[the 'stache]

Originally Posted by Tabe

This is the problem I have with WAR. The numbers seem so completely random at times. And there are some guys that seem to just tick all the right boxes for WAR and end up with a number that doesn't pass the smell test. To use a guy mentioned earlier in this thread - Gene Tenace in 1974 hit .211 with an OPS of .775. WAR of 5.0. Really?

Joe Morgan was a great player - for about 5 years anyway - and WAR just loves him. He ticks all the right boxes. But a WAR of 11.0 in 1975? OK. But then he goes out and improves in 1976 and his WAR *drops* to 9.6. Sure thing.

Agree with everything you just said.

The variance in his WAR between 1975 and 1976 appears due to defensive regression. He went from a 2.0 dWAR to a 0.4. But he won the Gold Glove both years, so the guys watching the games didn't see a precipitous drop off in his defense.

Sometimes I love WAR, but sometimes I hate it. I don't think there's ever been a more divisive metric in baseball. It's great for comparing players within one era, but I find it maddeningly (is that even a word?) flawed when you compare players from even one generation apart. And the defensive component of it....let me just say I don't get it, and I'll walk away.

The very fact that WAR is computed differently by BBR and Fangraphs only adds to my nervous tick. When it comes to statistical analysis, I'm a very precise guy (shocking, right?). I don't like that I have to qualify which site I'm taking my information from when entering into a discussion. The talking heads at BBR and Fangraphs need to lock themselves in a room, agree upon one formula, and then blow smoke for the world to see, like when the Pope has been named by the College of Cardinals.