INSIDE THE RANKINGS

EXPLAINING SI'S METHODOLOGY—AND WHAT THE DATA SHOWS

COLLEGE BASKETBALL values drama over predictability. This is the trade-off that defines the sport, that captivates the nation for three weeks in spring, when the champion is decided by a 68-team single-elimination bracket, a format that does little to ensure that the best team wins.

SPORTS ILLUSTRATED's College Basketball Projection System is, in a way, the anti--NCAA tournament. Our statistical model simulates a given season 10,000 times in order to find the most frequent No. 1—this year, that would be North Carolina—and determine where the other 350 Division I teams fall in line. This marks the second year that SI's preseason rankings have been decided by the projection system, a collaboration among economist Dan Hanner, SI producer and writer Chris Johnson and me. The results from year one were promising: The system forecast—in exact order—the eventual top four teams in adjusted efficiency (Kentucky, Arizona, Wisconsin, Duke). It also predicted that title-game opponents Wisconsin and Duke would have the most efficient offenses, and it ranked all eight of the NCAA tournament No. 1 and No. 2 seeds among its top 10.

For offense, the system projects every player's efficiency and shot volume by incorporating his past performance, recruiting rankings, development curves for similar Division I players, the quality of his teammates and his coach's ability to develop and maximize talent. Those stats are weighted based on the team's rotation—including human intel on who's expected to play—then used to produce each team's offensive efficiency projection. (The 10,000 simulations account for significant variance in individual performances as well as injury scenarios.) Team defensive efficiency projections are based on a blend of individual stats (rebound, steal and block percentages), roster turnover (if churn is low, then 2014--15 performances in areas such as two-point field goal percentage are given a lot of weight; if high, then a coach's historical defensive résumé matters more), experience (veterans have fewer lapses) and height (taller frontcourts make for stingier D).

This is what the projections tell us about 2015--16.

1 THIS IS NO YEAR FOR JUGGERNAUTS

The Tar Heels are No. 1, but the system hardly indicates they're a threat to run the table: This year's UNC team would have ranked eighth in SI's 2014--15 preseason projections. Along with Kansas and Kentucky, North Carolina makes up a cluster of good-but-not-great teams at the top. But while the high end of the rankings looks softer than it did last season, the back end looks stronger: SI's No. 25 team, Miami, would have ranked 20th in '14--15, and the No. 50 team, Tulsa, would have been 44th. The 2016 NCAA tournament should yield weaker No. 1 seeds than last season's but stronger six to 11 seeds.

2 DEPTH OF TALENT IS WHAT SEPARATES NORTH CAROLINA AND KANSAS

Often overlooked in Duke's title run last spring was the absence of a single meaningful second of playing time by someone who wasn't formerly a top 100 recruit (according to the Recruiting Services Consensus Index). North Carolina is set to follow suit this season: All 10 players in its rotation were top 100 recruits, providing injury insulation at every position. Kansas, likewise, could give minutes to 10 former top 100 prospects and feasibly endure an injury at any spot while still contending for the Big 12 title. Wisconsin, on the other hand, is projected noticeably lower (No. 34) by SI than in traditional polls, and that's based on a dearth of quality depth. The Badgers have two All-America candidates at the top of their rotation—forward Nigel Hayes and point guard Bronson Koenig—but no strong alternatives behind either. They could end up with as many as five freshmen in their rotation, only one of whom was a top 100 recruit. Wisconsin has a history of finding and developing underrated talent, but it's historically rare for a non-top-100 recruit to make a significant impact as a freshman on a major-conference team.

3 THE BIG-OFFENSE, ADEQUATE-DEFENSE MODEL COULD KEEP WORKING

Duke entered last season's NCAA tournament No. 3 in kenpom.com's adjusted offensive efficiency ranking but 57th in the adjusted defensive efficiency rankings—and we all know how that worked out. The Tar Heels ride a similar model to No. 1 this preseason; we project that they'll have the nation's No. 2 offense and just its 45th-best defense. Indiana, SI's No. 14 team, offers an even better test case, though, for big offense and marginal defense. Behind the talented backcourt scoring duo of Yogi Ferrell and James Blackmon Jr., the Hoosiers are projected to have the No. 1 offense but just the 113th D. If they can find a way to crack the defensive top 50—if, say, five-star freshman center Thomas Bryant emerges as a strong rim protector, or if they can ratchet up pressure after ranking 330th nationally in turnover percentage last year—then they're a dark-horse title contender.

4 MARYLAND WILL HAVE TO DEFY SOME HISTORY TO WIN IT ALL

SI's projections fancy the Terrapins, putting them fifth overall, but that's lower than nearly every poll out there. What could hold Maryland back? Coaching plays a significant role in our calculations, and while Mark Turgeon's teams have made the NCAA tournament six times over 15 seasons at Wichita State, Texas A&M and Maryland, his offense has never ranked in the top 25 in adjusted efficiency, and only three times has he led a top 25 defense. A fair follow-up question: Has Turgeon ever had this much talent at his disposal? Probably not. But he's had future NBA players to work with in Donald Sloan, DeAndre Jordan, Khris Middleton and Alex Len—not to mention one season already in College Park with two pros-to-be, forward Jake Layman and point guard Melo Trimble—and he has yet to turn that individual flair into a highly efficient unit. Conversely, there are coaches who keep producing elite teams regardless of personnel. A prime example: Bill Self, who's had top 10 defenses for eight of the past 10 seasons at Kansas. Thus, we peg the Jayhawks to have a top 10 D once again, even after losing two key rotation players to the NBA.

5 THE TERPS WOULD BREAK THE MOLD FOR TITLE-TEAM CONSTRUCTION

Among the past 10 national champs there have been four basic models of team construction. (Veteran denotes a sophomore or above.)

