David Pinto has an interesting take on a defense metric. He uses a probablistic model based on how hard the ball was hit, and the vector it was hit at. From this he calculates how many plays the players should have made compared to the number of plays he did make.  More can be found at http://www.baseballmusings.com/

Anyway, here are how the tribe's defenders stack up, with the number being the number of extra outs their defense provided over average (not replacement) defense. This us using the smoothed model.

1B - Broussard - saw avg before, can't find now
2B - Belliard - +9
3B - Boone - +40! (one of the best)
SS - Peralta - -28! (worse than Jeter, ecchh)
RF - Blake - +9
CF - Sizemore - +4
LF - Crisp - not posted yet

Michaels grades out better than Sizemore in CF (+10 in half the playing time) so he looks like an improvement over Crisp in left. 40 extra outs by Boone may expalin why Wedge stuck with him (or the lack of other options). Peralta made plenty of plays, this may be why he grades out fine in other systems, but this study says it is because he had many more opportunities (Westbrook anyone?). Or did Boone take some of Peralta's balls? (this may make more sense) By the way, Vizquel was not much better at SS. Third base is now not available in the future, so let us hope that Peralta can stay close to average.

Overall the tribe is 5th best in the majors, saving the pitchers roughly 120 outs. The previous year they were just average. Presumably, these improvements were from Boone replacing Blake who replaced Lawton (two for the price of one), and Sizemore allowing Crisp to shift to left. Peralta/Vizquel seems to be a wash defensively.