s only seven starts. Hideo Nomo struck out 55 batters over 41 innings in his first seven starts, on his way to whiffing an insane 236 batters in 191 1/3 innings as a rookie, before hitters began to gradually figure out his delivery. He settled in as a fine #2-3 starter, but wasn’t the dominant ace he first appeared to be. Fernando Valenzuela won – and COMPLETED – all of his first seven starts, posting a 0.29 ERA with a 61/16 K/BB in those 63 innings. He was very good for a few years afterward, but was never “that guy” again. We just might be seeing the best we will ever see from Tanaka right now, as he remains something new and different, as yet unseen by the majority of major league hitters.

We can certainly say this much about him, though. His splitter just might be the best single offering thrown by any starter in the game today. He has an amazing 25.7% whiff rate on it – over a quarter of the splitters he has thrown have resulted in a swing and a miss. His slider is a second viable whiff pitch (12.6%). He has utterly dominated the opposite hand, yielding a paltry .207-.225-.310 line to lefties to date.

Most of all, he just pumps strikes. Swings and misses with multiple weapons, a true, bona fide out pitch, a mastery of opposite-handed hitters – this is the stuff of which dominant starters are made. It’s early, but if he can stay healthy – a massive “if” for any pitcher these days – the Yanks should be able to chalk up this signing as a major win.




Print This Post



32 Responses to “Quarterly Report: Masahiro Tanaka’s Dominance”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. Aaron (UK) says:

    (This article was written prior to his eighth start, on Wednesday night.)

    Well a shutout with 4H, 0BB and 8K isn’t going to hurt too much.

    +20 Vote -1 Vote +1

  2. Kevin Towers says:

    Is he better than Yu Darvish? How close are they?

    Vote -1 Vote +1

    • JimNYC says:

      I think it’s too early to say, yet. We need more data on Tanaka to see if that 16.7% HR/FB is legit or just sample-size noise. I think it’s fair to say at this point that Tanaka’s K/BB is going to be FAR better than Darvish’s; that Tanaka is going to have a better strand rate because of his tendency to increase his usage of the splitter with men on base (ESPN today say that with men in scoring position and two strikes he’s thrown splitters 53% of the time); that Tanaka’s going to have a better GB rate; and that Darvish is going to have a better K rate.

      The question becomes whether Tanaka is going to stay so susceptible to the long ball; if so, he’s probably not going to be quite as good as Darvish. If not, he’s probably going to be the best pitcher in the American League.

      All of this, of course, is barring arm injury. He’s thrown over 100 pitches in every start but his first, and he’s averaging almost 107 pitches per game. That splitter puts a lot of strain on his forearm ligaments, so it’s hard to say how he’ll hold up over the long haul.

      Vote -1 Vote +1

      • Swfcdan says:

        He’s been a horse in Japan, so we can’t have any worry about the splitter cmon. The question is how he holds up to the overall MLB workload come the late summer months.

        But he’s has adapted to the US brilliantly, as has Abreu. These international signings really are the real deal these days, no more Iwamura’s or Dice-K’s/Igawa’s anymore.

        Vote -1 Vote +1

  3. Andy says:

    As always, thanks for the excellent analysis.

    His GB rate is only 115% that of the league average, yet it puts him in the 90th (actually you say 92d) percentile. Likewise, his FB and LD rates are 80-90% of league average, yet around in the 10 percentile. From which it seems to follow that there is a lot of bunching, with not that much difference in rates among a lot of pitchers, and small differences in rates resulting in large differences in percentile. This is even more evident from examination of the data of the other pitchers. E.g., Kershaw has a GB rate at just 103%, but a percentile of 60.

    So wouldn’t one expect to see a lot of movement in the percentile figures as more data accumulate? You conclude that the GB rate has probably more or less stabilized, but it seems even a relatively small change in it could have a big impact on percentile. Suppose it dropped to 48%, not that big a change, which I take it would be 110% of league average (assuming that is stable by now). I’m guessing (based on the data for the other three pitchers) that would drop him down to about the 70th – 80th percentile, still very good but not in an elite class. And the same argument with FB and LD (which you yourself say has not stabilized) going the other way.

    Also, I don’t understand how his adjusted production is > 100 on all BIP types, yet slightly below 100 on overall BIP. The same situation exists for Darvish and Hernandez, though not for Kershaw. That seems to imply that most pitchers have a very high adjusted production on one BIP type, so that someone with a fairly low but still > average rate on all three can end up below average for the total?

    Vote -1 Vote +1

    • Rex Manning Day says:

      Regarding your last point, I believe the answer lies in the BIP mix that each player allows.

      Within each BIP type, Tanaka allows a higher production than average, but his mix of BIP types is not average. He allows far more GB than average and far fewer FB and LD. Because GB are the least-productive BIP type, this makes Tanaka’s overall BIP production lower than average.

      If Tanaka gave up an average number of GB, FB, and LD, then his overall production would be higher. But he does not. Instead, almost half of his BIPs are GBs, and you can only get so burned by GBs.

      To make an exaggerated example, imagine if every one of Tanaka’s FBs went for a home run, but he only gave up 5 FBs in a full season. His production on FB would be way over average because even the most homer-prone pitcher doesn’t give up a homer on every fly ball, but his overall production would be far below average because he would basically never give up a FB in the first place.

