Chris Sale, before and after 75 pitches

Chris Sale is a fun pitcher to watch. He throws three solid pitches and has a plus fastball. He is one of those oddly proportioned left-handed pitchers, and he made the transition from reliever to starter look easy. As is not the case with most who share his physique, we currently have the luxury of seeing him pitch more than one or two hitters per outing.

Sale has been great as a starter, and has since silenced most of those who criticized his move from the bullpen. Sale’s transition to his current role was was aided by the fact that he already was already able to throw three pitches at the major league level as a reliever.

While he has been able to make it work as a starter without relying so heavily on his fastball, some may still believe that he doesn’t have the ability to go deep into games. Most of the damage done against Sale comes after he crosses the 50-75 pitch mark. Through his four starts so far in 2013, the trend has persisted. If you look at the table below (courtesy of Baseball Reference), you’ll notice this trend.

image

Sale strikes out more hitters and exhibits slightly better control as the game progresses, and his peripheral skills suggest that he should actually be a better pitcher after 50 pitches. His opponent OPS suggests otherwise, though.

The White Sox decided not to impose an innings cap on Sale in 2012, and manager Robin Ventura frequently allowed him to throw more than 100 pitches in a start. It is possible that White Sox know something that we don’t, because it doesn’t appear that Ventura has any plans to treat his ace differently in the near future. If we look at Sale’s mechanics over the course of a start, we might be able to better understand why the White Sox are so comfortable letting him pitch deep into games.

I’ll start with Sale’s release point. On the surface, his delivery looks pretty erratic. He slings the ball from a somewhat unconventional three-quarter arm slot, and some were worried about his ability to repeat this delivery as a starter. He says that he has adjusted his delivery in an effort to reduce the probability of injury. Sale throws across his body, and he adjusted the location of his landing foot by “three or four inches.”

In the graph below, I break down Sale’s release point by pitch count. I lumped his fastballs together and set other pitches aside to isolate the possible effect of fatigue on his release point. To be clear, this graph represents Sale’s four-seam and his two-seam fastball. The horizontal and vertical lines represent the average release point for each pitch count level.

image

It appears that Sale’s release point might actually be pretty consistent over the course of a start. A better model for consistency, however, is release point variance. In graphical terms, variance essentially measures the distance between the different points on the graph. The chart below compares variance levels across pitch count for Sale.

I did this same type of analysis last year on Daniel Bard’s stint as a starter, and found that his release point variance increased significantly as he approached 90 pitches. For comparison, I’ve also thrown up a chart of Bard’s horizontal release point variance last year. My sample size for Bard is much smaller than it is for Sale, and this explains part of the difference in baseline variance between the two (because variance decreases as the number of observations in sample increases).

image

Inconsistency in Sale’s horizontal release point over a start doesn’t increase quite as much as Bard’s does, but it is around 25 percent higher after 90 pitches. This might concern those banking on Sale to remain healthy and log 200+ innings this year and beyond. Bard’s release point variance jumped more than 45 percent as he approached 90 pitches, and we all know what happened to him.

What about movement?

The good news for Sale is that his “stuff” doesn’t worsen as he throws more pitches. The graphs below demonstrate that Sale is just about as effective on the 90th pitch he throws as he is on the first. Horizontal and vertical movement is plotted on each axis, and the lines represent averages for the different pitch count groups.

I left the change-up chart out of this article, because he is consistent in his release point for that pitch. The slider becomes a little flatter (good sliders have little horizontal movement), but not by much. The difference in movement between a Sale slider after one pitch and a slider he throws after 100+ pitches is only about an inch horizontally.

image
image

So what are we to do with this information on Sale’s response to fatigue? On the one hand, Sale’s effectiveness doesn’t seem to be affected by pitch count. Coaches monitoring Sale throughout a start are likely making the correct move in leaving him out there for the long haul often. The disproportionate number of runs and extra base hits coming in the late innings are probably just a result of randomness. Sale suffers from a high late-inning BABIP, and that would help explain why it looks like he struggles as he tires.

On the other hand, release point variance can increase the likelihood of injury. In 2011, Kyle Boddy determined that pitchers with higher vertical release point variance levels suffered from elbow injuries more often than those with lower variance levels. I don’t have any evidence to back this up, but I’d guess that varied release points cause even more damage when an arm is fatigued. Boddy didn’t find any issues with horizontal release point variance, though, and that was where we saw changes for Sale.

A Hardball Times Update
Goodbye for now.

If Sale’s vertical release point starts to move around, I’d hit the panic button. But for now, I don’t see any reason to worry about Sale’s ability to pitch deep into games. Rest easy, South Siders.


Noah writes for The Hardball Times. He has previously written for Beyond the Box Score and Fire Brand of the American League. Follow him on Twitter @woodwardps, or email him here.
3 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Kyle Boddy
10 years ago

Correct. How is this different compared to another population? Would be interesting to see.

MikeS
10 years ago

This is only meaningful if you compare it with other pitchers.

Noah Woodward
10 years ago

I’m in the process of working on “league averages” for this sort of stuff. It is tough work, because PITCHf/x data doesn’t come with pitch count information attached.