Last season, San Diego St. was 69th in Ken Pomeroy’s rankings, and 71st according to Sagarin’s Predictor. This year my model predicts they will jump up more than 25 spots in the rankings, improving significantly on both offense and defense. But with many people I respect tremendously picking the Aztecs to win the MWC this year, I am left scratching my head. Should my model be even more optimistic?

The next table shows the five luckiest teams in the NCAA tournament last year (excluding teams in the auto-bid range 13-16). I am using Ken Pomeroy’s definition of “luck” which indicates teams that won more games than their per-possession numbers would have predicted.

In the pro sports, lucky teams are the low-hanging fruit of prediction models. Teams with poor point differentials are usually a good bet to fall back the following season. The NFL’s Kansas City Chiefs went 7-9 last year but were outscored by 126 points. Not surprisingly, the Chiefs are off to a 1-5 start this season. The NBA’s Portland Trailblazer’s earned the 6th seed in 2010-11, but the team had the 9th best point differential in their conference that year, and the Trailblazers were forced to throw in the towel mid-season last year.

Lately, a number of results are reminding us that the Pythagorean standings are not everything. First, the New York Giants won the Super Bowl after being outscored in the regular season. Then, the Baltimore Orioles made the MLB playoffs after being outscored for most of the season. But opponent adjusted margin-of-victory is still very powerful. Much like the existence of bad beats in Poker doesn’t obviate the fact that Poker is a game of skill, a few teams bucking the trend shouldn’t convince us to doubt the general wisdom of margin-of-victory.

There is plenty of evidence to support the use of margin-of-victory in college basketball. As just one example, if we take the luckiest at-large level NCAA tournament teams in the last 10 years, the trend is overwhelming. 80% of the lucky teams either missed the NCAA tournament the following season, or had a worse seed. The teams that bucked this trend are listed in the next table:

Many of these teams were still disappointing in a relative sense. Syracuse was 6th nationally in the 2005 preseason poll, but earned only a 4-seed. Memphis was 11th nationally in the 2012 preseason poll, but earned only an eight seed. North Carolina was 1st nationally in the 2012 preseason poll, but never lived up to their preseason billing (even before the injuries hit.)

But one thing that helped these teams at least improve on the previous season was bringing back a lot of talent. In the last decade, the average D1 team returns only 60% of its minutes, and 59% of its possessions. And as the table shows, the teams that improved their NCAA seed a year after a “lucky” season, all had higher than typical returning minutes.

And that is why Creighton, Colorado St., and San Diego St. should still be smiling. With returning superstars like Creighton’s Doug McDermott, Colorado St.’s Pierce Hornung, and San Diego St.’s Chase Tapley, the top three teams in this column not only return major minutes, they return talented minutes. These teams also welcome exciting new players. Colorado St. adds transfers Colton Iverson and Daniel Bejarano from Minnesota and Arizona.  San Diego St. adds transfers James Johnson and Dwayne Polee from Virginia and St. John’s. And thus even if last year was a little bit of smoke and mirrors, this year still brings plenty of substance.

My only suggestion is to be cautious. On paper, San Diego St., Creighton and Colorado St. should all be better than last season. They should be a bigger threat to go deep in March. And yet they could have an equivalent or slightly worse win total than last season. These teams all deserve some hype, but perhaps not quite as much as they are getting.

Follow @DanHanner