From SI.com:

There is a lot of optimism concerning this year's UCLA team, and many pundits point to the team's 20 returning starters. However, if I remember correctly, there was a Purdue team a few years ago that was a popular preseason pick because it was returning a boatload of starters, but they ended up having a disappointing season. Can you please tell me that won't be the case for UCLA this year?
--Steve, Brea, Calif.

Stewart Mandel's Answer:

I can assure you of no such thing, as I have a hard time putting too much faith in any Karl Dorrell-coached team, but I can tell you that the reasons behind the Bruins' rosy outlook are slightly different from those of that 2005 Purdue team. As I recall, the Purdue team you're referring to did return a lot of starters (including all 11 on defense), but the single biggest reason everyone was so high on them that year was because they'd somehow had a schedule devoid of Ohio State and Michigan (Iowa now has the same luxury this year and next). Obviously, that didn't help, because Purdue wound up finishing 5-6 -- the only losing season in Joe Tiller's tenure -- primarily because a) QB Kyle Orton was not among those returning starters; and b) Those 11 defensive starters, it turned out, stunk.

UCLA, on the other hand, fielded an impressive defense last season that finished sixth nationally in sacks (40) and 35th in total defense. The Bruins did lose one stalwart in defensive end Justin Hickman, but everybody else is back, so it's reasonable to expect a similar performance. That said, UCLA did go 7-6 last year, its offense was inconsistent and that vaunted defense disappeared in the Emerald Bowl. Therefore, you've got to believe almost all of the Bruins' preseason hype is based on that USC upset last December. (Though strangely, that same game seems to be having no effect whatsoever on the Trojans' expectations.) If the Bruins play like they did against USC every week this year, then heck, book their trip to New Orleans, but Dorrell's track record suggests they'll probably lose at least one or two games they shouldn't. I'd consider a preseason ranking in the 20-25 range to be reasonable.

Emphasis Added.