Second Thoughts Game #122: Orioles 0, Indians 6
Saturday's match-up between former Clevelander Ubaldo Jimenez and current Clevelander Carlos Carrasco was intriguing because, had the Indians signed the former, the latter would not have been a rotation piece.
After Oakland swept up Scott Kazmir, the Indians rotation was, in no particular order, Kluber, Masterson, McAllister, Salazar and one question mark - had Jimenez been signed, he'd have been the fifth; since Jimenez was not signed, there was a nominal competition for the fifth starter position, which Carrasco won very nearly by default and which he promptly surrendered. This pushed Carrasco to the pen, a sterling half-season stint in which pushed the right-hander back into the rotation.
Carrasco has proven that he belongs on a major league club - if not necessarily in the rotation, at least in a bullpen role. His slider is a viciously effective pitch that few batters can contend with. To examine the renewed effectiveness of his recently-tweaked slider, one must introduce the concept of linear weighted run values. Since a home run is the worst possible outcome on any given pitch, and a swinging strike the best, it makes sense that each outcome in between - triple, single, ball, foul - has some tangible run value that helps or hurts a pitcher. Each of the values assigned to each outcome has a positive or negative weight, and if one adds up all the run weights, one can derive how effective or ineffective a pitch has been - this concept is called a pitch's Run Value.
Since not every pitcher throws an equal number of pitches, the rate of this effectiveness is typically expressed as Run Value per 100 pitches. For the sake of comparison, Corey Kluber's curveball, which is both an excellent strikeout pitch and extremely pretty, is ranked as the second best per-use curve in the majors, with a +4.12 run value per 100 pitches. Carrasco's slider has a per-100-pitch value of +4.50, which ranks second among all sliders from pitchers with more than 60 innings pitched.
The rate numbers cannot be taken as a one-to-one comparison, given that Carrasco has been operating in shorter bursts from the pen and that his lower total uses of the pitch means that the league hasn't had the chance to adjust to its recent pseudo-elite status. Going forward, it may not remain an elite pitch, but to date in 2014, his slider has certainly given an excellent impersonation of one.
The success of his slider means that, even if his 95+ MPH heater were to continue to be the net negative pitch that it has been over his career (-28.38 total), he has one genuinely plus-plus pitch that he can rely on. I've made the argument for Carrasco: Super Reliever here, and the argument remains largely unchanged.
Since the argument in favor of Carrasco is an easy one to make - he's thrown 12 consecutive shutout innings and has posted sub-2 FIPs in each of those games as well as having not walked a batter - what follows will be an argument against Carrasco's continued excellence as a starter. It's not an argument that Carrasco is condemned to be a bullpen man or a #5 starter, but merely that his results from Saturday are unlikely to be the norm. If it seems obvious to say that outings of seven shutout innings are unlikely to continue, perhaps it is; more concretely, the factors fueling his two most recent shutout starts and a 2.00 ERA since being pulled from the rotation are largely unsustainable. His ERA suggests he's an ace; it's entirely reasonable to pump the brakes on this rather wild conclusion and dig deeper. The biggest reasons for caution stem from his BABIP and his role on the team.
As a reliever, he can focus on his slider and change and merely allow his fastball to be a 97 MPH change-of-pace pitch. As a starter, however, Carrasco must be able to rely on his fastball and change as primary options, and while the results were good on Saturday, allowing only three hits off the fastball, the means to that end appear, at first glance, unsustainable. On Saturday, Carrasco used variations of his fastball 51 times; of those 51 times, 31 were strikes. Of those 35 strikes, 14 were put in play in fair territory. Of those 14 balls put in play, only three went for hits. On one hand, Carrasco's groundball rate (a well-above-average 54.7% of all balls in play) leads to - at face value - the sort of batted ball profile that results in good run prevention.
On the other hand, there were several extremely long outs in Saturday's game, illustrating that while he does not give up many fly balls, the fly balls he does allow can be damaging. This tenet is reflected by his Infield Fly Ball rate. Given that infield fly balls go for outs at nearly the same rate as strikeouts, the percentage of fly balls that are infield fly balls - IFFB% - determines how well a pitcher limits contact on those fly balls. In many cases, such as that of Jered Weaver or Chris Young, IFFB% is regarded as a repeatable skill. Relative to the league average of approximately 10%, Weaver and Young perennially run IFFB rates nearer 15%.