      Vote -1 Vote +1

    • JimNYC says:

      Your last point reminds me of my favorite anecdote from my first semester statistics class in college:

      In both 1995 and 1996, David Justice had a better batting average than Derek Jeter. In 1995, Justice hit .253 while Jeter hit .250; in 1996, Justice hit .321 while Jeter hit .314. However, Jeter had BY FAR the better batting average for 1995 and 1996 combined — Jeter hit .310 in the two years combined, while Justice hit .270.

      The reason being, of course, that Justice had 411 AB’s in 1995, and 140 in 1996; Jeter had 48 AB’s in ’95 and 582 in ’96.

      Here, Tanaka might have below average production on all BIP types, but since GB’s give far better production than line drives or fly balls, if he has a higher GB rate than other pitchers, his production will be better than theirs.

      Vote -1 Vote +1

  4. Tom B says:

    You could have just posted a picture of yourself watching the Ortiz home-run and save yourself 500 words. All googly-eyed and in love…

    -6 Vote -1 Vote +1

  5. Swfcdan says:

    He’s one of the few (only?) things Yankee fans have to look forward to watching nowadays!

    Vote -1 Vote +1

    • JimNYC says:

      Not true. I’ve already purchased a Yangervis Solarte jersey.

      Vote -1 Vote +1

      • Swfcdan says:

        Solarte’s a nice story, but will he continue? Hope so, not convinced though.

        Vote -1 Vote +1

        • JimNYC says:

          I’m not convinced either, of course — the history of 26 year old rookies maintaining status as impact players isn’t particularly a stellar one — but his BABIP is only .353, so it’s not like it’s a complete fluke… I hope.

          Vote -1 Vote +1

  6. vivalajeter says:

    It’ll be interesting to see how he performs not only the second time through the league, but at the end of this year and into next year. I was watching the highly-rated Mets broadcast last, rather than the awful Yankees broadcast, and Ron Darling obviously spent a lot of time discussing the splitter. From what he was saying, a lot of teams simply don’t allow their pitchers to throw a splitter so hitters rarely see it anymore. The vast majority of major league hitters may not have seen a legit splitter in years (or ever). Once they get more video on it, and get to see it in person more times, will it be quite as effective?

    It has terrific movement and seems to just drop off of a cliff, so if you’re expecting a fastball then you really have no shot at making contact. But as time goes on, will they be able to pick up the pitch a little better so they can see it coming before it’s too late? That’s where the rubber meets the road.

    Vote -1 Vote +1

    • JimNYC says:

      It worked for Curt Schilling and Roger Clemens for like twenty years each.

      Vote -1 Vote +1

    • Belloc says:

      If what you are saying is true, then Roger Clemens should have become a mediocre pitcher by 1987, and he should have been chased out of the league shortly thereafter. Curt Schilling should have been chased out of the league in the early 90s.

      Apparently Major League hitters who played in the 1990s and 2000s were absolutely dreadful since they never could solve a pitch that they saw over and over again.

      Vote -1 Vote +1

      • vivalajeter says:

        At what point did I say that Tanaka will be mediocre or be chased out of the league? There’s an enormous gap between regressing from where he is right now – potentially the best pitcher in the league – and being out of baseball.

        Vote -1 Vote +1

    • Bo Knows says:

      The thing is that Tanaka’s splitter comes from the exact same arm slot, with the exact same arm action and speed. It also travels the exact same trajectory and arc as his fastball for most of the way to plate before deviating. The point at which the split starts to deviate is said to be past the point the human eye is capable seeing the ball. So chances are the only way his splitter gets crushed is if it doesn’t move much and just hangs.

      Vote -1 Vote +1

  7. Hurtlockertwo says:

    Don’t pitchers only pitch once every 6-7 games in Japan? Is his workload now going to be problem in Aug/Sept?

    Vote -1 Vote +1

    • Bearman says:

      If u think of a 30% bump in workload that’s “supposed” to be the magic number, he should be good to go from 26 or 28 starts to 32. He’s thrown more than 200 innings before and because he limits walks he’s not stressing on baserunners.

      But then elbows exist so nothing is certain…

      Vote -1 Vote +1

    • Stank Asten says:

      He threw like 200 innings per season in a league that plays 144 games.

      Vote -1 Vote +1

  8. Helladecimal says:

    Great article.

    Vote -1 Vote +1

  9. pft says:

    Interesting article.

    Couple of notes.

    Tanaka has been unhittable vs LHB’ers, not so vs RHB’ers.

    His FB is getting crushed when batters put it in play, both the 2 seamer and 4 seamer, usually when he leaves it up in the zone, which is often. not to say the FB does not get some good results as it sets up other pitches

    His slider and split have been pretty unhittable. Tanaka is throwing the split far more often than in Japan and oone wonders how this plays out over time as pitchers can lose the split throwing it too much (cartilage loosens)

    One wonders how he fares when hitters figure out to take the same approach as with knuckleballers, if its low let it go, if its high let it fly. If a pitch looks as if its coming in at belt level, its probably going to be out of the zone (slider or split) or a tough pitch to hit in the zone. If it comes in higher its either a hanger or fastball. Sounds easy when you are at a keyboard

    Vote -1 Vote +1

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Current day month ye@r *