What a great GIF.

The third post ever posted to Views from 314 ft. was by Derek, called “Masahiro Tanaka’s missing splitter.” As you can imagine, it dealt with an unusual prospect: Tanaka, who has always relied on his trademark pitch to miss bats and get outs, simply couldn’t throw the splitter this year. He was open about it at the time, and that trend continued until August, when James Wagner of The New York Times reported that the new baseball’s lower seams made it more difficult for him to grip the pitch. A change was in order.

In fact, he’d already made one. A few days earlier, on the trade deadline, Dan Martin of the New York Post reported that Tanaka had made a “significant adjustment” to the pitch. The early returns were promising, too. As I noted here, Tanaka threw 28 splitters with 16 swings and 6 whiffs on deadline day, which is good for a 37% whiff-per-swing rate. It seems that trend is continuing, with Tanaka throwing 38 splitters in his last start and 31 in his prior start.

I think that’s worth diving into it, so with that context in mind, let’s get right to it, shall we?

Early Woes

It’s worth starting by covering just how bad the pitch was early on in the season. I looked into the data, stopping it right before his July 31 start against Arizona–the last start before, remember, Tanaka was tagged for 12 runs in a grotesque loss in Fenway Park. Here are some of the key numbers, with last year’s in parentheses:

  • Pitch Usage: 25.69% (31.52%)
  • Swing Percentage: 61.12% (63.14%)
  • Whiffs-Per-Swing: 17.31% (36.40%)
  • Batting Average Against: .284 (.220)
  • Slugging Against: .490 (.344)
  • Isolated Power: .207 (.125)
  • Vertical Movement: -26.79 inches (-29.40 inches)

There are a lot of numbers there, but it tells a pretty clear story. Tanaka’s trademark pitch, which he threw about a third of the time in 2018, declined in usage, attracted fewer swings (a sign the pitches were further out of the zone), and generated way fewer whiffs. The pitch was moving less, too. Add that all up, and you get more contact. Adding to the woes was the fact that the contact batters made was louder, too. There was a significant jump in power against it. Bad news all around.

As Derek noted back in May, a lot of that stemmed from the fact that Tanaka wasn’t locating the pitch in the same way. A greater percentage of splitters were left up and over the plate, which is obviously a bad recipe. All that meant that Tanaka started to use the pitch significantly less, which makes sense. Check out the usage graph on a month-by-month basis:

A slight increase from May to June, but then a nosedive after that. Through July 26, the day after his shellacking in Fenway, Tanaka had used his trademark pitch just 22% of the time in July. Making matters worse, the pitch had less downward movement than it had in year’s past. That meant that it wasn’t just that batters were seeing it better or anything like that–it meant that, overall, the pitch simply wasn’t as good.

The decline of the pitch, coupled with some atrocious starts, meant that he was sitting with a 4.79 ERA (4.56 FIP) in 120 IP, with just a 20.1% strikeout rate. He’d allowed more than 9 hits per 9 despite limiting walks, but this was not the Tanaka we all expected to see. It was a worrying trend.

The Adjustment

Mike Axisa, over at his Patreon, highlighted the difference in the grip that Martin highlighted at the post. It’s a stark difference. Check it out. Here’s the before:

That’s a pretty traditional splitter grip–the type of grip you would teach a young athlete learning to throw the pitch for the first time. No surprises to me that this was his preferred grip.

Here’s the pitch after the adjustment:

I’m no pitching expert by any means, but that is a noticeable difference. The grip is completely changed. It’s pretty amazing to think of what a seemingly subtle change to the baseball can do to players. At least it is to me.

Anyway, none of this is breaking new ground. Like I said, Mike pointed this out a while ago and it’s been a pretty standard talking point for a while. It still helps to see it so clearly, though.

The Results

So has it been working? The short answer is yes, even with the obvious short sample size caveat. It has only been a month. 42.2 innings in 7 games, to be exact. Again, that’s not a lot to play with. But you know what? It’s more than you think. Those 42.2 innings make up just about 25% of all of the innings Tanaka has thrown this year. It’s getting to the point where the sample will be at least noteworthy, in my opinion.

But anyway. Let’s dive into the data again, shall we? Here are the same metrics as above, this time with his stats from July 31 on. The numbers in parentheses are the same figures as his pre-deadline stats I highlighted above:

  • Pitch Usage: 34.38% (25.69%)
  • Swing Percentage: 64.35% (61.12%)
  • Whiffs-Per-Swing: 20.95% (17.31%)
  • Batting Average Against: .239 (.284)
  • Slugging Against: .310 (.490)
  • Isolated Power: .070 (.207)
  • Vertical Movement: -28.47 inches (-26.79 inches)

Again, that’s a lot of data, but the story is clear again. First things first: the pitch is significantly better since the tweak. Significantly better. Batters are swinging more, swinging and missing a greater percentage of the time, and overall inducing softer contact against the pitch. I mean, check out the drop in slugging and ISO. That’s huge.

He’s also clearly much more confident in the pitch, throwing it considerably more now than earlier in the season. In fact, he’s using the pitch even more recently than he did last year. Check out the usage graph, this time month-by-month for all of 2018 and 2019, and including all of his pitches:

Seems like things have normalized a bit to me. Aside from the massive drop in fastball usage last October, his three main weapons–fastball, splitter, and slider–are all being used the same now this year as they were last. Encouraging signs abound.

But the most encouraging sign of all is the vertical drop–he’s added about 2 inches of vertical drop to the splitter since the tweak. It’s difficult to overstate just how important that it. In fact, I’m pretty confident that the added movement is the major reason for the improvement.

Now, it’s also worth noting two things. First, the pitch still isn’t generating that many whiffs-per-swing. It’s still down quite a bit from 2018, actually. Second, despite the added movement, it’s still breaking about an inch less than it did last year.

I’m not too concerned, though. Both figures have improved since the adjustment, and every other metric has improved tremendously. The pitch may not be all the way back, but it’s on the path–and it’s already effective again, which is more than we could say in June. Not to mention, Tanaka has a 3.88 ERA (3.69 FIP) since the tweak, with a few real gems tossed in there. So hooray for that.

Looking Ahead

It’s hard to say what this all means going forward, but I think there’s reason to be optimistic. First, there was a real change. This isn’t smoke and mirrors. The ball was changed and Tanaka adjusted with it. Furthermore, we can pinpoint the date of the tweak and track the results. Causation, correlation, and all that, but in this case, I’m pretty sure that the correlation does imply causation here.

Second, this has always been his best pitch. The new trends are a return to normal, not a break from the past. What was unusual was the fact that the pitch was bad in the first place. That, too, is reason to be optimistic.

It goes without saying that Tanaka, Paxton’s surge or not, is critical to the Yankees postseason hopes this year. He’s always been a playoff warrior (he has a 1.50 ERA in 30 playoff innings across 4 series), but that only takes you so far. I’d feel so much more confident in Tanaka knowing that his trademark pitch is working again. My guess is that Tanaka, and the Yankees organization, feel similarly.