Pitching while injured: A case study with Ben Sheets

With C.C. Sabathia and A.J. Burnett signed, Ben Sheets is one of the top free agent pitchers left on the market and expects to be very well compensated for his awesomeness. Sheets has been the Brewers’ ace for years now when he has been healthy. Sadly, again at the end of the season Sheets fell victim to tearing of the flexor mass in his elbow. Sheets began feeling pain in a start on Aug. 26 against the Cardinals but continued pitching through five more starts.

A case study in injuries

Injuries to pitchers can be the difference between a team that makes the playoffs and one that ends up below .500. Sometimes, the problem for teams is identifying when a pitcher is injured. Just looking at results like ERA can be misleading. Sometimes the pitcher just needs a trip to Triple-A to work on his mechanics or develop a different approach. Even if the pitcher says he has soreness in his elbow or shoulder, differentiating between typical soreness between starts and an actual injury is sometimes difficult.

One possible new method will be to look at a pitcher’s stuff using PITCHf/x data to determine if something is wrong. Potentially, comparing a pitcher when we know he is right to his current form might tell if he is injured. To study this it is important to look first at cases where we know a pitcher was throwing while injured. This will tell us first if this method is likely to be useful and then what we should focus on to be looked at proactively. Unfortunately, it is often difficult to get information about injuries from teams that have every reason to not be forthcoming. This is why Ben Sheets’ injured elbow is so useful. We know when he began to feel the discomfort and, because he continued to pitch, we have a decent-sized data set to examine.

A look into Sheets

David Gassko and Dan Loeterman already have done an excellent job of breaking down Ben Sheets, so I am not going to talking about his stuff other than to say that his main pitches are his four-seam fastball and his curveball. To examine these pitches over the course of the season, we can look at his wear pattern.

What this shows is his average fastball and one standard deviation for each of his 31 starts. The date on the bottom shows the day of the start, with Day 1 being Jan. 1. So Day 90 was the season opener against the Cubs. Day 238 was the start in which he began to feel the elbow injury; I have placed a yellow line between that start and the last five starts that year. The black dots in the middle of the chart represent the speed of his fastball for that day and correspond to the vertical axis on the left.

Generally, Sheets was consistent with his fastball; its average speed usually was between 93 and 94 mph. The blue and red dots represent the movement of the pitch and correspond to the vertical axis on the right. Sheets’ vertical movement with his fastball was also consistent, usually just above 10 inches, which is good for a four-seam fastball. Because Sheets throws mostly over the top, he has little horizontal movement with his fastball; that tended to be around four inches, though that jumped around more as you can see on the chart.

It really is remarkable to see how consistent Sheets was with his fastball and then how inconsistent it was after his elbow injury. While the game on Aug. 26 didn’t appear to be trouble, his next two starts he saw a big drop in his fastball speed, down to around 91 to 92 mph. After one game in which he seemed to have his good stuff back, he ended the year with two more substandard speeds.

If you look left a few games from the 26th, you can see several other starts in which Sheets’ velocity was down. Was this a precursor to him being injured or just his arm getting tired? I can’t say for certain, but Sheets did experience issues with his triceps in a start on April 18 (Day 108), which caused him to miss a turn in the rotation. In his next game, on April 29 (Day 119) it appears that he still was feeling the effects. In both starts, his velocity was down.

Also notice that the horizontal movement of his fastball when its velocity was down tended to be less than his regular four inches (less negative movement into a right-handed batter—more straight. Pitchers sometimes compensate for injury by adjusting their mechanics. Raising your arm angle as a pitcher will cause the ball to have more pure backspin and less horizontal movement. It appears that is what Sheets was attempting to do in his last few starts of the season, with some success on Sept. 11 (Day 254) but not much success in the two starts after that.

A similar effect can be seen with Sheets’ curveball.

This is the same presentation as the previous plot, only now we are looking at Sheets’ curveball. Because his curveball is very 12 to 6, it has little horizontal movement, so I have removed that from the graph to better see the velocity and vertical movement.

Sheets throws a very hard curveball, usually around 80 mph. After the injury, his speed dropped with the exception of the start on Sept. 11. He threw only two curveballs in his last start on Sept. 27 (Day 270), so that point has been removed from this chart. Also, during his triceps problem in late April his curveball again lost some of its bite.

Also interesting to note on this graph is that while Sheets’ speed is rather consistent, the vertical movement of his curveball had good days and bad days. Generally, it appears that Sheets started the season with a rather strong hook but in June and some of July it lost some of its drop. Later in the year, Sheets regained that movement, but in each time period there were games in which he had a better or worse curve than usual. This is probably not too surprising, as a pitcher probably has more days with his good fastball than days with his good curve.

Conclusions

While a study of one pitcher is far from definitive, it does appear that when Sheets was injured, his fastball and curveball both lost a few mph. It is interesting that in the 24.1 innings he threw after getting injured, Sheets actually had a strong ERA, so if he hadn’t let on that his elbow was bothering him, the injury might have stayed completely hidden.

I hope that by examining more pitchers throwing with injuries, and throwing once coming back from injuries, we can develop some potential warning signs to identify future injuries quickly.