Publish date:

INSIDE THE RANKINGS

EXPLAINING SI'S METHODOLOGY—AND WHAT THE DATA SHOW
Author:

WE FORECAST WHAT we can, and happily cede the outcome to the players and the drama they create on the court. SI's College Basketball Projection System, now in its third year, simulates the season 10,000 times and ranks teams from 1 to 351 according to their average efficiency. Last year North Carolina was a weak No. 1 in a historically wide-open field—and five months later the Tar Heels led Villanova with 15 minutes left in the national title game. What transpired after that, culminating with Kris Jenkins's buzzer-beating three (POINT AFTER), was the greatest finish in the history of the sport. You don't get Jenkins's dagger without the exhilarating chaos of the NCAA tournament, and you don't get the country's most accurate preseason rankings—as ours have been for the past two years—without running 10,000 simulations.

This project starts from the ground up, assessing every player. For offense, the system projects efficiency and shot volume by considering past performance; recruiting rankings and advanced AAU stats; development curves for similar Division I players over the past 14 seasons; the quality of a player's teammates; and his coach's ability to maximize talent. Those stats are weighted based on the team's rotation—including scouting intel on who's expected to play—then used to produce each team's offensive efficiency projection. (The simulations account for variance in individual performances as well as injury scenarios.) Team defensive efficiency projections are based on players' projected rebound, steal and block percentages; height (taller frontcourts make for stingier D); experience (veterans have fewer lapses); and coaches' defensive résumés.

The complete rankings and player forecasts—and much more—can be found in the following pages and on SI.com, but here's a deeper look at what our projections indicate for 2016--17.

1 BREAK UP THE BLUE DEVILS

While SI's system predicted that no team would be dominant last year, Duke has the profile of a juggernaut in 2016--17, ranking first in offensive efficiency by a wide margin and 11th in defense. But as coach Mike Krzyzewski recently said of being ranked atop preseason polls, "It's like Internet dating. Until you see the real person, [you] can create a big, nice picture of who we're supposed to be." To bring that picture into focus, we looked at every No. 1 team from the past decade—using preseason rosters to project performance—to see where the Blue Devils stand among the greats.

Starting with 2007--08, only one team had a stronger projection than the current version of Duke: '08--09 North Carolina, which came into that season with a national player of the year candidate in power forward Tyler Hansbrough; a supporting cast of efficient veterans in guards Ty Lawson, Wayne Ellington and forward Danny Green; and two elite freshmen in Ed Davis and Tyler Zeller.

If that roster blueprint seems familiar, it's because it closely resembles what the Blue Devils have now. Junior combo guard Grayson Allen is SI's pick for national player of the year, and he's just the kind of high-volume, high-efficiency scorer that makes an elite offense possible. He has quality vets around him in guards Matt Jones and Luke Kennard and power forward Amile Jefferson. And Duke added two freshmen either of whom could be the No. 1 pick in the 2017 NBA draft (small forward Jayson Tatum and power forward Harry Giles) and two more who could be first-rounders (center Marques Bolden and combo guard Frank Jackson). Coach K can give every meaningful minute to players who were ranked in the top 35 of their respective recruiting classes, and Duke has the depth to withstand injuries. All this adds up to an otherworldly offensive projection; it just needs to work in the real world the same way it did for the Tar Heels in 2009, when they cruised to a national title. If issues arise—and Duke's lack of a pure, pass-first point guard could be one—then Kansas, a team with two veteran floor generals, a crop of talented underclassmen and a strong, No. 2 overall projection, will be right there waiting.

2 THE SOPHOMORE LEAP

The player data we've gathered from the past 14 seasons give us a unique window into how much production to expect from highly ranked recruits. We consider a valuable starter to be someone who plays at least 50% of his team's minutes with an offensive efficiency rating—adjusted for usage rate and defensive strength of schedule—of at least 110, or 1.1 points per possession. Freshmen ranked in the top 10 of the Recruiting Services Consensus Index have hit that valuable-starter mark 76% of the time. That bodes well for the two teams with multiple top 10 freshmen this season—Duke (Tatum and Giles) and Kentucky (shooting guard Malik Monk, point guard De'Aaron Fox and power forward Edrice "Bam" Adebayo).

Outside of the top 10, though, the majority of freshmen don't make a significant impact. Just 40% of those ranked in the 11-to-20 range were valuable starters in their inaugural seasons, and that number dropped to 24% for recruits in the 21-to-50 range and 13% for those in the 51-to-100 range. Most of those second-tier recruits don't break out until year two. Our data show that players make the biggest improvement in offensive rating as sophomores (a median increase of approximately five points, compared with just two to three points for juniors and seniors), and that the sophomore leap is even more pronounced for ranked recruits.

