Hey do you guys remember the salad days of Spring Training? Back when Greg Bird was batting .451/.556/1.098? And Masahiro Tanaka had a 0 ERA? Those were good times.

Here are Tanaka’s current numbers:

6.34 ERA, 5.18 FIP, 4.08 xFIP, 4.03 SIERA, 4.78 DRA. A whopping 2.07 HR/9, good for 4th worst in the sport, behind only Bronson Arroyo, Ricky Nolasco, and Jordan Zimmermann.

Yeesh. Not very good company.

Frustration, thy name is Tanaka

(Adam Hunger/USA TODAY Sports)

But even despite all of that, I’m not ready to give up on Tanaka. He did pitch a Maddux against Boston in Boston, and then chucked 7.1 innings of 1 ER, 13 K ball against Oakland at home. His K% hasn’t changed at all (literally, it’s 20.5% this year and last) and his BB% is only up two points to 6.2%, which puts him at #21 overall by that measure, right behind Max Scherzer. (Better company.) His O-Swing and Swing% are unchanged as well. Z-Swing% is up three points, but Z-Contact% is actually down by three, which I would actually say is a sign that his deception on the good pitches is intact.

So what’s wrong then?

The biggest thing that jumps out when you look at Tanaka’s stats is the pitch mix compared to last year:

Type20162017Diff
Fastball6.5%20.8%14.3%
Sinker21.1%24.5%3.4%
Curveball5.5%5.9%0.4%
Slider27.5%26.8%-0.7%
Cutter9.2%2.1%-7.1%
Splitter30.2%20.0%-10.2%

Most notably, he’s stopped throwing the splitter so much, and put the difference (and then some) into his four-seam fastball. Now, the four-seamer is actually why I buy there not being an underlying injury (at least not to the elbow,) because it’s clocking in at 97 mph this year. But the problem is that the pitch simply isn’t good. It’s never actually been very good: the closest the pitch has come to positive value according to FanGraphs’ PITCHf/x values /100 was in 2016, where it was still -0.36.

But the bigger issue for Tanaka this year is that none of his pitches are very good according to that metric:

Type20162017Diff
Fastball-0.36-2.67-2.31
Sinker0.79-3.36-4.15
Curveball-0.790.311.10
Slider0.370.680.31
Cutter-0.92-2.23-1.31
Splitter2.27-0.25-2.52

It is worth noting that the values/100 data is descriptive, not predictive, and we’re still dealing with a relatively small sample size. But this means the fix won’t be found simply by adjusting the pitch mix to come more into line with 2016’s numbers.

OK, but what do the tables mean?

The biggest issue that Tanaka simply isn’t executing good pitches. From an unnamed pro scout that watched him in Tampa this May, “He isn’t finishing his pitches, and he’s making mistakes with the fastball.”

To put it visually, here are two horrific mistake pitches thrown to Astros batters from Derek Jeter Night:

(Both videos come from Travis Sawchik’s article on this very subject which also includes heat maps and stuff. Check it out.)

That’s a sinker and splitter respectively, and they both do…nothing. You can see Gary Sanchez drop his mitt below the zone, then have to come way back up to try and receive the pitch. Look at his whole body language response to the splitter. He’s seeing that total lack of drop out of the hand, and so are batters. Again, 2.07 HR/9.

So wait, where’s the upside?

Well, like I said near the beginning: Tanaka’s K%, BB%, and various -Swing%s are pretty much right in line with his career numbers. This isn’t like what’s happening to Matt Harvey, whose walk rate has doubled and O-Swing% has dropped a full ten points (not to mention the major surgery last year.)

Tanaka seems to be just making some kind of mechanical mistake. It’s likely something small and subtle, or else it would already be fixed. At this point, even if he gets it together for another shutout against the Sox tomorrow and keeps it going, Tanaka will be working off these earned runs until September. But Tanaka and Rothschild know each other quite well by now, I would hope. If there’s a fix to be found, I trust them to get there. Hopefully sooner rather than later.