OK, the teaser is a bit misleading: Ross Detwiler was already in the positive in 2012, serving as a reliable (when in the rotation) starter for the Washington Nationals throughout the year. There are number of reasons why I didn't explain WAR in the pre-season PDBs, but by that measurement Detwiler (+1.8fWAR, +1.6bWAR) was what Fangraphs would call a "role player" and a borderline "solid starter." Sounds about right, but I'd bet Detwiler would be the first to say that he can do better.
Undoubtedly, it takes more than one number to really understand the hows and whys of pitcher performance, and since Reddington is "right there with his hawk-eyes on," let's review Detwiler's performance in 2012 and look forward to his 2013.
Past Scouting Assessments
"He was good but not as good as he was vs. Dallas Baptist. His fastball was 90-92 and touched 94. His command was just OK. He was behind a bunch of hitters. His breaking ball was very good at times. His change is definitely his third pitch. It was just OK. I think he has first-round stuff. The arm works and he is so thin that I think as he gains weight he can become even better."
Detwiler saw early returns on his prospect status, ranking out as the second-best prospect in the Nats' system in November 2007, five spots above Jordan Zimmermann. Baseball America awarded him best fastball in the organization, a harbinger of things to come.
In April 2008, the publication ranked Detwiler as the fifty-first best prospect in Major League Baseball, ahead of other hurlers like Max Scherzer, Justin Masterson, Luke Hochevar, Collin Balester, Neftali Feliz, and Jeff Niemann, and behind Joba Chamberlain, Rick Porcello, and Jair Jurrjens.
By January 2009, Detwiler had performed well enough for Baseball America to again list him as the second best prospect in the Nats' organization. Zimmermann, however, had performed better and grabbed the top prospect (and fastball) billing.
During the 2009 year, Detwiler pitched in the bigs for 75.2 innings. A 5.00 ERA made his performance forgettable, but a tremendous HR/FB rate (which, admittedly, varies in equally tremendous fashion) depressed his FIP to a ~3.85 mark. In this case, ERA probably told the more accurate tale: Rare is the pitcher who gives up just 3.7 home runs per 100 fly balls, particularly considering the sample size.
2010 was a lost cause, as National Det underwent hip surgery for injuries. He also had another DL stint after an attempted mid-to-late season return.
In March 2011, one of Baseball America's resident prospect experts, Jim Callis, had this to say about Detwiler:
Jay (South Riding, VA): Do you think Ross Detwiler can be a #2 or #3 type pitcher now that he is healthy and has stopped throwing across his body so much? I watched him at spring training and he looked great. Also, could Espinosa be the ROY of the year? He's a flat out stud with the glove at 2B.
Jim Callis: Not that optimistic about Detwiler. Not usually a good sign when they're changing your mechancics [sic] after you got drafted sixth overall with those mechanics.
The 2011 season was again unremarkable. 2012, however, was a meaningful step forward.
2012 Pitches and Pitch Movement
BrooksBaseball's manual classification system had Detwiler throwing a fastball, sinker, curve ball, and change in 2012. As the table shows, he liked the hard stuff:
|Four-seam fastball||39%||94.1 MPH|
|Curve Ball||13%||79.7 MPH|
|Changeup||7%||85.1 MPH (~10% difference from hard stuff)|
Regarding movement, Detwiler shows a noticeable difference between his fastball and sinker, despite the similar speeds, as the graph below indicates.
Where you see the vertical line of changeups (CH), you'll also notice the general movement of the fastballs (FA) and sinkers (SI). Specifically, almost every pitch to the left of those changeups is a fastball; almost everything to the right is a sinker. Over the course of the year, the difference averaged out to about 5 additional inches of both horizontal and vertical movement -- not insignificant when considering the similar speed and release point of the offerings.
Many here likely saw the following passage earlier this year, but Detwiler favored the fastball/sinker diet in 2012, with good results:
Detwiler accomplished the successful run by leaning on his sinking fastball. Last year, he threw fastballs on 80.3 percent of the pitches, more often than any other starter in the National League. One start, he shut out the Philadelphia Phillies for seven innings while throwing 85 sinkers in 88 pitches.
