The projections go far deeper at SI.com/CBBpreview, which has rankings of every team (from 1 to 351), forecasts of each conference race and video of the top scorers, freshmen, transfers and breakout players
This is an article from the Nov. 10, 2014 issue
PRESEASON POLLS have traditionally been an average of perceptions—of returning talent and key additions, and of a coaching staff's ability to convert all of that into wins. This season Sports Illustrated has adopted a rigorously empirical approach to its 1--351 rankings. Economist Dan Hanner has spent four years developing a statistical model that works from the bottom up, projecting advanced statistics—including offensive rating, possession usage, and rebound steal and block rates—for every Division I player, then using the results to forecast overall team performance. Hanner's model simulates the 2014--15 season 10,000 times, accounting for fluctuations in individual performance and health, and ranks teams according to their median efficiency.
This is the first season Hanner, working with SI's Luke Winn, has attempted to forecast (and publish) raw points-, rebounds- and assists-per-game totals for every player in the top 11 conferences. These projections are based on past advanced-statistical performance in the context of more than a decade of player data; the predictive power of recruiting rankings, both on immediate freshman performance and longer-term development; the quality of teammates and coaches; specific coaches' playing-time distribution tendencies; the estimated pace at which a team will play; and intelligence gleaned from coaches on how their rotations will be structured, which helped Hanner and Winn better understand how many minutes and shots will be available to each player. Read on to see how the top 16 teams stack up.
MEN'S SWEET 16
11 NORTH CAROLINA
13 OHIO STATE
14 SAN DIEGO STATE
15 IOWA STATE
16 WICHITA STATE