/cdn.vox-cdn.com/uploads/chorus_image/image/52662647/613572590.0.jpg)
The problem we face is that the two main measurements of overall pitcher ability approach the task from two different directions. Fangraphs WAR (fWAR) bases its number for pitchers on xFIP, a number calculated from a pitcher’s home-runs, walks and strikeouts only. The logic behind this decision is that those are the only three things which are strictly down to the the hitter/batter relationship. If you start looking at hits, runs, earned runs, etc. then those are dependent on how good the defense playing behind them is. If you’re trying to measure how good a pitcher is, you want to avoid that.
It’s true that xFIP has been shown to be a better predictor of future performance than regular ERA, since (for example) it removes the results of BABIP luck. But is it a better measure of actual performance? I’m less convinced, because I tend to think that, for example, generating weak contact is part of a skill-set. Baseball Reference WAR (bWAR) embraces defense, saying “Our view is that, while the pitcher may have been unlucky or lucky in certain ways, we are trying to measure the value of the recorded performance--not its repeatability.” Thus, the base for their numbers is different. Think of it as roughly based on ERA, and I’ll leave it at that: details here for the interested.
What this means, in effect, is that the same pitcher can be evaluated radically differently. fWAR values high-strikeout pitchers much more. To some extent, this makes sense, because high-K guys usually have low ERAs: of the five qualifying NL pitchers with the best K-rate in 2016, four had ERAs below three. However, the fifth was on the 2016 Diamondbacks in Robbie Ray, whose K-rate of 11.25 trailed only Jose Fernandez, but had an ERA of 4.90. bWAR ranked him only seventh on the D-backs this year, below Matt Koch. fWAR had him first, ahead of Zack Greinke in second by a margin that was greater than Ray’s entire bWAR value.
When it comes to picking five nominees, however, it’s a problem. Because the top five by bWAR is mostly people that are not in the top five by fWAR, and vice versa. There are only two common names, Greinke and Archie Bradley. Everyone else finishes outside the top five by the other metric: two of Arizona’s top five pitchers by fWAR, aren’t even in the top twenty by bWAR. For the record, however, here are the top 10 by both fWAR and bWAR. The numbers in brackets are the value and rank by the other metric.
bWAR
- Zack Greinke, 2.3 (2.2, #2)
- Brad Ziegler, 1.2 (0.5, #6)
- Jake Barrett, 1.0 (0.2, #12)
- Archie Bradley, 1.0 (1.8, #3)
- Matt Koch, 0.9 (0.3, #11)
- Rubby De La Rosa, 0.8 (0.4, #8)
- Robbie Ray, 0.7 (3.0, #1)
- Randall Delgado, 0.5 (0.2, #13)
- Tyler Clippard, 0.4 (0.1, #15)
- Tyler Wagner, 0.4 (0.4, #10)
fWAR
- Robbie Ray, 3.0 (0.7, #7)
- Zack Greinke, 2.2 (2.3, #1)
- Archie Bradley, 1.8 (1.0, #4)
- Daniel Hudson, 0.6 (-0.6, #25)
- Shelby Miller, 0.5 (-0.7, #26)
- Brad Ziegler, 0.5 (1.2, #2)
- Patrick Corbin, 0.5 (-0.9, #29)
- Rubby de la Rosa, 0.4 (0.8, #6)
- Andrew Chafin, 0.4 (-0.5, #21)
- Tyler Wagner, 0.4 (0.4, #10)
So, pick the bones out of this lot, and try to figure out which five names should make it onto the final ballot. At this point, I’m not even going to make any suggestions, and am entirely open to any convincing case!