Posted by loudawg10 on 10/4/2012 2:08:00 PM (view original):
the point of WAR is to be a tool, not a be all end all statistic to determine a players "value".
Fun with WAR part 2: TEAM WARS!
Houston 55 wins - team WAR 7.6. - a team of "replacement" Houston Astros wins 47-48 games
Chi Cubs 61 wins - team WAR 11. - a team of "replacement" Cubs wins 50 games
Wash Nats 98 wins - team WAR 41 - "replacement" Nats win 57 games
NYY 95 wins - team WAR 48 - "replacement" Yanks win 47 games
SF 94 wins - team WAR 35 - "replacement" Giants win 59 game
so the Yanks win 1 more game than the Giants with "real players", but if each team is stocked with "replacements" the Giants win 12 more games than the Yanks?
Um, that's exactly how it is supposed to work. Replacement level players are calculated to be waiver wire/AAA guys. A team of replacement players would be absolutely horrible.
The discrepancy comes from lack of context. Timing matters in games, it doesn't in WAR calculations (or batting average, rbi, slugging, OBP, runs scored, stolen bases, etc etc etc)