Tanner Roark and the southpaw struggles

Is it fair to say that 2015 did not go exactly as the Washington Nationals had hoped? OK, maybe that is best served as a rhetorical question. I mean, come on. ‘Well, Shawn, did you even watch the Six Flags-style charade the fighting Matt Williams (Williamses?) displayed this season?’ would be the correct way to answer such a profoundly obvious rhetorical (side note: excuse me while I mentally compare the similarities of Matt Williams and the dancing Six Flags man). In short, I sorta did?

If by ‘sorta’ you mean watch a team crush your soul whilst maintaining a decent deficit behind a team from New York and running a guy out to left field who is woefully ill-equipped defensively to do such a thing. Should that be the case then yes, yes I did. The team I watched just happens to be over 400 miles away from the team you think I watched.

Anyway, much like the Nationals, one pitcher on their team in particular did not have quite the season they hoped he would. If you can remember back long ago to the year 2014, Tanner Roark was spectacular for the Nationals. How good? Well, in his first full MLB season Roark threw 198.2 innings and made 31 starts for a NL East-winning Nationals team. In those innings, the former 25^th round draft pick posted a 2.85 ERA and 3.84 xFIP—walking less than five percent of the batters he faced and striking out 17.1 percent. That was on his way to a noteworthy 3.1 fWAR campaign, which was good enough for 16th out of the 47 NL pitchers who threw more than 150 innings. So Roark was primed for a breakout onto the national scene—no pun intended…ok pun intended—in 2015, right? Well, not so much.

Before the season even started, the signing of Scherzer pushed Roark into the uncoveted role of being the swing-man—a 6^th man of a 5-man rotation. The righty would still make 12 starts in 2015, throwing significantly fewer innings (only 111 innings, to be exact). Combine that with an ERA, FIP, K%, BB%, and fWAR that were all worse than their predecessors, and it doesn’t take much to infer that this season had a less-than-desirable result for the 29 year old.

The headline to this article can largely summarize why Tanner Roark struggled in 2015—not to mention that it also has the added bonus of sounding like the name of a late 20^th century band. So while you fire up some Tom Petty and the Heartbreakers, Hootie and the Blowfish, or Huey Lewis and the News, I’ll toss you some Roark splits:

Left to out to dry

	TBF	K%	BB%	HR/FB%	wOBA	BABIP	FIP	Soft%	Med%	Hard%
2014 Roark	410	13.2%	6.8%	7.8%	.299	.248	4.38	18.1%	59.2%	22.7%
2015 Roark	225	12.4%	7.1%	14.3%	.369	.304	5.49	20.5%	48.3%	31.3%

(If ‘Free Falling’ was the song you went with from the three bands listed above, then congratulations. You can claim your "I have above-average timing" shirt when Ian Kennedy signs his next contract—here’s a hint, the shirt is a tank top since that is a very popular type of shirt for summertime weather.)

So there is a distinct difference in how Roark has fared against left-handed hitters, but it took a jump in a not-so-good direction this season.

Certainly a right-handed pitcher struggling against lefties is nothing new to the world of baseball. However, the extent of the platoon split is different for each individual pitcher. As FanGraphs notes, a pitcher's platoon splits stabilize* between 500-700 plate appearances against both hands—meaning Roark recently passed the threshold for stabilization.

*Meaning there's more signal than noise, not necessarily a magic switch point. There is still noise.

Will he stabilize against lefties at his current level of performance? Maybe. Although he has passed the stabilization point, coming out and saying ‘of course Roark is going to continue struggling mightily against lefties for the rest of his career' would not be exactly true. As would saying ‘of course Roark is going to improve greatly against lefties and begin to dominate his opposite handed colleagues’. It is not because I think that there is a flaw in any of the statistics that suggest he very well may. No, my reasoning is that there has been a flaw in the way Roark has approached left-handed hitters that began to impact him negatively in 2015.

To give you some context, check out this year-and-a-half-old interview Roark had with FanGraphs’ Eno Sarris. In the sit down, Sarris talked with Roark about a plethora of things. See if you can decipher what the righty's plan of attack was against lefties:

[…]When he’s faced with lefties, Roark knows he’s trying to get it into the hitter "so that I don’t start it somewhere where I leave it right over down the middle where they can rip it." The solution? "Have to start it pretty much on the front hip and make them get them out of the way and then have it run back over. You have to make sure you get it there, or else it’s going to get hit hard," Roark admitted.

[…]For the slider, the work is situational. Roark is working to find ways to use the pitch against lefties. "I’ve just got to bury it in the back foot," for the most part, but he’s also been trying to work on going front door to lefties. He wants them to "give up on it on the outside corner."