• One-and-done-freshmen-led, veteran-augmented: Duke 2015, Kentucky '12

• Veteran-led, freshmen-augmented: UConn '11, North Carolina '09

• Veteran-led, transfer-augmented: UConn '14, Louisville '13

• Purely veteran-led: Duke '10, Kansas '08, Florida '07, Florida '06

This year's Terps? They're none of the above. They look more like a sampler platter from every talent pool. They have vets in Layman, Trimble, swingman Jared Nickens and guard Dion Wiley. They have a potential one-and-done center in Diamond Stone. They have two high-impact D-I transfers in guard Rasheed Sulaimon (from Duke) and forward Robert Carter Jr. (Georgia Tech), both of whom we project to start. They have a top 100 juco transfer in point guard Jaylen Brantley, who will back up Trimble. They even have two internationals who could contribute frontcourt relief minutes in Ivan Bender (Bosnia) and Michal Čekovský (Slovakia). Aside from finding a D-II or D-III transfer, Maryland seems to have worked every angle.

6 GET READY FOR THE MID-MAJORS TO MAKE SOME NOISE

Rare is the season in which we can claim that the nation's most efficient backcourt and most efficient frontcourt both come from mid-majors. No. 7 Wichita State has an unmatched 1--2 combo in senior point guard Fred VanVleet (projected to have an All-America level 121.1 offensive rating while using 25% of the Shockers' possessions when he's on the floor) and senior shooting guard Ron Baker (120.1 on 22%), who also defend their positions better than any other backcourt duo. No. 9 Gonzaga is reconfiguring its lineup so that 6'10" senior Kyle Wiltjer, whom SI projects as the nation's best high-usage scorer (127.9 offensive rating on 26%), can play small forward and share the frontcourt with two high-efficiency imports: Domantas Sabonis (from Lithuania) and Przemek Karnowski (Poland). All three are polished scorers with NBA potential.

Valparaiso (at No. 40) has a nation-high 98.5% of its minutes back from last season, when the Crusaders made the NCAA tournament as a No. 13 seed. That makes them the strongest Horizon League team since Butler in 2011. The gap between the blue bloods atop SI's rankings and Wichita State and Gonzaga is not large at all. And Valparaiso, by being mentioned in the same sentence as Brad Stevens's last great Butler team, has at least a shot at a title, right? We're not quite ready to declare this the season that a mid-major wins it all. What we are saying is that there's a reasonable statistical possibility.

NOVEMBER MADNESS

ROUND OF 64

ROUND OF 32

SWEET 16

ELITE EIGHT

FINAL FOUR

TITLE GAME

EAST

NCAA CHAMPS

1NORTH CAROLINA

16BU/MONTANA

8LSU

9BAYLOR

5GEORGETOWN

12CENTRAL MICHIGAN

4UTAH

13LOUISIANA-LAFAYETTE

6PURDUE

11SYRACUSE

3OKLAHOMA

14STONY BROOK

7MIAMI

10VALPARAISO

2MARYLAND

15HOFSTRA

ROUND OF 64

ROUND OF 32

SWEET 16

ELITE EIGHT

FINAL FOUR

TITLE GAME

WEST

1DUKE

16N.C. CENTRAL

8VANDERBILT

9OREGON

5TEXAS

12STEPHEN F. AUSTIN

4CALIFORNIA

13UC-IRVINE

6SAN DIEGO STATE

11PITTSBURGH

3GONZAGA

14BELMONT

7MICHIGAN

10UCLA

2VILLANOVA

15NEW MEXICO STATE

TITLE GAME

FINAL FOUR

ELITE EIGHT

SWEET 16

ROUND OF 32

ROUND OF 64

MIDWEST

1KANSAS

16MOUNT ST. MARY'S/TEXAS SOUTHERN

8CINCINNATI

9WISCONSIN

5XAVIER

12BOISE STATE/DAVIDSON

4INDIANA

13IONA

6LOUISVILLE

11RHODE ISLAND

3ARIZONA

14SOUTH DAKOTA STATE

7TEXAS A&M

10N.C. STATE

2WICHITA STATE

15WOFFORD

TITLE GAME

FINAL FOUR

ELITE EIGHT

SWEET 16

ROUND OF 32

ROUND OF 64

SOUTH

1KENTUCKY

16N.J. INST. OF TECH

8WEST VIRGINIA

9BUTLER

5UCONN

12MARQUETTE/VCU

4MICHIGAN STATE

13UAB

6NOTRE DAME

11FLORIDA

3IOWA STATE

14HIGH POINT

7OHIO STATE

10FLORIDA STATE

2VIRGINIA

15PRINCETON

Top 20

1 North Carolina

2 Kansas

3 Kentucky

4 Duke

5 Maryland

6 Virginia

7 Wichita State

8 Villanova

9 Gonzaga

10 Iowa State

11 Arizona

12 Oklahoma

13 California

14 Indiana

15 Michigan State

16 Utah

17 SMU

18 Georgetown

19 Connecticut

20 Texas

SI.COM

For the complete ranking of all 351 D-I teams, conference breakdowns and expert analysis, go to SI.com/CBBpreview

PHOTO

SIMON BRUTY FOR SPORTS ILLUSTRATED

PHOTO

CHUCK BURTON/AP (MEEKS)

Kennedy Meeks

PHOTO

GREG NELSON FOR SPORTS ILLUSTRATED (SELDEN)

Wayne Selden Jr.

PHOTO

TIM CLAYTON FOR SPORTS ILLUSTRATED (JEFFERSON)

Amile Jefferson

PHOTO

GREG NELSON FOR SPORTS ILLUSTRATED (ULIS)

Tyler Ulis