Bart Torvik has his
2021 T-Rank projections out and they are... interesting.
1. Gonzaga
2. Virginia
3. Baylor
4. Creighton
5. Texas
6. Wisconsin
7. Kentucky
8. West Virginia
9. Florida
10. Kansas
11. Texas Tech
12. Ohio State
13. Villanova
14. Houston
15. Michigan State
16. Purdue
17. San Diego State
18. Duke
19. Oregon
20. Florida State
21. Indiana
22. North Carolina
23. Penn State
24. UCONN
25. Iowa
In general, I do not see too many outliers here from other Top 25 lists. Duke is obviously lower than I expected. More on that later. Texas is basically returning everyone, but 5th seems absurdly high to me. Florida also seems high to me and Villanova is a little lower than everyone else has them. T-Rank doesn't have Saddiq Bey returning, so 13th might be a good place for Nova based on the roster he is projecting.
T-Rank also includes his projected Adjusted offensive and defensive efficiencies. He has Duke with and ADJOE of 111.2, which is 39th in the nation. That seems way off to me. Duke has never finished a season with an ADJOE below 10th in the history of T-Rank's database, which goes back to 2007-08. The worst Duke has been is last season, when the ADJOE ended at 115.3, 9th best in the country. I just don't see a Duke team coached by Coach K being unable to score. Duke is projected to have the 4th best ADJDE, for what it is worth, with a 91.5 rating. The projected record is 22-9 on the season with a 13-7 conference record. That seems conservative to me.
T-Rank also has player projections. Matthew Hurt is the top player with a projection of 16.6 pts/7.4 rbg/1.6 ast stat line playing 82% of the minutes. The core rotation consists of Hurt, Moore, Goldwire, Jalen Johnson, Jeremy Roach, JD Steward, Mark Williams, Jaemyn Brakefield, and Henry Coleman, in descending order. Patrick Tape is projected to play 7% of available minutes. I think you could flip Tape and Coleman.
Looking at the player projections, I think T-Rank is overly conservative on projecting Jalen Johnson , who he has getting 13 pts/6.3 rpg/1.6 ast. I think Johnson might be closer to 15/8/4, but T-Rank has a tendency to be conservative in player projections in general. I also don't think Jordan Goldwire is going to play as much as he is projecting. We'll see, but I assume Jeremy Roach is going to start at PG with a frontcourt of Wendell Moore, Jalen Johnson, and Matthew Hurt. At SG, I expect to see a good amount of DJ Steward. Goldwire can and should play the PG position when Roach sits, but I think he's closer to a 20-25 MPG player than a 30 MPG player as T-Rank is projecting.
The other thing about this group is that it does not include Cassius Stanley. If Stanley does return, I wonder if that bumps the team's rating up considerably. Goldwire, Baker, Stanley, Hurt, and Moore would represent some of the most returning minutes a Duke team has had in quite a long time.