In contrast, Carrasco's career IFFB% is 6.4%, meaning that when Carrasco does surrender fly balls, they typically go to the outfield, where they are both less likely to go for outs and more likely to go for extra bases. The fact, then, that Carrasco had an in-game BABIP of .167 on Saturday - and .214 against his fastball - is exceedingly and anomalously low. In fact, since Carrasco has been removed as a starter, his .254 BABIP is an opposite and nearly-equal extreme of his .355 BABIP in his first four starts. What's more, while his IFFB% changed slightly from 5.0% in his first starting stint to 7.7% since, his line drive rate - the primary driver in BABIP - has not substantially changed, with his first four starts posting a 14.3% LD% and having posted a 15.5% rate since. Along with his LD%, his ground ball rates and fly ball rates have likewise remained relatively stable. There are factors within Carrasco's control that impact his BABIP; none of those factors, however, explain the 100-point shift between Carrasco's BABIP before he was pulled from the rotation and after that point but sheer volatility
The primary driver in Carrasco's low BABIP, then, is the exceedingly low line drive rate. For comparison, the league average is 20.6%, and even saber-savvy pre-season critics of Kluber pointed out that, despite his extremely good xFIP, his LD% had been as high as 24% in 2013. Small differences in line drive rate carry serious implications for BABIP, and Carrasco's low line drive rate explains how his BABIP is as low as it is. Simply because an explanation exists, however, does not in turn imply sustainability. Line drive rate, as posited by Baseball Prospectus writer Derek Carty, is an exceedingly volatile stat, stabilizing only after 2000+ balls in play - stabilizing, in other words, only after four full seasons have elapsed. Irresponsibly oversimplified, this states that unless Carrasco were to carry a 15% line drive rate even until 2016, 'randomness' would be a reasonable explanation for his low line drive rate. Responsibly oversimplified, Carrasco's currently-low LD% does not predict an equally-low LD% going forward. One can expect his characteristically good ground ball rates, but his line drive rate is unlikely to lead to the same suppressed BABIPs as it has to date in 2014.
BABIP aside, one should note that his generally positive year-to-date peripheral stats include his relief appearances. If one sets ERA to the side, one notices that Carrasco's xFIP is 3.18. This rate, if run by a starter over a full season, would absolutely suggest at the very least a solid #2. However, not only do these numbers take place over only 77.0 innings, they also include 43.0 innings of relief. As my previous article on Carrasco's reliever status noted, the approach taken by Carrasco in relief is tangibly and dramatically different from his approach as a starter, so reliever numbers should likely be excluded from an analysis of Carrasco the starter.
What remains after removing relief appearances is 34 innings, a 4.50 ERA, a 2.98 FIP and a 3.43 xFIP. All of this suggests that Carrasco, as a starter, is approximately a solid #3 or fringe #2 starter, but 34 innings is simply too few to say anything about Carrasco's future effectiveness with any certainty. Uncertainty is not synonymous with deficiency, but it is likewise far from synonymous with acedom. While his fielding-independent numbers look good, 34 innings remains an exceedingly small amount of time over which one can judge Carrasco's performance as a starter.
Those are some arguments against Carrasco, the ones that most immediately relate to Saturday's performance. They were not even-handed and they did not address the other side of the issue, which side might most effectively be summarized as 'he's thrown twelve consecutive shutout innings.' He's been extremely effective in his last two games, and there exist real reasons for optimism. Carrasco allowed zero walks over his last twelve innings; on Saturday, he induced 11 whiffs on 79 pitches for an exceedingly high 13.9% swinging strike rate, and a 72.1% strike rate overall on the game. Ultimately, one can take the Carlos Carrasco argument in nearly any direction while still having a reasonable case.
Until such a time as we know for certain what Carlos Carrasco is, however, there exists more than enough time to enjoy baseball and take in a one-sided pitching performance that featured excellent baserunner prevention and seven shutout innings. Cleveland will take the win, and fans might be best served just letting the unpredictable road that is baseball run its wandering course.
John can be reached on Twitter at @JHGrimm. He can also be reached by e-mail at email@example.com.
It also seemed earlier in the year that he was afraid to pitch inside. That he pretty much only worked the middle to outer half of the plate. I wonder whether the suspensions played into this at all, that he was worried about hitting someone again. Or it could just be confidence, and learning to attack hitters better. Or maybe he was trying to pitch inside, but his command was still not what it used to be pre-surgery. Will be interesting to see how he performs for the rest of this year.