MacPhail:"this week we would put the finishing touches on" the roster."

[Frobby]

I agree that every personnel move has a context and a ripple effect (if that's what you're saying). And how a player is used helps to determine how valuable he is. So if Miggy allows the O's the cut Andino loose, and that allows the O's to keep another pitcher in the bullpen, and keeping a 13th guy in the pen keeps the bullpen from reaching that tipping point of overuse and exhaustion in mid-August...

...I can see the possibility that he would be valuable beyond his numbers. It's hard to refute. (Also hard to prove).

And extreme success in high-leverage situation can mean that a player has an actual W/L effect greater than his totals imply. (A really really good NL pinch hitter, maybe).

But in general, you have to compare a player to the player he replaces, or to the player who replaced him, or to the faceless baseline player who would be plugged in his spot if he weren't there. And in general, WAR seems to be a good way to do that.

I'm not suggesting there's anything better available. I just think we need to keep in mind that baseball is not a set of WAR numbers. Sometimes one trade unlocks the potential of a team in ways WAR never would suggest.

[El Gordo]

I'm not suggesting there's anything better available. I just think we need to keep in mind that baseball is not a set of WAR numbers. Sometimes one trade unlocks the potential of a team in ways WAR never would suggest.

Don't know what Frank's WAR was in '66, but I bet he was worth more than whatever it was, to the O's.

[DrungoHazewood]

My gut instinct is that this is a little optimistic. For one thing, you've done the pitching based on fan projections, and I'd like to know if the fan projections are better or worse than what CHONE and the other systems say. My guess is that the fans are on the optimistic side.

In a post today on Tango's blog it was mentioned that the fan forecasts end up something like 30% high (in WAR, based on recent historical data), so that the fans forecast an average team to win 94 games. So you'd probably do well to cut the O's number here by about 1/3. Which leaves them in the mid to high 70s.