Oliver Perez and Release Points
NOTE: This is a guest post by Matt Himelfarb.
Jerry Manuel deserves credit for usually putting a humorous spin on another of Oliver Perez’s dreadful outings. I always picture him lying back in the manager’s office at Citi Field, his hands clasped behind his head, telling reporters “Ollie’s in a funk right now” and sounding a bullish note about Perez’s eccentrics with an autographed Jonathan Malo photo in a gold-plated frame in the background.
Manuel didn’t have to defend Ollie on Friday night, despite a crushing 6-2 loss to the Padres. Since returning from the DL on July 8th, Perez has been hampered by control problems; entering his last start, he compiled 26 walks in 27 innings of work, giving him a 20% walk rate. To put that in perspective, among qualified pitchers, Pirates southpaw Tom Gorzelanny has the highest walk rate in baseball at 14.29%
As the convenient SNY narrative goes, Dan Warthen advised Perez to take an extended pause when he lifts his leg. As a result, Perez’s mechanics were smooth on Friday night, enabling him to allow just two walks over 6.1 innings of one-run ball.
I’m no pitching guru, so I really wasn’t all that skeptical about this. Using pitch f/x info, I looked for any concrete adjustments Perez made. The first thing I looked at was the consistency of Perez’s release point. The general consensus is that a consistent arm slot contributed mightily toward Perez’s solid outing. After all, control is all about finding a solid, comfortable release point.
To do this, I took his vertical and horizontal release points, and found the standard deviation between each. I did this for his Friday night start, as well as all his other walk-laden outings since he came back. (For those not familiar with this, Josh Kalk wrote a nice primer over at The Hardball Times last September concerning repeatable release points and pitcher effectiveness.)
To begin with, having almost two inches of variation is undoubtedly below average. Despite all the talk of keeping his front shoulder closed, his horizontal release point is still a major problem. That being said, if you look at his vertical release point, his last start certainly wasn’t his worst outing in terms of consistency. Regardless, as you can see, it does not appear it should make much of a difference anyway. His release point was actually more consistent against Colorado, during which he walked four batters in five innings, and against the Dodgers, where he allowed seven walks in five innings.
Next, I went to see if his actual release point changed throughout all his outings:
Again, Perez didn’t seem to make any adjustments with his horizontal release point. It does look, however, that he lowered his vertical release point 2-3 inches. I found this kind of surprising, since we usually associate Ollie’s wildness when he classically drops his arm angle down.
[NOTE by Joe: This could be a flaw in the two-dimensional nature of the pitch f/x technology. When a pitcher throws more overhand, he must push his hand further forward and DOWN in order to direct the ball to the strike zone. That said, it makes sense that the release point was lower. (This path tends to create more leverage, momentum, and velocity, BTW.) The pitch f/x charts don’t account for depth, only vertical and horizontal axis points.]
To be honest, I don’t if or how this would affect his command. I’m just logging the results. Going back to the pause, I do know that properly planting your plant foot toward the plate plays a major part in throwing strikes- although I guess in turn it would reflect his release point. [Note by Joe: Matt is correct on both points — see Isaac Newton for more info.]
If not, though, there are reasons to think Perez’s successful outing was a mirage, and doesn’t say much about him going forward. For one, the Padres can be your prototypical collection of overly aggressive youngsters. They are ranked 14th in the National League in OBP. (.313), just two points ahead of Cincinnati and San Francisco, who are tied for last, and are ninth in walks (one ahead of the Mets in fact). To their credit, they take a decent amount of pitches per plate appearance (3.85 P/PA- league average is 3.82)
Friday night, they seemed very patient (4.04 P/PA). Perez’s average P/PA since returning, however, is 4.08.
Of course, maybe Perez was simply due for an outing like this; a 20% walk rate is probably unsustainable over the long run, even for Ollie. If he can string a few good outings together, we should be able to make some better observations.