Which schools will benefit most from this boost in production? The five teams that gave the most minutes to top 100 freshmen who are returning are Texas A&M (30.0%), Texas (25.7%), San Diego State (23.6%), Oregon State (22.5%) and Louisville (22.3%). The Cardinals are an especially promising case: Even though they lost Damion Lee and Trey Lewis, their top two scorers from last season, we project them to rank 13th and finish fourth in a loaded ACC. That's because they have two former top 40 recruits in sophomore wing Deng Adel and combo guard Donovan Mitchell, both of whom flashed breakout-star potential at the Adidas Nations camp in Garden Grove, Calif., this summer. Adel is forecasted to make a surge in efficiency by slashing his high turnover rate (29.0%), and Mitchell looks ready to emerge as the Cardinals' leading scorer and one of the country's better lead guards. Although he mostly came off the bench during ACC play as a freshman, Mitchell used a team-high 23.0% of possessions with decent efficiency, suggesting he has the DNA of a go-to guy. Those sophomore leaps, along with coach Rick Pitino's extensive track record of creating elite defenses, make our system bullish on the Cardinals.

3 THE 90% CLUB

In 2009--10, when Butler made the first of its back-to-back runs to the national title game, the Bulldogs had more than just wunderkind coach Brad Stevens and future NBA draft picks Gordon Hayward and Shelvin Mack on their side. They also brought back every rotation player from a team that had finished 38th in efficiency in the previous season. These players made substantial improvements and were a half-court heave from upsetting Duke in the title game.

Three teams in our projected top 50 are bringing back at least 90% of their minutes played from last season: Wisconsin (99.7%), Princeton (99.0%) and St. Mary's (97.9%). While schools in this situation run the risk of plateauing—it can be hard for a new player to break out when every role is already set—our system predicts that the Badgers can improve enough to be in the mix for their third Final Four in four years. After finishing 38th in efficiency last season (the same position Butler occupied in '08--09), Wisconsin projects at No. 8, with the No. 1 D.

The Badgers' defense was already elite last season; their offense, which we forecast to jump from 90th in efficiency to 27th, is their prime area for growth. Wisconsin gave 28.9% of its minutes to freshmen last season, many of whom committed turnovers at an uncharacteristically high rate for the program, which has a history of sure-handedness. Power forward Ethan Happ (a TO rate of 19.0% last season), wing Khalil Iverson (28.7%) and power forward Alex Illikainen (15.9%) are expected to become less mistake-prone with experience. The Badgers should benefit from coaching continuity too; last year they endured the midseason retirement of coach Bo Ryan, and after longtime assistant Greg Gard took over, he switched their offense to the classic (motion) version of the swing to better integrate the young players. Gard is keeping the swing, and he's tried to prevent complacency by mixing up players' roles in the off-season. "I've had the teams jumbled all summer and fall, and I think it's helped everybody," Gard says. "They've had to fight and scrap."

4 CAN VILLANOVA HOLD ON TO THE NCAA TITLE?

It will be next to impossible for the Wildcats to top last season's drama and highly unlikely for them to beat last season's level of efficiency, but they are bringing back 72.2% of their minutes—the most of any defending champ in the past decade—and look like one of the most viable title contenders, after Duke, Kansas and Kentucky.

Our projection system does not, however, consider Villanova to be the strongest defending champ of the past decade: That would be 2010--11 Duke, which came into that season having added future NBA All-Star point guard Kyrie Irving and transfer Seth Curry to a team that already had five future NBA players in Nolan Smith, Kyle Singler, Mason and Miles Plumlee, and Ryan Kelly. Losing Irving to a foot injury weakened those Blue Devils, and they eventually lost in the Sweet 16. The second-strongest defending champ was 2013--14 Louisville, a team that finished No. 1 in efficiency that season but was knocked out in the Sweet 16 on a dagger by Kentucky's Aaron Harrison.

Villanova will be led by senior Josh Hart, who we project will challenge Duke's Allen for player of the year. Sophomore point guard Jalen Brunson and small forward Eric Paschall, a sophomore transfer from Fordham, are both expected to make substantial offensive gains. 'Nova projects to have the No. 3 overall offense and be a No. 2 seed in the NCAAs—just as it was in 2016, when what mattered was not whether the Cats were the best team all season, but whether they were the best team over the final three weeks.

MEN'S SWEET 16

1. Duke

2. Kansas

3. Kentucky

4. Oregon

5. Villanova

6. North Carolina

7. Virginia

8. Wisconsin

9. Gonzaga

10. Arizona

11. Purdue

12. Xavier

13. Louisville

14. Indiana

15. Syracuse

16. UCLA

GO FIGURE

23.6

POINTS THAT SI PROJECTS DAVIDSON SENIOR POINT GUARD JACK GIBBS WILL AVERAGE—THE MOST IN THE NATION.

18.2

PROJECTED SCORING AVERAGE FOR WASHINGTON POINT GUARD MARKELLE FULTZ, TOPS AMONG FRESHMEN AND 10TH IN DIVISION I.

6.3

ASSISTS PER GAME FOR IOWA STATE SENIOR POINT GUARD MONTE MORRIS, WHO WILL ALSO AVERAGE 15.5 POINTS TO BECOME A DARK-HORSE PLAYER OF THE YEAR CANDIDATE.