A Starter Or A Reliever? A Second Look At Hisanori Takahashi Through PitchFX

In my first article here at Amazin' Avenue, I wrote about the surprisingly strong performance of Hisanori Takahashi. At that time, Takahashi had just turned out some great long-relief work and two terrific starts. Having looked at his stuff through pitchf/x, I concluded the following:

Takahashi's statistics are amazing against left-handed batters and decent against right-handed batters, as seen by the following chart:










vs LHB









vs RHB


















Takahashi's fastball is basically mediocre in terms of movement and speed. Against left-handed batters Takahashi throws the pitch only on the outside part of the plate, and is reasonably accurate. Against right-handed batters, Takahashi uses both sides of the plate but is far more inaccurate. In order to have success against right-handed batters, Takahashi needs to improve his accuracy against these batters.

Takahashi's changeup is his best pitch and is more or less exclusively used against right-handed batters. The pitch has an amazing swinging strike rate of 17.5% and is his out pitch. He ONLY throws the pitch away and low, to the point where he misses the strike zone extremely frequently. There is a worry that batters will wise up to his usage of the pitch and start to let these pitches go instead of swinging at them. If that's the case, his walk rate will increase and his effectiveness against right-handed batters will drop.

Takahashi also has a slider and a curveball. The slider is used instead of the changeup as the off speed pitch of choice against left-handed batters and it's fairly effective. Takahashi's curveball is his worst pitch and is used mostly against right-handed batters. He does not use the pitch on one ball counts and beyond very much at all.

Since that time, Takahashi has pitched 61 more innings. His statistics have of course dropped as he made the change from starter to reliever. The following table shows his new total statistics split by batter-handedness.










vs LHB









vs RHB


















As you might have expected, the extremes seen in Takahashi's original small sample size have almost completely disappeared. His strikeout rate is down from being a batter per inning or better, but is still fairly good. His walk rate against lefties has gone down, but his walk rate against right-handed batters has dropped. The only extreme number to still exist is his 0 HR/9 against LHBs though it is notable that his sample size against LHBs still is very small (as a starter, lineups are loaded with right-handed batters against him).

Still, Takahashi retains his amazing numbers against left-handed batters that he showed on May 31st. His performance against right-handed batters has suffered, though not as much as his ERA would suggest if we take the advanced metrics (FIP and xFIP) at face value. What's changed? What's remained the same? And what has improved? These are the questions this article will attempt to answer.


Takahashi's Fastball:



Reading the above graphs:

The above graphs show the location in the strike zone each pitch crossed the plate from a catcher's point of view. Thus a pitch on the left of the graph went in on a right-handed batter and away from a left-handed batter. The square in the middle of the graph represents an approximation of the strike zone.

The top graph shows Takahashi's fastball against left-handed batters. The bottom graph shows Takahashi's fastball against right-handed batters.

Pitch Type Batter Handedness # Thrown Run Value Above Average (RVAA) RVAAe Whiff% Swing % Swinging Strike % In Strike Zone % Ground Ball % Called Strike % SLGCON
FF L 216 0.0616 -2.7602 10.47% 39.81% 4.17% 68.98% 38.24% 29.17% 0.618
FF R 550 3.161 0.0545 12.33% 41.27% 5.09% 62.00% 25.84% 22.18% 0.719

Reading the above table:

Run Value Above Average: The total runs saved or given up above that of the average pitcher on that pitch. A negative value is GOOD (it means he saved more runs than average) while a positive value is BAD.

RVAAe: The run value above average for a pitch controlling for bad luck on balls in play and for an average HR/FB rate. In other words, this is the EXPECTED run value for that pitch.

Whiff %: Percentage of pitches that batters have swung at resulting in swinging strikes.

Swing %: Percentage of pitches that batters have swung at.

Swinging Strike %: Percentage of pitches that have resulted in swinging strikes.

In Strike Zone%: Percentage of pitches that have been in an approximation of the strike zone (2 foot plate).

Ground Ball %: Percentage of balls hit in play off a pitch type that resulted in ground balls.

Called Strike %: Percentage of pitches of a type taken for a called strike.

SLGCON: Slugging percentage on balls put in play; In other words, this is the total bases gained per ball put in play.

As we saw before, against left-handed batters Takahashi aims his fastball at the outside and low corner. He's also decently successful at hitting the strike zone, nailing it roughly 69% of the time. Batters don't really swing and miss at the pitch, but they do take it for a called strike 29.17% of the time. The end result is that the pitch is an average pitch, which is able to set up his "out pitch" when he's got the hitter in a good count (more on this in a moment).

