ESPN Article on WAR

[qwikdraw]

http://espn.go.com/mlb/story/_/id/8959581/why-wins-replacement-mlb-next-big-all-encompassing-stat-espn-magazine

Great read, I've been following WAR (as most have you) for a while now, but this article puts a great perspective on things, especially the uneasy faith we have to put into certain websites.

[MikeM]

WAR is alright I suppose. Gives a very general sense of a player's worth IMO.

I don't trust the defensive metrics that go into the stat that can really balloon a player's WAR. I'm supposed to believe Bourn is the 13th best player in baseball? He had a 22.4 FLD WAR last year and a -6.4 the year before. WTF is that.

The Rangers led the league in Runs but were 9th in batting WAR.

COL and CWS were 6 and 7 in runs scored and 21 and 22 in batting WAR.

Obviously much more goes into baseball than just collecting high WAR players.

[Skin Blues]

WAR is alright I suppose. Gives a very general sense of a player's worth IMO.

I don't trust the defensive metrics that go into the stat that can really balloon a player's WAR. I'm supposed to believe Bourn is the 13th best player in baseball? He had a 22.4 FLD WAR last year and a -6.4 the year before. WTF is that.

The Rangers led the league in Runs but were 9th in batting WAR.

COL and CWS were 6 and 7 in runs scored and 21 and 22 in batting WAR.

Obviously much more goes into baseball than just collecting high WAR players.

Adam Dunn went from a 138 OPS+ to a 54 OPS+ in the span of one offseason. Ricky Romero went from a 2.92 ERA to 5.77 ERA. Same team, same defense, same opponents. The same could be said for any stat. I agree that defensive stats have wider error bars but they are still meaningful.

As to the second part, the Rangers, Rockies and White Sox all play in a high run scoring environments so their runs are inflated. The Rays play in a terrible run scoring environment and therefore have a much better rank in wRC+ than in runs scored. This is very useful information and shows the glaring deficiencies in just looking at plain old ERA and runs scored numbers. In our heads, everybody knows that a 4.00 ERA is much better in Colorado than in, say, San Diego. But when the difference isn't so enormous we have a hard time mentally adjusting for park/league. Advanced metrics do an excellent job of that.

[TwistedLogic]

+1 Skin on a perfect post. Exactly what needed to be said. @Mike, Your "WTF is that" on Bourn's defense is the same "WTF is that" that goes for Bautista's offensive spike in 2010 and the same "WTF is that" that someone out there must have said on Lincecum's incredible 2012 regression. Player's production fluctuates in all facets, year to year. If a player like Wells can play like crap one year, great the next, crap the year after that and good again the following year, why can't a player's defense fluctuate just as much from year to year? Your mention of the Rangers scoring the most runs proves yet another huge flaw in traditional ways of looking at the game. The Rangers score so many damn runs because they play in a bandbox. It's unfair to compare teams based on something as simplistic as "runs scored". In Arlington, you might need to score atleast 7 runs in every single game of a series to sweep the other team. In San Diego, you might get the sweep in 7 total runs over all three games. Both teams win three games, both teams make the same showing on the standings but one scored far more runs. Is that always because that one team is better than the other? No. It's oftentimes thanks to factors that were previously intangible, but now have a way to be calculated. Take the same Rangers team and move them to Target field and chances are, they'll score fewer runs.