Workshop: Pitcher Injury Factors

Projecting pitcher injuries and their effect seems like the Holy Grail for fantasy analysis. From years of research on the subject, I find it’s just a frustration filled enterprise with no firm resolution. Until a start to the 2020 season has been agreed upon (or I eventually find an acceptable answer), I plan to continuously grind for a workable understanding of pitcher injuries.

First, this article will be a work in progress as I try to find answers to various questions. I can’t fill the RotoGraphs article list with a new article every time I make a change or add more information (Ed. note: Sure you can, Jeff, we’ll post all of ’em!). Every few days or so, I’ll summarize the findings from the previous article’s work and keep moving forward. The series will come to an abrupt end if the framework exists for a start to the season since other analysis will then take priority.

I’ve started down this path already with a couple of recent articles on pitcher injuries. Here is a summary of the findings:

• Accumulation of previous IL days leads to increase aging and injuries compared to other pitchers.
• Accumulation of arm injuries leads to increase aging and injuries compared to other pitchers.
• High fastball velocity leads to an increase in injuries compared to those who throw slower.
• Age doesn’t seem to be a factor with increase injuries beyond the baseline rate.

I need to combine these factors and I don’t feel confident about the results at all, but that’s why I’m asking for help. Using the benchmarks from the two articles, a pitcher gets one injury factor for:

• Throwing over 93 mph.
• Having two or more career arm injuries.
• Over 120 accumulated IL days with a bonus point for over 460 days.

And here are the 2019 starters (min 10 GS) grouped by the total injury factors.

What seems off?

Two areas of exploration I see from just putting words to the ideas are age and recent injuries. I just need to make sure age doesn’t have a greater effect than is already factored in. Second, Carlie Morton near the top intrigues me. I wonder if a IL recency factor needs to be added in. I know that the two aren’t the only questions, but they’ll keep me busy for a bit. What else is missing and can be explored? Let’s find as many answers to this to pitcher injuries.

Append #1: Aging

I took all the Steamer Projections and look how it did at projecting IP and stats for different ages. I didn’t find much if anything. First is how the projections held up for players under and over 30 (projected GS/G >= .5 and IP >= 10)

The older starters underperformed their projections by five innings but generally overperformed in the talent stats when compared to the Younger hitters.

Innings: Actual Minus Innings Projected Grouped by Age

Rate Stats: Actual Minus Innings Projected Grouped by Age

I sort of see some trends … maybe.

6/19/20 update: While working with the Il days grouped by total days and the past two seasons, I found no signs that age was not properly accounted for.

Append #2: Previous Durability

The durable pitchers lose a few more innings but almost match their projections. It’ time to make a Morton clone. Years of injuries but a recent clean slate.

6/19/20 Update: I’ve been cutting and dicing the data based on the total IL days and the ones accumulated over the past two seasons. I ran the data while accounting for age. Age doesn’t seem to be an issue. Here is the most simplified data at this point:

While these starters with over 120 career IL perform worse, being healthy over the past two seasons seems to stop a decline in talent. Now, I just picked 30 days out of thin air. My next step is to see if there is a better designation point.

Update: And 30 days was a damn fine guess. I’m not sure where to go next.