The Year in Tanner Roark

August 7th, 2013: Jordan Zimmermann lasts only four innings in his start against the Braves, giving up seven hits and a couple of runs before Davey Johnson goes to the bullpen. In need of a bridge to the team’s middle relievers, Johnson calls in a rookie, Tanner Roark, to make his Major League debut. It is about as nondescript an appearance as one could imagine, as he faces six hitters, giving up a hit to B.J. Upton — okay, so one remarkable thing happened — but erasing it with a double play, and facing the minimum six batters over his two innings of work.

No one really thought much of it. Even in this post-game interview, Roark looks about as excited in his debut as everyone watching was. He was just a guy, a 25th round pick by the Rangers who was traded to Washington as half of the return for Texas’ acquisition of Cristian Guzman back in 2010. Fun fact; after the trade, Guzman hit .152/.204/.174 and was worth -0.7 WAR in just 50 plate appearances. Whoops.

But, really, giving up Roark wasn’t anything to lose sleep over. When Texas traded him, he was a guy with 75 strikeouts in 105 innings in Double-A, and he wasn’t even avoiding walks that well. He was an organizational guy, a non-prospect with no obvious upside who looked like a career minor leaguer. Even when he got to Washington, he didn’t really have any kind of major breakthrough. He just progressed through up the chain, got to Triple-A as a 25 year old, and then threw enough strikes in Syracuse last year that the team called him up when they needed a long man in the bullpen.

Well, he’s not a long man in the bullpen anymore. Since Roark’s debut one year ago today, here are the top 10 qualified starting pitchers by ERA-.

Clayton Kershaw 196.0 5% 30% 53% 6% 82% 0.289 50 52 58 6.8 7.8
Johnny Cueto 183.2 7% 25% 50% 10% 82% 0.227 52 82 83 3.5 6.4
Jon Lester 219.2 6% 23% 42% 4% 75% 0.300 61 66 87 7.0 6.3
Chris Sale 194.1 5% 27% 45% 9% 80% 0.279 64 71 75 5.6 6.4
Zack Greinke 212.2 5% 26% 48% 10% 82% 0.294 65 76 72 4.8 6.3
Tanner Roark 194.2 5% 19% 46% 6% 79% 0.262 69 85 96 3.5 5.4
Yu Darvish 211.1 9% 31% 36% 11% 82% 0.308 70 76 79 5.4 6.1
Felix Hernandez 217.0 6% 28% 56% 7% 70% 0.278 71 60 65 7.1 5.9
Hisashi Iwakuma 187.2 4% 21% 50% 11% 80% 0.276 71 81 78 4.0 5.8
Cole Hamels 208.1 6% 25% 47% 8% 78% 0.285 71 75 83 4.7 5.5

There’s Roark, right in the middle of a list of the best pitchers in baseball. Over almost 200 innings, runs have been scored against Roark at a lower rate than they have against Felix Hernandez. Few things in baseball will remind us of the unpredictable of the game as much as that one.

Of course, this stat also says as much about ERA as it does about Roark or Hernandez. If we sort by FIP-, he’s 25th, not 6th, but still, that puts him in a tie with Madison Bumgarner and Sonny Gray. That’s not bad company either. If we look at xFIP-, which factors in that he’s likely been a bit lucky in allowing home runs, he winds up 42nd among qualified starters, tied with Doug Fister. This is a more realistic comparison for Roark, as while he’s a great story, he’s not Felix Hernandez. But he might be Doug Fister.

Tanner Roark 194.2 5% 19% 46% 6% 79% 0.262 69 85 96 3.5 5.4
Doug Fister 170.1 5% 17% 49% 9% 80% 0.315 79 93 96 2.5 4.0

On walks, strikeouts, and groundballs, Roark and Fister have been very similar over the last few years. Like Roark, Fister was also a non-prospect with mediocre stuff, and his development into one of the game’s best starting pitchers has been a significant surprise to just about everyone. If there’s a career path to copy for Roark, Fister is probably about as good as it gets, within reason anyway. The good news for Roark is that he’s basically been a carbon copy of Fister over the last year, especially when you look at their plate discipline numbers.

Name IP O-Swing% Z-Swing% Swing% O-Contact% Z-Contact% Contact% Zone%
Tanner Roark 194.2 29.4% 59.7% 45.1% 66.4% 90.3% 82.8% 51.6%
Doug Fister 170.1 28.3% 58.8% 44.3% 68.8% 91.3% 84.5% 52.5%

The key for these kinds of pitchers is pinpoint location, and Roark certainly has pitched like a guy who can put the ball wherever he wants. And when you look at his location grids, it becomes clear what his preferred method of attacking hitters is.

Screen Shot 2014-08-07 at 12.16.29 PM

That’s Roark versus right-handed batters this year. You could basically draw a straight line from the upper left-hand corner of the strike zone to the far bottom right, and note that Roark pitches almost exclusively along that line. He’s coming at right-handers either up-and-in or down-and-away, with a bit more of the emphasis on the low-outside corner of the zone.

But what’s interesting about Roark is that the up-and-in pitch is really where he’s making his mark. For all the talk about the need for middling stuff guys to “keep the ball down”, Roark’s relative effectiveness is most obvious on those up-and-in pitches. Here is the RAA/100 map for Roark’s locations against right-handers this year.

Screen Shot 2014-08-07 at 12.23.53 PM

His down-and-away numbers are above average, but his up-and-in numbers are excellent. He’s dominating hitters in this part of the zone, generating a ton of weak contact in that area. Let’s switch to the 5×5 grid to make it more clear. Here’s Roark’s rate of contact allowed to right-handed batters this year.

