A Baseball Weblog

Wednesday, August 3, 2011

Another new curveball from Hughes

Phil Hughes had what was probably his best start of 2011 yesterday, throwing six shutout innings while walking none and striking out four.  If it hadn't been for the rain during the seventh inning, Hughes certainly would have pitched more, since he only had 66 pitches in his outing.  In addition to hitting 95 mph, something that we had not seen this year, Hughes unleashed a new variation on his curveball, which he seems to be constantly tweaking.  The first one he threw was a nasty one with two strikes to A.J. Pierzynski that registered at 82 mph, which is harder than Hughes has ever thrown his curve in the majors. Hughes confirmed after the game that it was indeed a hard curveball, as opposed to the slider he incorporated in March and ditched after two starts.  This is already the fourth time there's been some detectable change in Hughes' curveball:

  • During Hughes' dead-arm period in April, he was using the same spike grip on his curve that he had in 2010, but his lack of arm strength only allowed him to throw the ball in the low-70s, more than 4 mph softer than his 2010 average.
  • After his reinstatement from the DL on July 7th, Hughes started throwing his curveball harder, though not as hard as he did in 2010.
  • Before his July 17th start against the Blue Jays, Hughes announced that he would be switching to a new grip without a spiked finger in order to get a tighter break and more velocity, though the velocity didn't change a whole lot over his next three starts.
  • Yesterday, Hughes showed a pitch that looks more like a power curve, averaging around 80 mph.
This chart shows the vertical spin deflection (the number of inches the ball drops relative to a spinless pitch) plotted against velocity for Hughes's curveball this year.  You can see that the new grip (starting July 17th) gave the pitch less movement without much added velocity until yesterday.

What's strange is that the pitch he was throwing yesterday looks like what he was trying to throw since the 17th, but the velocity spike didn't come until now.  My guess is that Hughes sticks with this new variation of the curve, since he's said that he's been trying to add velocity to it.


  1. Have you ever written on the question of how accurately the pitch categories describe the variety of pitches thrown? In other words, are sliders more alike than they are like changeups or hard curveballs? A statistical technique called factor analysis could be a useful tool.

  2. That's a really good and important question. You can only do so much with labels, and the labels don't tell you anything about the movement/velocity variance of a pitch type (which if I'm getting you right, you're essentially talking about the variance within a pitch type name, like slider, curve, change, etc). I haven't seen extensive research on it (hopefully I haven't missed it), and it's certainly something that should be done.