It's time for a detailed look at how our Giant Killers statistical model sees the NCAA tournament brackets. We'll go region by region, and, as we did last year, we will sort potential upsets into four categories: Best Bets, Worth a Long Look, Not Completely Crazy and Stay Away. Hopefully, those names are self-explanatory; Best Bets have a good chance to win outright and Stay Aways are likely to lose no matter what. But we can't tell you exactly how to fill out your brackets in between, because that depends on how heavily your particular pool rewards upsets. (The more points you score when underdogs win, the more you should be willing to pick high-risk teams.)
We began with the East Region, and now we're headed West. In terms of upset potential, a trip West has been good for the Giant Killers model.
Note: Numbers in parentheses represent a team's Giant Killer score. The higher the number for underdogs, the more likely they can pull the upset. The higher the number for favorites, the more likely they are to fall. For a deeper explanation, click here.)
WEST REGION
Worth a Long Look
No. 12 Memphis (24.0) vs. No. 5 Arizona (57.4)
UPSET CHANCE: 27.9 percent
America, meet the second-most vulnerable Giant in the field.
Yes, Arizona boasts exceptional forward Derrick Williams, but his brilliance alone can't hide several fundamental flaws that put the Wildcats in the company of past slain Giants. Start with this subtle deficiency in what appears to be a top-notch offense (effective field goal percentage of 54.4 percent, eighth in the NCAA): Arizona generates only 45.7 percent of its points on 2-pointers, 312th in the country, and does a terrible job of defending teams inside, allowing teams to shoot a staggering 50.5 percent on 2-point FGs. They're getting outscored by 7.07 points per 100 possessions on 2-pointers, by far the worst differential of any Giant. That might not matter much against other big-time teams, but past numbers indicate it leaves them vulnerable to a little guy on the wrong day; it's the steady-Eddie teams that dominate inside the arc that are traditionally upset-proof. And because the Wildcats are just average offensive rebounders and well below average at forcing turnovers they can't make up for it by generating extra possessions.
Arizona's flaws are especially problematic against Memphis, because the way to beat the C-USA champs is -- you guessed it -- by hitting the offensive glass (the Tigers are 266th in the country in defensive rebounding) and forcing turnovers (they give it up on 21.9 percent of possessions). Meanwhile, Memphis just happens to be fairly efficient from inside, and its defense is strong inside and out, giving up just 95.3 points per 100 possessions (60th in the NCAA), limiting the areas where Arizona can make hay.
The Tigers aren't a great team by any stretch of the imagination, but their strengths as well as their weaknesses match up well against a Giant with Arizona's vulnerabilities. Memphis starts four guards and gives significant minutes to five, and -- very unusually for a Killer-Giant matchup -- this contest looks like it will turn on how well the Tigers' little men can penetrate against the Cats. It's a live one if you're looking to pick a 5/12 upset.
Not Completely Crazy
No. 13 Oakland (21.1) vs. No. 4 Texas (17.3)
UPSET CHANCE: 11.7 percent
After four days of commenters and e-mailers demanding that we cover Oakland, here goes! In asking what makes a team special, it might seem absurdly simple to answer, "Making baskets," but that's what the Golden Grizzlies do best. They're constantly in motion (72.9 possessions per game, seventh most in the NCAA). They nab offensive rebounds (36.3 percent of misses, ranking 39th). They get a ton of blocks (12.2 percent of opponent 2-point shots). And then they put all those balls they collect through the hoop, shooting an amazing 55.8 percent inside (second in the country only to Kansas) and 37.7 percent on 3s (39th in the NCAA). In the Summit League tournament, the Grizzlies dropped 45 field goals -- 45! -- on South Dakota State, then laid 35 on Oral Roberts one night later.
So there's no question Oakland is an offensive machine. Its center, 6-foot-11 Keith Benson, is a dominator who will play in the NBA next year, and the Grizzlies have three players besides Benson who have effective field goal percentages higher than 60 percent. It all adds up to 115.9 points per 100 possessions, the 13th-best scoring efficiency in the country. All season long, if Oakland happened to fall behind in a game, all you had to do was wait a couple of minutes, and it'd go on a big run -- in the Summit final against Oral Roberts, it was almost like the Grizzlies were toying with the Eagles. Oakland also scheduled the fourth-toughest slate of nonconference opponents in the NCAA this season. And to top it all off, coach Greg Kampe, who has been at Oakland for 27 years, is not only quotable, he once made a list of the 50 sexiest men in college basketball despite a wardrobe heavy on windbreakers and a mop of curly hair that many fans wrongly assume is a bad toupee.
