Burch Smith and the Problem of Holding Velocity

Right-hander Burch Smith has been traded from San Diego to Tampa Bay. “Will he start or not?” is a question a person might reasonably ask about that. What follows is an attempt to answer the question — in part, if not in whole.

At some point during during April or May of 2013, after the latter had produced some conspicuously excellent numbers with Double-A San Antonio, the present author developed a fascination with then-Padres right-hander Burch Smith — including that pitcher, for example, in multiple editions of the Fringe Five.

When Smith was finally promoted to the Padres, it was not unlike Christmas on May 11th. And even after Smith conceded six runs over a single inning in his debut, I remained curiously enamored of him.

Smith’s first three major-league starts didn’t weren’t a total disaster. He struck out exactly 25% of the batters he faced, for example, which is a very strong figure for a starting pitcher. He also threw this changeup once to Denard Span, which pitch offers quite a bit in the way of aesthetic pleasure, if nothing else:

Ultimately, though, Smith was unable to prevent runs. He allowed 15 of them (all earned) in just 7.1 innings over those first three starts and was demoted to Triple-A Reno before he could make a fourth.

There were likely a number of problems with Smith’s first three appearances, but at least one of them is illustrated by the three images below.

First this:

And then this:

And, finally, this:

Those are Smith’s velocity charts (care of Brooks Baseball) from the aforementioned starts — and what each illustrates is that, after touching 96-98 mph over the first 20 or 30 pitches of each appearance, Smith was descending more into the 92-ish range. In the one start of the three during which he recorded an out after the first inning, on May 17th, Smith was having difficulty reaching even 92 mph.

This was a problem: the enthusiasm regarding Smith’s minor-league success was founded in no small part on multiple reports which suggested that he was touching 98 mph. Anecdotally speaking, a pitcher who’s touching 98 mph ought to be sitting more comfortably in the 93- or 94-96 mph range. Which, Smith was doing that, but only for two or three innings’ worth of batters.

So, Smith’s inability to hold his velocity over those early starts seemed like a problem and also seemed to be more pronounced than for other pitchers. But what is the baseline for velocity loss over the course of a start?

Surprisingly, for how widely accepted the notion is that pitchers do lose velocity over the course of a start, there’s hardly any extant research or even documentation of the phenomenon. Probably the best work on the matter, though — brought to my attention by FanGraphs’ Jeff Zimmerman — was written by Jeremy Greenhouse for Baseball Prospectus in May of 2011.

Greenhouse found that, among a sample of major-league starting pitchers, that, while it was common to lose velocity over the course of a start, the average starter lost less than 0.5 mph over even 100 pitches.

Regard, a graph illustrating all that information:

Only four pitchers in Greenhouse’s sample lost as much as 2 mph over the course of 100 pitches: Tommy Hunter (-2.0), Zach Duke (-2.2), Rick Porcello (-2.5), and Jonathan Sanchez (-2.7). It’s perhaps unsurprising to find that two of that group have moved to the bullpen, while a third (Sanchez) has had considerable difficulties since Greenhouse published his piece.

Courtesy Greenhouse, here’s a graph of Sanchez’s fastball velocity plotted against pitch count:

As Greenhouse notes, Sanchez represents essentially the worst-case scenario in terms of lost velocity over the course of a start. Keeping that in mind, regard the following table, which features the average velocity over ten-pitch intervals for Burch Smith during his seven career major-league starts (i.e. the aforementioned three May starts from 2013 and then four additional ones from September of that year):

And the attendant, hastily made graph:

As the above data illustrate, Smith has lost nearly 3 mph on his average fastball velocity between his 1st and 100th pitches. Which is to say, pretty similar to the amount lost by Jonathan Sanchez, according to Greenhouse’s study. Jonathan Sanchez, that is, who represented the worst-case scenario for velocity loss over the course of a start.

The Rays, who employ a collection of people who are smarter and better than me, are likely aware of the challenges facing Smith. They’ve either identified possible solutions to Smith’s loss of velocity, or otherwise are content allowing Smith to work at 94-96 mph as a reliever. If Smith pitches in starting capacity for the Rays, however, and also exhibits the ability to maintain his velocity over a 100-plus pitches, that will be a development — at least so far as his major-league resume is concerned.