"He has an idea what he wants to do," pitching coach Steve McCatty said. "Sometimes, he can be stubborn. But he’s learned how to change what his game plan is, because he’s seeing what’s going on, reading the situation of the games."
Check out his September start against the Dodgers. Many of the outs he earned were via either the fastball or sinker, but note that the curve ball makes a few late-count appearances too (later, we'll see this was generally true throughout the year). Working up, down, in, and out, Det turned in a really good performance, getting 5 strikeouts in 6 innings pitched.
But the team isn't allowing Detwiler to rest on his laurels this year:
The Nationals, especially Johnson, want Detwiler to throw more curves and change-ups, to keep hitters guessing and to set them up for later in the game. "I don’t think it’s about developing. I think it’s about trust," Suzuki said. "He’s got it. It’s there. You have to be confident when you’re throwing it. I think towards the end of the year, he’s got more success with it, which allows him to be more confident. If he mixes some off-speed pitches in, I think it’ll put him over the top."
Detwiler built a 10-8 record, and pretty good sabermetric season, by leaning on the fastball. Yet the team wants more diversity in pitch selection. What might be the reason for this?
First, some of the core stats that help to evaluate performance, with league averages on bottom:
Above-average performance in walks and home runs per nine innings earned Detwiler a 3.40 ERA, which was 14% better than the NL average in 2012. FIP, which only really looks at strikeouts, walks, and home runs, saw Detwiler as closer to an average pitcher, with a mark of 4.04 -- 4% below the NL mean. xFIP wasn't much of a fan either. Detwiler's 4.34 mark in this statistic, which uses a league average HR/FB rate, was higher because the league average HR/FB rate was higher than his.
Part of the reason Detwiler succeed in limiting the long ball -- and in his 2012 performance generally -- was his ability to get ground balls. 50.8% of balls in play against him burned worms, against a league average of 45.7%. He also gave up around 5% fewer line drives than the average pitcher, suggesting that his .263 BABIP wasn't all "luck" because: 1). Less line drives generally means a lower BABIP; and 2). Extreme ground ball pitchers have lower BABIPs on grounders than normal ground ball pitchers. 2012 is what happens when you post career best marks in GB% and LD%.
Below is a GB% graph for qualified NL starters last year, which helps place Det's 2012 in context.
The other number in his batted ball profile that stands out is the amount of contact made by opposing hitters outside the strike zone. While the average pitcher saw 63.6% of pitches swung at outside the zone contacted, Detwiler was hit 71.5% of the time a batter swung at a wayward offering. I know more swings and misses would result in more K's, but it's unclear how hittable these pitches were. Additionally, pitch location information can vary with left handed pitchers. So make of this what you will.
Split-wise, Detwiler was death to lefties, who slashed .165/.255/.259. Right-handed batters fared much better, batting to a .258/.320./414 line.
Finally, recall that SIERA is a pitching stat that incorporates batted ball data into its calculus and is scaled to ERA. Detwiler's SIERA was 4.35. Basically, Det's above average GB performance couldn't overcome the lack of strikeouts for SIERA's affection. But what if he had a league average strikeout rate? Well, that would translate to a mark of ~3.76, giving him a performance that was better than league average and a peer group that includes Zimmermann, Edwin Jackson, and Mat Latos.
Fangraphs does not use Brooks' classifications, so sadly this isn't a strict apples to apples comparison. However, confirming the scouting reports and general observations on the effectiveness of his hard stuff, Detwiler's fastball and sinker both sat at 7.4 runs above average, a good value considering that the best pitches in the league usually return around 15-25 runs above average.
Unfortunately, Det's curve ball (which Fangraphs labels a slider) hurt him, as reflected in his -5.2 run value. The change was around average, at 0.
I can't do it any better than Brooks:
Det's love of the hard stuff wanes late in the count when he is ahead or has two strikes: lefties see fewer straight four seamers while righties dodge the sinker. Yet pretty much any other time, as a hitter, you can bet you'll see a hard one or two.
