#13 Oregon (-3.5) 21 UCLA 17
Oregon is one of the best defensive teams in the nation, allowing just 4.1 yppl to a good schedule of offensive teams that would combine to average 6.0 yppl against an average team (including Boise State, Purdue and Cal). Oregon should have no trouble defending UCLA's mediocre offense (5.0 yppl against teams that would allow 5.0 yppl to an average team) regardless of which Kevin (Prince or Craft) is at quarterback for the Bruins (backup Craft has actually been better).
Oregon will be without starting quarterback Jeremiah Masoli, but most experts consider backup Nate Costa a better passer - although Masoli's scrambling will be missed. Oregon has only been 0.4 yppl better than average offensively, but that number has been better since their opening game loss at Boise State.
UCLA's defense has yielded just 4.3 yppl (to teams that would average 5.3 yppl), so the Bruins have an edge over the Ducks' attack unless Costa proves to be even better than expected.
My math model only favors Oregon by 2 1/2 points but the Ducks apply to a solid 125-51-1 ATS momentum situation. UCLA's Rick Neuheisel is 11-3 ATS as a home underdog in his career, so I'm not really interested in betting either side in this game. I do, however, like the under.