WAR Doesn't Make Sense; 2018 Topps Edition

Beginning in the 2014 Topps set, the WAR or Wins Above Replacement statistic became a big thing not just for baseball cards, but for baseball as a whole. The statistic serves as a way to evaluate how good a player is while compared to the average player at that position. Case and point; Albert Almora and his 1.2 WAR in 2017 which brings me to my problem with WAR in general. 

Not only do I have no clue how Wins Above Replacement is calculated, but there are numerous examples of how faulty it can be. Maybe it's the home run-heavy era of baseball that we're currently living through, but I found multiple examples of WAR undervaluing some players while overvaluing others which I will show today. 

I'll start by comparing 2 closers; Roberto Osuna and Edwin Diaz. If you look at each player's 2017 stats (particularly innings pitched, ER, strikeouts, and ERA) they have basically identical stats. They each pitched roughly the same amount of innings, have very similar ERAs, and both gave up 24 earned runs. With that being said, Osuna's WAR in 2017 is almost 3 times greater than Diaz's. If the gap between the WAR was around .5 or lower, I'd understand. But such a drastic difference between the WAR of 2 players who put up almost the same numbers proves that WAR isn't always reliable. 

Additionally, when you look at a player like Dylan Bundy's ERA, it doesn't make you think of a very good WAR. Granted, there are more dingers being hit now than ever so a 4.24 ERA isn't considered as bad as it used to be. But with that being said, a 2.7 WAR is very generous for a guy with a slightly above-average number of strikeouts for his innings pitched and 152 hits given up. In my opinion, he still had a pretty good year, as well as he could do as a decent Orioles starter in 2017, so I'd give him a WAR closer to 1.7 than 2.7.

I'm sorry, but Greg Holland wasn't just a guy who started a couple games in 2017. Holland was a routine starter with 135 innings pitched for the White Sox and a 7-14 record. He likely had a spot in the rotation for most of the season, so there's no excuse for a ridiculous 6.20 ERA. Which leads me to my point about WAR. How is his WAR only -0.9? Nothing against Greg Holland, but how is it not much lower. I've seen position players and pitchers who've put together far superior seasons whose WAR is barely higher than Holland's in 2017. It really should be at least -1.5, maybe even -2.0.

For my final point as to why WAR isn't a very reliable statistic, I present to you Gerardo Parra's 2017 stats through his 392 at-bats. Not only did Parra have a .309 average in his limited playing time for the Rockies in 2017, but he had 71 RBI which is incredibly high for such few at-bats. Despite putting up very strong numbers, his 0.8 WAR certainly does not reflect what he did last year. Like I said before, I'm not sure of everything regarding WAR, but it seems like it doesn't do a good job at representing player's who didn't have very high at-bat totals and that's an issue. 

However, despite the points I've made against WAR thus far, I will say that it's not all that bad. Michael Taylor is one of MLB's most underrated players, especially due to his speed which is very undervalued nowadays. Last year, Taylor hit a respectable .271 with 19 dingers and 53 RBI and 17 stolen bases through 399 at-bats. While I do feel his 3.1 WAR is slightly excessive, I appreciate that a player of his caliber, often overlooked in today's game, is given the recognition he deserves because of the WAR statistic. 

Maybe WAR isn't too bad after all, but I'm still not too convinced.