Detwiler's success is predicated on the effectiveness of his hard stuff. If the fastball and sinker aren't generating grounders in exchange for fewer line drives or home runs, he will struggle to be effective. Duh.
But what's up with the lack of strikeouts, and how is that related to the 2013 "more breaking/offspeed" initiative?
As Baseball America noted, he had good K/BB ratio and the best fastball in the organization at one time. He also had respectable strikeout numbers through 2010 while in the minors. His Pitchf/x card, and some other information, helps to see how his 2013 more-breaking/offspeed-heavy approach could pay dividends.
First, Detwiler's fastball was whiffed 19.09% of the time when batters swung. That's better than Stephen Strasburg, and placed Detwiler 24th out of 118 qualifying pitchers.
Second, Detwiler's sinker generated 12.47% swings and misses - not elite overall, but 33rd out of 82 qualified pitchers.
If he's throwing these pitches often, and they are getting more swings and misses than his peers, why is he below average when measuring strikeouts?
Recall Detwiler's approach was more curve-centric when he was ahead in the count, or had two strikes. His curve whiff rate? 26.45%. The MLB average, however, was 29.17%, and Detwiler's rate placed him only 67th out of 103 qualified pitchers. This may explain in small part his below average strikeout performance.
On the curve, there is cause for optimism: he threw the bender far, far less frequently than the hard stuff overall, yet still earned nearly the same amount of total strikeouts with the pitch (31) as he did with either the sinker (32) or fastball (38).
Maybe there was a reason he didn't go to the curve a lot, though. When the curve was in the strike zone, batters didn't swing and miss all that much. While some of these looks may have been on two strike counts, his negative run value for the pitch suggests that contact was made. Moreover, Detwiler's called strike to ball rate on the curve ranked an unimpressive 93rd out of 103 among qualified pitchers, at .31 strikes per called ball. Throwing one strike per three balls isn't ideal, and it will decrease the value of the pitch.
The chart below shows Detwiler's whiff rate against lefties and righties in 2012 on breaking pitches and the relative lack of cuts and misses:
Now, here is the swing rate to put the above numbers in context:
Batters were hacking at the curve fairly frequently, and that (maybe unsurprisingly) went for out of the zone, too.
So if the curve was in the zone, and the batter didn't whiff or take the pitch, what happened? The next graph shows batters' slugging percentage with the same parameters as before:
Detwiler sustained some damage on curve balls in many of the places you want the pitch, and in all the places you don't.
In sum, batters could reasonably figure they were getting a curve late, didn't worry about it being unhittable, and also weren't concerned with his ability to throw the pitch for a strike on demand. The last chart displays the product of this calculus, and may be the consequence of going to a pitch that neither generates a lot of swings and misses relative to his peers nor is thrown for a called strike as often as it is a ball.
Perhaps illogically, I believe the more breaking and offspeed-focused approach is the correct call. One reason I think this is because Detwiler is still a young pitcher who has worked through injuries and doubts concerning his ability to succeed in the bigs. As shown above, he built a pedestal of confidence on the strength of favorable results on hard stuff. Now appears to be as good a time as any to attempt capitalize on his promise.
The second reason, and it's a bit of a corollary to the first, is that in his current mold Detwiler won't approach the long-term success of pitchers who work with better strikeout rates. And while that is perfectly fine for many big league twirlers, the Nationals (and perhaps more importantly Detwiler) don't seem content with "role player" or "solid starter" performance.
If Detwiler does go to the curve and change more, he must show that he can command those pitches to keep batters honest. This is especially important against right-handers, who performed much better than lefties against Detwiler in 2012.
Considering the small curve ball sample from last year, I wouldn't be surprised to see an increased whiff rate and corresponding increase in strikeouts, all else being equal. If he doesn't develop those offerings, I worry that Detwiler will have to maintain his 2012 performance on his hard stuff to approach ~1.5-2 WAR. It's not an unreasonable proposition, but even with just the variability of HR/FB rates, it may not be the best idea to throw cheese all day long. Fortunately, it appears that the Nats and Detwiler think the same thing.