So, hey, we get it, Oakland is a lot of fun. And we've enjoyed its ride. Our model, however, is a killjoy. It notes that the Grizzlies force turnovers on just 16.6 percent of opponent possessions (327th in the NCAA), something that's easy to overlook when a team plays at Oakland's pace. Oakland also allows opponents to shoot 47.5 percent on 2-point FGs, a dangerous sign at this time of year, when underdogs are typically going up against Giants with big inside games. And their overall defense just ain't that hot (103.4 points per 100 opponent possessions, 200th in the country). Still, the Grizzlies would have been a fine speculative pick had they drawn most Giants as a first-round opponent, but they got Texas. The Longhorns are exactly the kind of team most analysts won't like as a tournament pick: they're young, and they have fared poorly in March over the past couple of seasons. But by advanced metrics, Texas is one of the top five teams in the country, and the Longhorns are certainly a tougher opponent than any other 13-seed, or in fact any 14-seed, has to face. Moreover, their top strength is stopping Oakland's strength: Texas allows opponents to shoot just 41.5 percent on 2-point FGs and only 29 percent on 3s. Overall, the Longhorns are giving up just 86.5 points per 100 possessions, tops in the entire country.
Our hearts are rooting for Oakland. Our model sees a buzz saw straight ahead.
Stay Away
No. 14 Bucknell (less than 2.0) vs. No. 3 Connecticut (23.1)
UPSET CHANCE: 9.3 percent
You know how we keep writing that Giant Killers need to maximize the number of their possessions? Bucknell ranks 282nd in the country at generating turnovers (TOs on just 18.2 percent of opponent possessions), and is even worse, 319th in the NCAA, at grabbing offensive rebounds (ORs on only 26.5 percent of misses). If our model were a fat lady, it, or she, would be singing. The fact the Bison rank even this high is due to a couple of danger signals beeping around the Huskies. Our model doesn't know whether UConn is juiced or gassed from its stunning Big East tournament run, but it sees that Connecticut doesn't generate many turnovers or defensive rebounds. So an opponent who makes something of its possessions could be dangerous for the Huskies. While Bucknell shoots better than 40 percent from downtown, however, the Bison don't take enough 3-point shots to scare Connecticut. They don't take enough of any kind of shots to scare any Giant.
No. 15 Northern Colorado (11.6) vs. No. 2 San Diego State (4.3)
UPSET CHANCE: 5.6 percent
If our model could reproduce, it would take parental pride in watching San Diego State grow from a small-conference squad snubbed by the NCAA in 2009 into a Killer we liked last year (when the Aztecs were edged by Tennessee in the first round) and now into a full-fledged Giant. With its big frontcourt, San Diego State's rebounding prowess (ORs on 37.5 percent of misses, 18th-best in the NCAA) may not be a shock. But the Aztecs also protect the ball (turnovers on just 17.4 percent of possessions) and defend the perimeter, allowing opponent to shoot just 30.5 percent on 3s. And the Bears aren't exactly cut out to be conquistadors: they turn the ball over on 21.8 percent of possessions (264th in the country) and are heavily reliant on free throws (25.5 percent of all points, 12th in the country), a bright red flag at this time of year. All we ask is that San Diego State remembers which statistical models were nice to it on its way up.
No. 16 Hampton (7.0) vs. No. 1 Duke (31.8)
UPSET CHANCE: 4.0 percent
Year after year, our model has viewed Duke as an incredibly safe Giant. This season, the Blue Devils' statistical profile makes them vulnerable to the right kind of GK, but that team isn't Hampton. Duke isn't forcing turnovers at its usual high rate, and is a poor defensive rebounding club (allowing opponents to grab rebounds on 32.8 percent of misses, 202nd in the NCAA). Key result: Opponents are shooting poorly from inside (just 42.6 percent) against the Blue Devils, but scoring the bulk of their points there anyway (60.4 percent of points from 2-point FGs, second most in the nation). So Duke is outscoring opponents by a whopping 14.56 points per 100 possessions on 3s, but by just 1.40 points per 100 possessions on 2-pointers, which is terrific for regular-season success but not so great for producing the kind of steadiness that protects Giants from Killers in the early rounds of the NCAA tournament. Well, from many Killers, anyway. Hampton is 286th in the nation in 2-point field goal percentage (45 percent) and a lousy offensive rebounding team (29.3 OR percentage, 276th in the country). These Pirates' swords aren't nearly sharp enough to force Duke to walk the plank. Arrrrr!