MASH Report (9/1/16): More On Exit Velocity & Hitter Injuries

Last week, I examined hitters who had a known injury they were playing through and then determined how their Stat Cast exit velocity changed. Today, I am moving on from back looking at injuries (reactive) to forward-looking (predictive). Any important conclusions drawn this early in the process should be taken with a grain of salt, but I will trudge along on anyway.

One issue I will continue to run into when looking at this data is the lack of data. Having less than two season’s worth of data is frustrating to work with. Five to six years from now, we will likely be laughing at the work that was done with so little information. Just be skeptical of the information provided as I am trying to be at the tip of the spear so I am likely to get cut a few times.

The more I looked at the data, it seemed a large drop of around 5 mph to 6 mph was indicative of a major injury. The one problem I ran into was players like A.J. Pollock who have both speed and power. Charlie Blackmon and Starling Marte as other similar players. They have wider normal ranges in exit velocity, so instead I went with looking at a half of the standard deviation of the player’s exit velocity to help flag injured players. Most of the time this value is around 5 to 7 mph.

The second issue I dealt with is how to select a time frame to examine for the decline and how to weigh the drop. For now, I am going to keep the process simple and just look back over the past week for the average exit velocity. I thought of some other weighting options, but the coding for them would be quite time-consuming. I will start small for now and if I find any bias’s in the data, I will adjust them at a later date.

Another related issue to two of the previously stated issues is the correct time frame in which to compare the weekly exit velocity. With less than two years of data, I will use all the data available and weigh it equally. Maybe in the future, I can determine some yearly weighting, but the data is not available yet to test.

So, I went ahead and looked for possibly injured players. The full list is available here, but here are the players with over 300 total batted balls from 2015 and 2016 and five this past week who have seen their average exit velocity drop enough using the above criteria. The 300 and five batted ball values are again pulled out of thin air and could see them adjusted over time. Besides the table, I tried to find any recent injury information on the players listed.

• Jonathan Lucroy and Stephen Vogt: They are catchers and I always expect them to be hurt. The list of players who just missed the cut was littered with catchers (Zunino, Ianetta, Chirinos, and Saltalamacchia)
• Jonathan Schoop: I found nothing on an injury, but he has been struggling with the bat hitting .250/.276/.393 over the past week.
• Christian Yelich: He had some mid-season back injuries which could be flaring up, but nothing being reported now.
• Yasmany Tomas: He has been dealing with back stiffness.
• Nick Markakis: No injury information, but is really struggling at the plate hitting .231/.355/.308 this past week.
• Brock Holt: Just returned to the lineup about a week ago after an oblique injury.
• Neil Walker: He has been dealing with a back injury since June and just went on the DL for it.
• Gerardo Parra: Came off the DL in early August from a sprained ankle. Also, he just made the cutoff of 5 batted balls. Looking at the past two weeks, his average exit velocity for the time from 85.7 mph.
• Eddie Rosario: Nothing on an injury.

The results aren’t as good as I would like, but the player and/or team could be hiding the injury from the public. I have some ideas like looking at contract rate as another factor, but I will wait for now.

Besides trying to find players using the injury rate, I can use player comments and see how the Injury value holds up. Here are some more players who have recently admitted to playing through injury and their five, 10, and 15-day Injury values (just doing a little testing here for the right length to examine).

Lorenzo Cain (wrist)
Days: Injury
5: 1.8
10: 6.2
15: 5.8

Cameron Maybin (thumb)
Days: Injury
5: 10.4
10: 9.6
15: 9.6

Albert Pujols (foot)
Days: Injury
5: 10.0
10: 8.6
15: 8.6

The Lorenzo Cain injury and exit velocity drop fits the time frame. With Maybin, the thumb may not be affecting him or the injury could be limited in scope. With Sir Albert, I just assume he has always been hurt and his overall exit velocity is lower than previous values. The lack of information is again showing its ugly head.

I am done for this week. I will play around a bit with the numbers and get back later with some more results. Additionally, I need to change the scale of the Injury value to be more understandable. Finally, the “injury value” needs a name. Any suggestions?