Screen Shot 2014-08-07 at 12.27.09 PM

Hitters are putting the bat on the ball on their up-and-in swings at nearly the same rate as they are when they swing at his pitches that are grooved down the middle of the plate. Roark gets his strikeouts down-and-away, but up-and-in, he’s a pitch-to-contact guy. But look at what happens on that contact.

RHBs batting average against Roark this year:

Screen Shot 2014-08-07 at 12.29.11 PM

RHBs isolated power against Roark this year.

Screen Shot 2014-08-07 at 12.30.12 PM

Up-and-in is just as effective for Roark as down-and-away, giving him two areas of the strike zone to attack hitters where he can generate weak contact. As Chris Young explained to Eno Sarris a few months ago, hitters have a hole at the top of the zone, and pitchers with great command can exploit that hole, but it’s a fine line to walk.

This is basically the way that Matt Cain limited home runs forever, and how guys like Justin Verlander have lived as fly ball pitchers without being home run pitchers. If you can locate your fastball up-and-in, you can pitch up and avoid home runs. Of course, Roark doesn’t have a Verlander or Cain fastball, but then again, Chris Young throws 85 and lives at the top of the strike zone, and he has a career 8% HR/FB ratio. Jered Weaver has shown that this same approach can work with a below average fastball. Roark wouldn’t be the first guy with mediocre stuff to limit hard contact by exploiting the very thin hole in hitter’s swings on up-and-in pitches.

Of course, he also wouldn’t be the first guy with mediocre stuff to throw 200 decent innings and then regress quickly. There are some parallels between Roark now and the 2010/2011 version of Vance Worley, for instance. We don’t have enough data to forecast Roark as the kind of pitcher who is going to beat his FIP and xFIP by significant margins.

But we do have some evidence that he might be one of these guys, and the good news for the Nationals is that he’s showing the kinds of tools that will allow him to succeed even if the BABIP and HR/FB ratios both regress back to league average. If he’s not an exceptional contact manager, and he’s just a guy who pounds the strike zone and gets ground balls, then he’s an above average big league pitcher. If he can keep exploiting the up-and-in hole, however, generating significant amounts of weak contact, then he’s not just a fun story anymore, but legitimately one of the better young arms in baseball.

Predicting the rest of Roark’s career is likely as futile a process as predicting that Roark would rise from a nothing prospect to this level of pitcher in just a year’s time. Perhaps Jay Robertson — the Nationals scout who recommended Roark in the Guzman deal — has a crystal ball and knew this was coming, but I’m guessing he’s just as surprised at how well this kid has developed as the rest of us. Given what he’s done over his first full year in the big leagues, though, we might want to stop being surprised when Tanner Roark dominates big league hitters. He might just be pretty good at it.

Print This Post

Dave is the Managing Editor of FanGraphs.

12 Responses to “The Year in Tanner Roark”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. Al says:

    Small question, did you just know off hand the scout who recommended Roark in the Guzman deal?
    Find it in a write up about the deal years ago?
    Or touch base with someone in the organization in order to give them credit?

    It’s a single sentence in a decently long article but I liked it. Thanks for the article, appreciate details like this.

    Vote -1 Vote +1

  2. Sam says:

    Great Stuff

    Vote -1 Vote +1

  3. The Lorax says:

    So can it be argued that his trade value would never be higher if we expect regression to the Vance Worley level? I know it’s a bit early to tell but I’d always rather trade early then later

    Vote -1 Vote +1

    • David says:

      His value is probably at it’s local peak.

      Going into next season there will be an expectation of regression, and his value will be lower.

      If he does this again next year then his value will go back up and likely stay there.

      Vote -1 Vote +1

    • Will says:

      Sure, but according to the market, he’s worth a AAAA utility player, a LOOGY, and a fringe top 100 prospect, using Fister as his comp.

      Vote -1 Vote +1

  4. ALEastbound says:

    Don’t give Trevor Bauer any ammunition Dave! Pitching up and in?

    Vote -1 Vote +1

  5. Jeremy B. says:

    Regarding the batting-average heat map, is there a way to narrow those results to balls that are actually put in play? I understand that the heat maps are valuable because they account for the value of every pitch thrown, not just those where a result happens (hence the RAA/100P map). But the raw numbers in the batting-average heat map don’t resemble batting average as we’re typically familiar with the stat, because they account for pitches taken (and fouled off, and swung-on-and-missed). For BABIP purposes, I’d be curious what the average is on high-and-inside pitches for balls actually put into play.

    Any way to do that? Or are we just supposed to look at the relative differences in the hitting zones (i.e., blue vs. red) and draw conclusions from that?

    Vote -1 Vote +1

  6. BDF says:

    Does this suggest anything about a possible method for identifying pitchers likely to perform better at the major-league level than they did the in the minors?

    Vote -1 Vote +1

  7. Moses says:

    It’s funny that the final two heat maps have the “red” area where he’s allowing a .136-.154 AVG and a .091 ISO, respectively. Can I assume Marco Estrada’s ISO/P heat map is “blue” at .250?

    Vote -1 Vote +1

  8. Pig.Pen says:

    Harper over at Nationals Baseball made a good point about Roark though, ( He has one of the lowest aggregate OPS of opponents faced this year at .667. Compare that with Fister who’s opponent’s aggregate OPS has been .703. It’s not the end all be all of stats and he can still be a very good pitcher, but as a Nats fan I must admit that I keep waiting for him to run out of black magic and pixie dust.

    Vote -1 Vote +1