Steve at Was Watching poses the following question:
Since July 8th, Phil Coke has pitched in 18 games for the Yankees – throwing 13.6 IP and facing 61 batters. In the process, he’s allowed 22 base runners. His ERA during this time is 9.22 and the opposing team’s BA/OBA/SLG line is .315/.377/.519 (against him).
Are you nervous about Coke? Based on these numbers, maybe we should be?
A closer look at the numbers suggests that two awful outings are skewing these numbers. Against the White Sox and Angels, he allowed 11 of those baserunners and 10 ER in 1.1 innings. In the remaining 12.1 innings over that span, he allowed just 11 baserunners. Looking at the numbers on relievers outside of their context can often lead to faulty conclusions. My question is, can we develop a stat to judge the reliability of relievers, something that can better provide us with information on a reliever’s consistency and his ability to notch scoreless appearances? Maybe we could grade each performance on a scale, add up the total grade and then divide by appearances. Do you have any ideas on this front?
If you enjoyed this article, please consider sharing it!
4 Responses to Discussion: Finding A New Stat For Relievers
Leave a Reply
TIQ IQ
Bill James was saying in the 80s that ERA wasn’t a useful stat for relievers. I’m not sure there can be a simple formula to rate them given the different aspects of the job.
You can try to use the variance or standard deviation off an ERA and WHIP to see how close to thos average numbers they really are. Players with high variances are hit or miss and players with low variances tend to consistently pitch to what their numbers are.
I have a system I use to rate all Yankee pitchers (majors and minors): I give a rank from 1-5 for each performance using the box score as reference. I know it’s not perfect, but it’s been a fairly good indicator of who best to promote and demote. Right now Phil Coke has an average performance of 3.93. Just to compare, Phil Hughes in relief is at 4.70, Mariano is at 4.49 (he was at 4.30 last year) and Bruney is at 3.70. The “best” AAA reliever has been Sanit at 4.34. I’d love to refine this someday…
This is the kind of thing I was thinking about.