Against right-handed batters, Takahashi's fastball has not been very successful. He obviously aims the pitch at both the inside and outside part of the plate, but, well, he isn't very accurate, hitting the zone only 62% of the time (a very low amount for a fastball). As a result, the pitch gets taken for a called strike only 22.18% of the time; 7% LESS than against LHBs. The pitch isn't getting swinging strikes either and most importantly is an EXTREME fly ball pitch (25.84% GB Rate). When batters have hit the pitch (and they've done so often), they've hit it hard (seven of the 12 HRs he's given up have come on the pitch).

Now it's important to realize, a pitcher can't simply stop throwing a fastball. And Takahashi already throws the fastball less than all but 13 pitchers in the majors who have thrown 800 pitches this year (and two of those others are knuckleballers). So I'm not sure he can really reduce the amount he throws this fastball. What he might want to do is pick a side of the plate against RHBs he CAN hit the strike zone on (probably away) and just aim there all the time, like he does against LHBs.

Takahashi's Changeup:


Pitch Type Batter Handedness # Thrown Run Value Above Average (RVAA) RVAAe Whiff% Swing % Swinging Strike % In Strike Zone % Ground Ball % Called Strike % SLGCON
CH L 38 -1.8168 -2.1055 57.14% 55.26% 31.58% 31.58% 100.00% 2.63% 0.333
CH R 502 -8.7589 -7.4499 36.93% 57.17% 21.12% 44.42% 46.24% 9.96% 0.441

As a starter, Takahashi uses his changeup just as he did as a reliever: As an out pitch against RHBs. And it's a very good pitch, no matter what measure you use to describe it. Back on May 31st, he had a 17.5% swinging strike rate on the pitch. Since then, this rate has gone UP to 21.12%!! The changeup may be the best pitch at getting swinging strikes (it is) but this is still amazing. What's even more impressive is that Takahashi only gets the pitch in the strike zone 44% of the time against RHBs! And yet, despite this being about the same as it was back when I wrote my last article batters STILL cannot lay off. And of course, because the pitch is always either away and low or way out of the strike zone, the pitch is his best ground ball pitch by far (46%) and is rarely hit for much power (SLGCON of .441).

Takahashi has gotten worse against RHBs, but that is certainly not because of his changeup. Instead, the pitch has seemingly improved as a starter.

Takahashi's Slider:



Pitch Type Batter Handedness # Thrown Run Value Above Average (RVAA) RVAAe Whiff% Swing % Swinging Strike % In Strike Zone % Ground Ball % Called Strike % SLGCON
SL L 125 -3.551 -1.0838 23.21% 44.80% 10.40% 61.60% 34.78% 18.40% 0.217
SL R 79 3.1078 0.7242 23.26% 54.43% 12.66% 65.82% 21.43% 13.92% 1.071

The big thing to talk about now is Takahashi's slider usage. Back in May, when Takahashi was a reliever and only had two starts under his belt, the slider was only used as a replacement for the changeup against LHBs. And it was pretty good against these batters. But since then, Takahashi has also started to use the slider a bit more against right handed batters. And it's been a real failure.

Against LHBs, Takahashi uses the slider to hit the outside corner (just like his change-up to RHBs) and is very successful (per 100 pitches it's even more successful than the change-up is against RHBs). The pitch has a decent (but not great) swinging strike rate of 10.40%, but gets most of its value from the fact that batters simply cannot hit the ball hard at all (SLGCON of .217). To put this into perspective, for every five sliders that left-handed batters put into play, they manage to get only one single. Now I can't expect such an extreme result to continue (thus why there is the discrepancy between the slider's RVAA and RVAAe), but I don't express this to regress much. It is the SLIDER which has been part of the reason why Takahashi is simply death to lefties this year.

But against RHBs this is simply a different story. Takahashi's location with this pitch has been mostly in the middle of the plate and he has been hit HARD. His SLGCON for this pitch has been a whopping 1.071. To put that into perspective, for every slider that Takahashi has thrown to a right-handed batter that has been put into play, the batter has gotten essentially at least a single out of it. Now, once again this is an extreme result which we should expect to come down. But with the control seen on the chart above, it shouldn't regress too much. And meanwhile, Takahashi doesn't get enough swinging strikes (a decent but not great 12.66%) or ground balls (a miserable 21.43%) to justify using this pitch so much (6.3% of the time). And to make things worse, Takahashi has INCREASED his usage of this pitch over the last two months at the expense of his usage of his curveball and fastball. This would seem to be a pretty big mistake.

