It should not be the sole determining factor, on that we can agree. But if your statement goes further than that, and that it shouldn't even be one of the factors, then mind explaining what aspect of bWAR for pitchers makes you feel that? bWAR for pitchers has a black box feel that often doesn't pass the smell test. Look at Chacin's numbers and Arrieta's numbers and explain why Arrieta is a 3.3 bWAR and Chacin is 1.9 bWAR. Chacin has better numbers for Games, IP, ERA, FIP, WHIP, K9, H9, HR9, and WP. Arrieta leads only in BB9, K/BB and HBP. The reason I specifically mentioned pitching bWAR is because it's something that really isn't a black box, bb-Ref quite clearly outline how it's calculated themselves. And their version of WAR is based on RA/9, trying to separate the pitchers part in that from the defense but still looking at the result, not how (i.e high strikeout numbers or some other way) they got there. In the case of Arrieta and Chacin the difference comes from adjusting for defense. Chacin has had the second best defense in majors (by DRS) behind him, Arrieta has had the absolute worst. Brewers are at +85 DRS, Phillies are at -112. So essentially it's RA/9 adjusted for defense, park and opposition. Which, other than extreme outliers, means that pitchers with a low ERA and a lot of innings pitched will accrue a lot of WAR. ERA and IP being two of the primary stats many people use for Cy Young consideration. So I wanted to see what the arguments were behind it. Wanted to see if it was a case of, say, trashing xFIP as a metric but in the same breath praising K/BB ratio. Baseball Prospectus' Deserved Run Average (Which is the basis for their WARP for pitchers) takes a rather different approach, trying to incorporate absolutely everything into it. But that rates Oliver Drake as one of the best performers on the Brewers this year, so I don't think most Brewers fans would prefer that. That one is pretty simple. Defense and positional adjustment. Braun was better offensively, but not by all that much. He's obviously the better hitter, but he had a down year last year and Sogard had an unlikely career year. bWAR is descriptive, not predictive. Braun played poor defense in left field, which is near the bottom of the defensive spectrum. Sogard played average to above average defense at 2B and average to below average defense at SS, two of the more important defensive positions. One way to look at it would be how much better or worse the team would be if Braun played SS/2B and Sogard played LF. Which is why it's a good idea to, when they differ a lot, take the average. When it comes to position players, differences nearly always stem from defense. DRS (used in bWAR) tends to produce more extreme differences between good and bad defenders than UZR, even when they both agree on who the best and worst defenders are. Generally though the problem with WAR isn't so much WAR itself, but how it's used. And a lack of understanding of what it is and how it's calculated. Small differences in WAR over short periods of time should be ignored (Which their creators also are very clear with), but larger differences over time are a whole other matter. That's largely due to defensive metrics needing a long time to stabilize. What WAR essentially does is package together offense, defense, baserunning and playing time to give an overview of roughly how much a player contributed. It's a way to compare a baserunning phenom, a defensive wizard and Chris Carter. It's a good starting point, you can then look further to see in what manner they accrued those WAR. Lorenzo Cain isn't an MVP candidate because he leads the league in bWAR. He's an MVP candiate (or should be) because he's good at the plate, a good baserunner and a good defender. WAR simply sums that up, and shows how valuable that is. It's also evolving all the time, and as defensive metrics get better it will get more accurate. Statcast data, when/if it gets implemented, will likely be a big step forward.