An Investigation (and Validation) of Rubby’s Improvements

Rubby de la Rosa twirled a gem Monday night.* It was a night of several unlucky fantasy pitching performances: Corey Kluber, Chris Sale, Matt Harvey and John Lackey together struck out 34 batters and allowed only three runs in a combined 32 innings pitched, and the Twitterverse made sure everyone knew.

Yet there was de la Rosa, plugging away against the Marlins (and against Dan Haren, who admirably tossed eight innings of two-run ball in defense), needing only 94 pitches to get through a full nine innings of play. Rubby, too, settled for a no-decision, his due to a poorly timed two-run homer allowed to J.T. Realmuto in the 7th inning.

I was reluctant to invest in Rubby at first, remembering his wholly disappointing 2014 season. (In his defense, he was never really a touted, let alone highly touted, prospect, even within the Dodgers and Red Sox systems, so the term “disappointment” is used loosely here.) However, I have bought several shares of Rubby in the past couple of weeks due in large part to a smattering of injuries but also a series of respectable performances by de la Rosa.

Starting with the obvious: de la Rosa has produced the same amount of WAR in 2015 as he did in 2014 in about half the time. His ERA and various FIPs have only seen marginal improvements, but he has greatly reduced his WHIP, from 1.49 to 1.11 (which, yes, is probably caused in part by a suppressed batting average on balls in play (BABIP)). Most importantly, he has improved his strikeout rate by more than 5 percentage points while shaving another 2 ticks off his walk rate.

We can attribute these augmentations primarily to two adjustments: he improved his slider, and he is pitching more efficiently — details that are not necessarily mutually exclusive.

Let’s start with the former. Depending on whom you ask, de la Rosa’s slider has been anywhere from the 22nd-best of its kind (PitchF/X) to the 10th-best (FanGraphs) per 100 thrown. (In nominal terms, aka total value earned, his slider ranks 27th- and 16th-best, respectively.) Again, depending on whom you ask, the value of his slider has leaped from -2.18 to +1.57 (PitchF/X) or from -3.00 to +2.89 (FanGraphs). No matter whom you ask, however, they unanimously agree: Rubby’s slider has never been valuable before, and Rubby’s slider is, in relative terms, fairly valuable now.

I have depicted below a comparison of Rubby’s FanGraphs plate discipline statistics between this year and last strictly for his slider:

The pitch is simply better by all accounts. One, he’s coercing more swings on pitches outside the strike zone with less contact on such swings to boot. Two, he seems to be spotting the pitch incredibly well, daring to throw it in the zone almost 60 percent of the time, also while inducing far less contact. This all manifests itself in more than twice as many swinging strikes on the pitch in 2015 than in 2014.

Thanks to Eno Sarris, who was working on some Rubby analysis at the same time for a front-page post, I present you with side-by-side GIFs comparing the 2014 slider (left) to the 2015 incarnation (right).

Update (2:18 pm EST): Read Eno’s post about the Diamondbacks’ rotation here, where he keenly assesses the footage of Rubby’s sliders. It’s a good read, per usual.

The two pitches don’t look profoundly different — perhaps a release point by fractions of an inch and his arm coming loaded a bit sooner and longer than before. The new slider, however, appears to move less than it did before, a notion validated by the PitchF/X movement metrics (MOV) in the table above. It could be a matter of frame rates, but the pitch simply looks faster, the break coming much later, rather than hanging and, with great fortune, fooling Mike Trout. Indeed, PitchF/X notes a shrunken variation on his sliders’ velocities: they ranged from 72.8 to 90.0 mph in 2014 but now range only from 79.4 to 87.0 mph — a range almost identical to that of his change-up (81.2 to 87.8 mph), arguably his most effective pitch prior to this season and one with much different movement.

As aforementioned, de la Rosa is working much more efficiently in general, as demonstrated by this table:

He’s pounding the zone to the tune of the 3rd-highest rate of all qualified pitchers (9th-highest, per PitchF/X) while managing less zone (and overall) contact in the process. And although his rate of first-pitch strikes (F-Strike%) is nothing special compared to the league average, it is for him, given its 12-percentage point improvement. The resulting decrease in pitches per batter of 0.28 equates to roughly two additional batters faced per 100 pitches — potentially two extra outs per game, surely something appreciated by Chip Hale and fantasy owners alike.

De la Rosa has struggled with his control throughout his professional career. It appears that he has not only improved his control but also bolstered the command of a legitimate put-away pitch, one which has induced strikeouts on more than 30 percent of plate appearances that end with it.

This is all to say that your window to buy on de la Rosa (or simply add from him from waivers) may close soon. Mike Podhorzer’s pitcher expected K% (xK%) formula says de la Rosa’s strikeout gains are legitimate, and Rubby’s plate discipline stats validate his stinginess with free passes as well, leaving you with a 16.4 K-BB% that ranks among the top 30 for qualified pitchers, ahead of Ubaldo Jimenez, Zack Greinke, Cole Hamels and Stephen Strasburg, consecutively.

Oh, and he’s owned in only 21 percent of Yahoo! fantasy leagues. That’s interesting, man. That’s interesting.

* I didn’t realize this was a groaner until I remembered his name is actually pronounced Ruby, identical to the name of a gemstone, and not the phonetically correct and more amusing way. Apologies to those who may have groaned.