Remember, that was Roark in June of 2014, who had been doing this at the time. Attack lefties on the inside portion of the plate with the two-seamer, work either the inner or outer portions of the plate with his slider. Sounds like a simple enough plan. Here is where it got derailed—Roark never did either of those two things this season.

Let’s first approach the former. Here is his change in two-seam usage in order from 2014 to 2015:

For the most part, Roark stayed away from left handers on the inner part of the plate, but let’s not jump the gun. While it looks like this pitch was used on the middle to outer third of the plate to lefties a lot more often than 2014, that doesn’t necessarily mean it didn’t work. Then again, I’m sure you can count the amount of right-handed pitchers that have had success leaving a two-seamer over the middle of the plate to lefties on one hand.

As was—and likely always will be—predictable, Roark’s fastball took a beating in the new location. I could say that the BAA on the pitch was raised from .243 to .303, or I could focus on how opposing lefties have seen a significant power increase (from .149 ISO to a .220 ISO). I could. However, I think the best example of how the sinker has changed lies in the linear pitch weights offered by FanGraphs.

A couple disclaimers, though. For starters, linear type pitch weights don’t differentiate between a pitch thrown to a LHB or a RHB. This means that, since we are talking about just the famous Roark v. Lefties of Major League Baseball case, it isn’t ideal to include how he also pitched against RHB. While it isn’t ideal, it is the best we have, not to mention that he actually pitched better against righties with most of his pitches. It is not likely that any decrease in his linear weight is due to pitching worse against righties.

Another thing I should mention is that Roark throws only a two-seam fastball. How do I know this? He said so in his interview with Eno Sarris I mentioned earlier. The problem here is that PITCHf/x has categorized the pitch as two separate pitches, one four-seam fastball and one two seam. When it says that, in 2014, he owned a 10.1 wFT and 6.4 wFA know that he did not actually own two fastballs graded that high. He did, however, own one really good fastball that likely graded out somewhere in the middle of the two. When in 2015 both pitches fall to 0.1 and -0.2 in their linear weights, respectively, that should attest to the change that the pitch has undergone.

How on Earth could his two seam go through that kind of change in one year, though?

Well, look no further than how Roark's release point has shifted this season. All the veteran did was move from the first base side of the rubber to the third base side:

Normally this would not be a big deal. Pitchers do this all the time, and there is some question as to how much of an effect this type of move can have on a pitcher. My hypothesis was that Roark had decided to move to the third base side of the rubber so that he could have more access to the outer portion of the plate against lefties. This move made it tougher for him to work inside to lefties because his fastball would move back over the heart of the plate more often than he intended, so he decided to stick with the outer third.

However, that was just an initial guess. Realizing that I could always benefit from a second opinion from someone who has a greater knowledge of this area than myself, I reached out to my BtBS colleague Dan Weigel. After a chat with Dan, we came to a couple conclusions. Immediately Dan had picked up something I hadn’t when watching video of Roark pitch—the righty steps across his body a good bit.

When he threw from the first base side of the rubber he would end up landing with his front foot on the third base side. This would, in turn, cause the dragline created by his back foot to be centered (a good thing). When Roark moved to the third base side of the rubber it threw off his dragline, moving it away from the center (a bad thing).

With this explanation, Dan also sent an excerpt from the Tom House-Doug Thorburn book "Arm Action, Arm Path, and the Perfect Pitch: Building a Million-Dollar Arm". It went like this:

Any position on the rubber that ends up causing a pitcher’s head and spine to be off the center line when his shoulders square up to the catcher’s target will cause a late posture change, an inconsistent release point, and excess stress on the throwing arm, shoulder/elbow complex. Not one of the aforementioned is good for a pitcher’s level of performance or health.

In reality, the best pitchers in the game find that their most efficient and effective starting position on the rubber is that position which results in a dragline (no matter what its configuration) that finishes on the center line between the middle of the rubber and the middle of home plate."

This led us to the conclusion that the move to the third base side of the rubber made it tougher for the right hander to pitch inside to left-handed batters. While the initial reason for the move was likely to attack the outer portion of the plate against lefties and inner portion against righties, what Roark did was make it much tougher to reach the opposing side. This was something he could easily do from the first base side of the rubber, where he also proved he could have access to the outer part of the plate against left-handed hitters.

Right now, if you’re still with me you might be asking yourself "but Shawn, it’s just one of his pitches! Surely this couldn’t have had that big of an effect on Roark overall." To that I would say thank you for sticking with me, and don’t call me Shirley. Also, I don’t think you realize how much he throws this pitch. This season he owned the highest overall percentage of two-seam fastballs thrown (FanGraphs registers the pitch at 63.8 percent usage, although it was really closer to 66 percent) out of pitchers with at least 100 innings. Against left-handed hitters, he used it nearly 68 percent of the time. Translation: He threw it a lot.

