The Duke Blue Devils are not the favorite to win this year's college basketball national championship.
No, according to BPI, the favorite is Wichita State, who we dubbed as BPI's top team in July and now project to have an 11 percent chance to win the national title, more than any other team.
You can read more about why we like Wichita State here, but let's go back to Duke because BPI's skepticism is noteworthy. While the model believes that the Blue Devils are one of the best teams in the nation, it ranks them 14th. This is even though the model considers the track record of Mike Krzyzewski and has full knowledge of Duke's recruiting class, which boasts four five-star recruits and is headlined by Marvin Bagley III. While we're on the subject, the model isn't all that high on Kentucky (19), another team including a number of expected one-and-done players.
Keep in mind: BPI doesn't dislike one-and-done teams. It just dislikes these one-and-done teams.
In fact, over the last three seasons, either Duke or Kentucky has ranked No. 1 overall in preseason BPI. What's the difference this year? A lack of returning talent to complement the high-end recruits.
This year, Duke is taking one-and-done to an extreme by only returning 18 percent of its adjusted minutes (which considers transfers and previous-season injuries). A year ago, when Duke was BPI's preseason No. 1 team, that number was at 75 percent. And though the Blue Devils only returned 31 percent adjusted minutes at the start of the 2015-16 season, those returners were more efficient than their current crop of returnees.
Remember, Grayson Allen is the only returning Duke player who played more than 10 minutes per game last season. And he's coming off of a down year.
Even accounting for the skill of the modern elite freshman, they don't always pan out. Anyone doubting that need look no further than Kentucky's 2012-13 team (featuring freshmen Nerlens Noel, Willie Cauley-Stein, Archie Goodwin and Alex Poythress), which was ranked No. 3 in the AP poll heading into the season but failed to even qualify for the NCAA tournament.
All this is to say that Duke may turn out to be excellent, but we shouldn't anoint them yet. As we pointed out in our July writeup, the model is wary of the Blue Devils' abilities on the defensive side of the floor.
Wichita State (and Duke) aside, who else is a contender?
Look, it's early. There's a whole regular season to be played before a wild, 68-team single-elimination tournament occurs, so no team is going to have a particularly high chance to win the whole thing at this point. But here are the favorites to cut down the nets and win it all in March.
There are some of the usual suspects up there. But if Duke at 14 seems crazy, what does that make Cincinnati at 6?
Here's what going on:
The Bearcats were a solid team last year: a No. 6-seed in the tournament that was knocked by UCLA in the Round of 32.
Part of BPI's calculation considers the minute-weighted efficiency of the returning players. And it turns out that Cincinnati has the enviable combination of a high percentage of returning minutes and sneaky efficiency from those returning players. The Bearcats' returning players' offensive rating, adjusted for opponents and minute-weighted, ranked 13th best in the nation and their defense was 17th-best. That kind of production, from players like F Gary Clark and G/F Jacob Evans, on both ends of the court is a good sign for the Bearcats' 2017-18 prospects.
TCU is in a similar situation. For a school that hasn't made the NCAA tournament since 1998, BPI is awfully high on the Horned Frogs' prospects. TCU's returning players were just a bit less efficient than the Bearcats' but are returning even more adjusted minutes.
One thing to note in both situations is that the difference between BPI's fifth-ranked team, TCU, and its 12th-ranked team (West Virginia) is less than .6 points per game.
The Pitino (and Bowen) effect
The scandal that rocked the college basketball world, cost Rick Pitino his job and resulted in the suspension of Louisville's Brian Bowen had dramatic consequences on the Cardinals' BPI ranking.
In our initial run of BPI back in July, Louisville was ranked second in the nation behind only the Shockers. Now, post-Pitino firing and Bowen suspension, the Cardinals have dropped all the way to 31st.
The model projects a Louisville record of 19.3-11.7 and gives the Cardinals a 67 percent chance to reach the tournament. But it also gives them less than a 1 percent chance to win the national title, a number that surely would have been much higher if the status quo had remained.
Kansas leads strong Big 12
It's no surprise to see Kansas as BPI's best team in the Big 12, but it's a little surprising to see how close their competitors are behind them.
Kansas ranks third in the nation in preseason BPI but is trailed by TCU (5), West Virginia (12), Baylor (17), Iowa State (20) and Texas Tech (22). Despite the strength at the top of the Big 12, the Jayhawks still are solid favorites to win the conference and are more likely than not to win or share the regular-season title for the 14th consecutive season.
But make no mistake: the Big 12 is strong this year. BPI projects an average of over six teams from the conference to make the tournament, which is particularly impressive given that it has just 10 teams.
ACC up for grabs
Partially because of BPI being down on Duke, the model also sees close competition in the ACC. And, as is the case nationally, BPI also doesn't have Duke as the favorite to win the ACC's regular season. That honor goes to Virginia, which BPI believes is the best defensive team in the nation. The Cavaliers hold a small edge over Notre Dame, North Carolina and Duke.
Paul Sabin contributed to this report.
For more from ESPN Analytics, visit the ESPN Analytics Index.