Takahashi's Curveball:


Pitch Type Batter Handedness # Thrown Run Value Above Average (RVAA) RVAAe Whiff% Swing % Swinging Strike % In Strike Zone % Ground Ball % Called Strike % SLGCON
CU L 18 -0.471 -0.3139 28.57% 38.89% 11.11% 50.00% 100.00% 11.11% 0.000
CU R 106 0.5156 0.2332 24.24% 31.13% 7.55% 58.49% 30.77% 25.47% 0.615

Takahashi's other breaking pitch is a curveball that he uses near-exclusively against right-handed batters. As I said in my previous article, Takahashi seems to have little confidence in this pitch as he uses it more-or-less only in favorable counts. On 0,0 counts, he uses it 14.4% of the time; on 0,1 counts 14.4% of the time, on 0,2 counts 11.5% of the time. On no other ball-strike counts does Takahashi throw the pitch even 7% of the time. On three ball counts, he NEVER throws the curveball.

And yet, despite this lack of faith in the pitch, it's not terrible. Yes it IS true that he does only get roughly five inches of drop (once gravity is factored out) on the pitch and he does hit the middle of the plate a little more than he should. However, his accuracy with the pitch isn't terrible (in the zone 58.49% of the time) and he manages to frame the outside corner of the plate better than he does with the slider to right-handed batters (but worse than the change-up). More importantly, batters take the pitch for a called strike a strong 25.47% of the time (note that this rate is bumped up due to the pitch being thrown most often on 0,0 counts) and they haven't hit it anywhere near as hard as they have the slider (SLGCON of .615). If Takahashi believes he needs a breaking pitch to augment his change-up and fastball, THIS is the pitch he should focus on against RHBs.

How should Takahashi be used going forward? As a starter or reliever?

It looks like Jerry is planning on using Takahashi as a reliever full time, perhaps as an "eighth inning guy". That would be a mistake. Takahashi clearly has the tools for dealing with left-handed batters and also DOES HAVE THE TOOL to handle right-handed batters: his change-up. If he was to abandon his slider against RHBs, replacing it with mostly change-ups and some curveballs, he should be able to be reasonably successful against these guys. And I think that a great anti-LHB pitcher who is decent pitcher to RHBs can easily suffice as a #5 starter, if not a #4. And really, do we have confidence in anyone else (Pat Misch?) to be the #5 starter right now?

Now, the argument for using Takahashi as a reliever is two-fold: 1) With it, he can reduce the number of non-fastball/non-change-up pitches he throws and thus be more effective and 2) As a reliever, opposing managers can't stack the lineup with right-handed batters (unless they pinch hit, reducing their later options). The second reason is indeed valid and is a major reason why Takahashi should be more effective as a reliever. But the first reason I don't think is true; I don't see why, as a starter, Takahashi cannot simply switch to using two pitches against LHBs (fastball/slider) and two against RHBs (fastball/changeup).

And REMEMBER, a starter is more valuable than a reliever. And Takahashi has the stuff to be a valuable piece of the rotation this year. Sure, the team might decide to use him out of the bullpen during his bullpen-session days (maybe to give Pedro Feliciano a day off here or there), but he can be most effectively used as a starter for this team.


Regardless of how Takahashi is used, he DOES have the tools to succeed. The key is for him to use these pitches more effectively. Against RHBs, he should simply DITCH the slider entirely. He hasn't shown great control over the pitch and he has another breaking ball over which he's shown greater control. Moreover, he has the changeup which is especially deadly against right-handed batters. So he can easily replace the slider with say 2/3 change-ups and 1/3 curve balls and it would seem very likely that he would improve. He should certainly stop INCREASING his usage of the slider against these batters.

Against LHBs, Takahashi should simply continue what he's doing (more or less all fastballs and sliders). It could hardly be more effective. And if he switches his pitch-selection to RHBs, there's no reason why he can't be fairly effective against them as well. He'll still have his weaknesses: for one, he still hasn't shown that he can accurately aim his fastball against RHBs enough to make it an average pitch. That said, by relying on his changeup and reducing the usage of his slider against these batters, he can compensate for this weakness.

Garik16, AKA Josh Smolow, can be contacted by twitter under the original name: garik16. Okay it's not very original. Anyhow, if you have any PitchFX Met Questions unrelated to this article, feel free to send me a tweet.

This FanPost was contributed by a member of the community and was not subject to any vetting or approval process.