Discussion

Top teams that are vulnerable to upsets

Updated: March 13, 2013, 7:42 PM ET
By Peter Keating

Ahhhhhh. Did you hear that relaxing sigh? It's the top seeds in the NCAA tournament breathing easier. But not every Giant can afford to kick off its sneakers.

First, the good news for Goliaths: The past 96 hours have been brutal to would-be Davids, as eight mid-major and small conferences lost their shot to send their best Giant Killer to the big dance. On Friday, we wrote that our Giant Killers statistical model was showing historically high odds, from 27 percent to 30 percent, that a No. 16 seed would, for the first time, knock off a No. 1 seed.

But while Robert Morris, Northeastern, Stony Brook and Charleston Southern all had the potential to be unusually strong 16-seeds, each of those regular-season champions lost in its conference tournament. A little further up the brackets, Middle Tennessee (GK Rating: 25.9) got knocked off too. And while we identified Detroit (33.3), Canisius (24.2) and North Dakota State (18.3) as killer candidates, none of them sneaked off with auto bids.

In their place: the Western Kentucky Hilltoppers (whom we liked in 2009) and a bunch of awful teams, like James Madison, Liberty and LIU. Not to beat a dead horse, but the quality of NCAA tournament teams is going to keep taking a hit as long as conferences keep sending championship-week flukes instead of their regular-season champs to the Field of 68. Our model estimates that Stony Brook was more than 8 points per 100 possessions better than any other team in the America East this season, but that conference's silly postseason rules forced the Seawolves to travel to Albany for their championship tournament, where they lost by two points to ... Albany, which is a mere 85 spots lower than Stony Brook in the BPI rankings.

Meanwhile, the top line for Giants has rounded into shape. Gonzaga, Indiana, Duke and Louisville (in whatever order) are pretty clearly consensus picks as No. 1 seeds. While our spreadsheets (and other analytical models) would argue that Florida deserves a place among their ranks, that's a formidable group even without the Gators. We now estimate that potential 16-seeds are, on average, 1.5 points per possession worse than average; their "secret sauce," or tendency to play like past successful Killers, rates at an average of -0.8 points per 100 possessions; and there's just a 10.5 percent chance that any of them will beat one of this year's No. 1 seeds. These numbers are all much more in line with historical norms than the odds we reported last week.

But beyond the No. 1 seeds, there's a big drop-off in Giant quality this season. Our statistical model sees little difference in fundamental strength between teams that are likely No. 2 seeds and No. 5 seeds or between probable No. 6 seeds and No. 8 seeds. Moreover, among teams heading for the upper brackets, half a dozen lose three or more points per 100 possessions in our calculations because they play like squads who lost to deep underdogs in the past. Take a No. 6 seed that plays like an 8 and add some sour secret sauce and you get one curdled Giant.

So here's a look at how our spreadsheets rate the 10 most vulnerable Giants, including a look at the factor that measures their similarity to slain favorites of the past. (We analyzed all teams that Joe Lunardi currently projects as first through seventh seeds. Current No. 7 seeds might bump a slot or two higher by Selection Sunday, and actual No. 7s have played tournament games as Giants against No. 15s, as Florida did last year against Norfolk State.)

1. UCLA Bruins

Giant Rating: 55.5 (percentage likelihood of beating a generic Killer, on a scale from 0 to 100)
"Secret Sauce": -3.8 (points per 100 possessions)
Key Metric: Offensive rebounds on 30.2 percent of missed shots (ranking 224th in the NCAA)

MORE MEN'S COLLEGE BASKETBALL HEADLINES