The initial problem I have with relief pitchers is defining them. I didn’t want to be looking at pitchers who split their time between a relief and starting role. I decided the pitchers needed to have at least 95% of their appearances as a reliever. Also, I didn’t know what to set the minimum innings limit to. Here is a chart of the DL% depending on the number of innings pitched the previous season. Matched season data from 2002-2003 to 2012-2013 was used.
|IP||Pitchers||Pitchers Who Went on DL||DL%||Total DL Trips||AVG Days per DL Trip||AVG Days lost per pitcher|
Not a huge difference in values until a pitcher reaches the 70 and 80 IP level. For all the rest of the analysis, I will use a 20 IP limit (33%) as a baseline. Basically one out of every three relief pitchers should expect to go on the DL for around 55 days. With the average bullpen being seven pitchers deep, the team will need to have two more pitchers who can fill in as the others get hurt (7 * 33% = 2.3 pitchers)
Additionally, it may be counter intuitive that the more innings a pitchers throws their DL chance drops. As I found with starters, if a pitcher has proven they can throw a bunch of innings once, they can probably do it again.
Now on to see about the age component. I looked at the DL chances and days lost with two year time intervals.
|Age||Pitchers||Pitchers Who Went on DL||DL%||AVG Days per DL Trip|
Some trends can kind of be seen. I went a little further and grouped the data into 5 year grouping where some dividing lines can be seen clearer.
|Age||Pitchers||Pitchers Who Went on DL||DL%||Total DL Trips||AVG Days per DL Trip||AVG Days lost per pitcher|
Two trends stick out now. Younger pitchers are less likely to go on the DL, but when they do, they are more likely to stay longer. The average number of days lost per pitcher (not DL trip) is relatively constant until a pitcher reach the big 40.
So age and previous innings pitched don’t really change a pitcher’s DL chances. Now on to an item which does, previous DL stints.
|DL Status||Pitchers||Pitchers Who Went on DL||DL% year2||Total DL Trips||AVG Days per DL Trip||AVG Days lost per pitcher|
|No DL year1||1746||517||30%||568||55||18|
So relief pitchers who went on the DL in previous season see a 43% increase in their DL chances and when they go on the DL it will be 16% longer compared to the baseline 20 IP values. This increase is important for a bullpen full of injured relievers (see the Rangers). If a team has all seven of its relief pitchers on the DL in the previous year, it will need three pitchers lined up to fill in (42% * 7 = 2.9).
Well that is it for now. Next up, I will run a logistic regression against the pitcher data to come up with a formula to give each pitcher a DL. Until that date, at least a person now should have a basic understanding of a relief pitcher’s DL chance and the main factor which can increase it, previous injuries.
Print This Post