Two seamers also have one of the biggest platoon splits against opposite-handed hitters. Don’t believe me? Check out this, this, this, and this. The only way Dave Allen and Max Marchi could shout this louder would be to buy advertising space on every MLB backstop. "DO NOT, FOR THE LOVE OF GOD," the ad would dictate, "DO NOT THROW THIS LEFTY A TWO SEAM. JUST DON’T DO IT."

When looking at the slider part of Roark’s comments, it should be noted that he threw only five sliders to left-handed hitters all season long. Five. Out of the 837 pitches he threw to left-handed batters this season that total amounts to a whopping 0.28 percent. Even in 2014 Roark did not throw a bunch of sliders to southpaws, doing so only 1 percent of the time. Also, not throwing sliders to lefties isn’t necessarily a bad choice, as it ranks up there with two seamers on the list of ‘probably shouldn’t be thrown to lefties from a righty’.

So not throwing sliders isn’t too much of a big deal; how about looking at all of his offspeed pitches in general? Other than his slider, Roark throws a changeup and a curve. Against lefties, the changeup was used around 14.3 percent of the time while the curve was thrown about 17.8 percent of the time.

Similar to what we did with his fastball, here is where each pitch was thrown the most in 2015. The top being his curveball, the bottom being his changeup:

He, apparently, had a problem with leaving curveballs over the heart of the plate. Without spewing more stats, would it surprise you if I said his curveball got hit hard? As for his changeup, it actually had a mild amount of success. It was the only pitch Roark owned with a pitch type linear weight above average (it owned a 1.0 wCH), and it had a 36% whiff/swing rate.

Why did his changeup have more success, you ask? Well, because it masks itself very well with Roark’s two seam. Both move horizontally away from a lefty—the changeup moves 1.2 inches more—while owning varying amounts of vertical drop. Not to mention that the changeup is also 10 mph slower than his fastball.

So putting all this together, why did Roark struggled against left handed batters? Pitch selection and tracking. In his interview with Eno Sarris, Roark noted that the scouting report had him throwing three pitches to lefties. However, it likely also denotes that two of these three pitches move away from a left-handed hitter. Not to mention that the move to the third base side of the rubber gives left-handed hitters a better opportunity to track these pitches—especially his two-seam.

Yes he has a slider, but it was thrown less than half of a percent of the time. The only pitch that Roark throws moving toward a lefty is a curveball, which is 16 mph slower than his fastball and 6 mph slower than the change—meaning that lefties can easily assume faster pitches will be on the outer portion of the plate. Surely the scouting report notes this.

Not having to worry about a pitch coming inside means that hitters have a much smaller area to look for a pitch, increasing their odds of making contact with a pitch they were already given the ability to track better. So, what is the solution?

I have two small-ish requests that the Nationals should try with Tanner Roark in 2016. First, moving back to the first base side of the rubber will surely allow him to regain meaningful access to the inner portion of the plate against lefties—the group that did the most damage against him. This one is obvious, as he had far more success on the first base side of the rubber anyway.

The second request, well, remember how Shelby Miller had success for a large portion of this season? Although lefties hit him hard, and he struggled down the stretch, the cutter he had been using more and more was a huge weapon for him against left-handed hitters. What if Roark did the same thing? What if he learned a cutter?

Not only would it give Roark an element of surprise for the first portion of the season—hitters wouldn’t expect a fastball to move in on them with the track record Roark has—but there is a fair amount of research that suggests it is one of the most successful pitches to use against opposite handed hitters. The four pieces I linked earlier all suggest that when facing a lefty as a righty (or vice-versa), cutters have a more favorable outcome when used than any other fastball type.

Now, the caveat with this one is pretty obvious. It is not easy to learn a Major League-ready cutter overnight. It isn’t easy for a pitcher to learn any pitch that is Major League-ready overnight. The ends could justify the means, however, and a cutter would vastly improve a repertoire that is currently two-seam fastball top-heavy.

Tanner Roark was one of many reasons the Nationals took home the NL East crown in 2014, but he was pushed to the bullpen at the beginning of the 2015 season through no fault of his own. Now with Jordan Zimmermann and Doug Fister departing to free agency, it is likely he will step back into the rotation for more than 12 starts. Should Roark move back to the first base side of the rubber, pitch to lefties on both sides of the plate, and add in another pitch that moves toward left-handed batters, look for him to bounce back for a Nationals team in 2016 that could really use him.

. . .

Shawn Brody is a contributor for Beyond the Box Score as well as a sophomore pitcher at Howard Payne University majoring in Business Management. He has the current misfortune of being a Red Sox fan. If you would like to get a hold of him, please feel free to email him at Shawnbrody9@gmail.com or follow him on Twitter @ShawnBrody.