José Urquidy’s Offspeed Makeover

José Urquidy delivered a relative gem in Game 2 of the World Series, punching out seven and giving up two runs with no walks in five innings. Included in that was a big bump in velocity across the board: Urquidy’s fastball gained 0.6 mph from its regular-season average (93.1 versus 92.5); his changeup spiked nearly two miles per hour (86.2 versus 84.4); and his slider saw the largest increase, from 79.3 mph to 81.7. Dialing up the velocity to this extent can effectively create new pitches, and with that an element of unpredictability, particularly when it comes to a pitch’s movement. That can be a plus in the postseason, but what about the flip side; could that extra velocity and movement make things worse?

On the surface, Urquidy throwing his hardest fastballs of the year in his most important start can’t be a bad thing, assuming it doesn’t compromise his command. Can he so precisely regulate the velocity of his offspeed offerings? Amazingly, we might be seeing the reverse from Urquidy, who seems to be throwing the offspeed stuff harder but holding back on the fastball.

We wouldn’t expect each pitch to have the same gain in velocity, but the jumps for the changeup and slider are notable. And with that extra velocity come changes to each pitch’s shape. On the slider, Urquidy lost five inches of drop and six inches of horizontal movement compared to his seasonal average; the changeup, meanwhile, lost four inches of vertical movement. Given that home plate is just 17 inches wide, changes of six, five, and four inches are certainly substantial. But location is important, too, and that’s where Vertical Approach Angle comes in.

The concept of Vertical Approach Angle (or VAA) has become more prevalent in the past few years as a way to simplify vertical movement in the context of release height and location; it measures the steepness of the angle as a pitch crosses the plate. (For more on the subject, check out these excellent primers from Ethan Moore and Alex Chamberlain.) VAA isn’t publicly available, but it can be approximated with public data: release point, location, and pitch movement. For context of the measure itself, the best “rising” four-seam fastballs at the top of the zone will have flatter vertical approach angles at something like -4 degrees, whereas 12–6 curveballs below the knees will come in around -15 degrees. While the measures may not read as intuitively as inches of movement, know that a degree or so within a pitch type is substantial.

As you can see, the slider and in particular the changeup have markedly different VAA measures in the postseason (and even from his lone start in the ALCS to his Game 2 start on Wednesday). So what’s the net effect of that?

VAA should be considered in the context of velocity, and using run values, we can see how sliders and changeups like Urquidy’s perform in aggregate. His regular-season slider is highlighted by the black dot on the left, and the sliders from his two playoff starts are represented by the triangles to its right:

Unsurprisingly, adding velocity to your slider is generally a good thing, and the postseason version of Urquidy’s has a vertical movement shape and velocity pairing that is closer to the league’s better-performing sliders.

The changes in horizontal and vertical movement together also appear to be a positive for Urquidy. Horizontal Approach Angle is much the same concept as VAA, but for left-to-right or right-to-left pitch movement. As before, here’s a plot of run values that include his regular-season and postseason sliders, once again represented by a black dot and black triangles, respectively:

The differences in HAA for Urquidy’s two playoff starts are much more defined by location, with a gap of only an inch or so in horizontal movement. This isn’t to say that flatter sliders are necessarily better — obviously velocity is heavily intertwined here — but rather that the horizontal approach or changes in horizontal approach still leave him with a relatively successful shape.

Where the increase in velocity covers for a drop in vertical movement or inconsistencies in horizontal location for the slider, the same cannot be said about Urquidy’s changeup:

The gains in velocity don’t hurt, but either a loss in vertical movement or a higher location bring that VAA up, giving the pitch a shape that batters tend to punish. The horizontal movement has held in the playoffs, and more velocity is not messing up what was an already great pitch; the issue to watch out for is increased flatness. You can see that in the graph below: the more changeups stray from that -6 to -9 VAA range, the worse they tend to perform.

Despite his changeup and slider undergoing a lot of change, Urquidy has maintained a good combination of movement and velocity, and that extra heat on all three pitches may have played a part in his strong World Series start. We know that there are teams that prep for pitchers by using machines that can mimic whatever movement and velocity you could imagine, and every team is doing advance work centered around what a pitcher’s arsenal looks like. So what better way for a pitcher to counter than to be markedly different from what he was during the year — and on top of that to maintain a movement profile that has been successful nonetheless! I don’t know if Urquidy and the Astros have been consciously doing this, but in a postseason where starters are getting jumped and not lasting long, it’s no doubt useful to appear as something new.