Before I share the projections, I want to comment on a few places where my rankings disagree with some other experts. I explain the statistical reasons why my model is more skeptical of Texas, SMU, San Diego St., and Oklahoma. But I also do not completely agree with what the model is currently suggesting. The lineup-based statistical projection seems to be falling in love with teams with a lot of unproven high school talent.

Where My Model Disagrees with Other Experts

Texas (Too High Elsewhere): Texas returns 100% of their minutes from last year, they have super-recruit Myles Turner joining the front-court, and they made the round of 32 last year. But I think it is important to emphasize that Texas finished with the 6th best margin-of-victory in the Big 12, and the 39^th best margin-of-victory nationally. In terms of NCAA at-large teams without major injuries, Texas was the luckiest team in the nation last year, winning a ton of close games. Heck, even their NCAA win came by the slimmest of margins on a last-second buzzer beater. If you start from the premise that Texas was a Top 25 team, you could justify a Top 5 ranking. But based on how they really played, and how inconsistently their lineup performed individually, I have them just behind the other elite teams in 9th.

SMU (Too High Elsewhere): A lot of SMU’s players had huge jumps in efficiency last year.  Marcus Kennedy had an ORtg of 88 at Villanova and an ORtg of 106 at SMU last season. Nic Moore had an ORtg of 106 at Illinois St. and an ORtg of 118 at SMU last season. But big leaps in efficiency are usually followed by players slipping back some. AAC coaches are studying film of Kennedy and Moore’s game this off-season and figuring out ways to slow them down. The addition of Emmanuel Mudiay is why I have SMU improving its margin of victory from 30th to 16th, but without Myles Turner, I think it is premature to put them in the Top 10.

San Diego St.  (Too High Elsewhere): Most people account for Xavier Thames impact on the offense, but not enough people are recognizing how special it was to have an elite defensive rebounder like Josh Davis. Remember what happened to Oregon’s defense after Arsalan Kazemi graduated? Oregon fell from 10^th to 88^th on defense.

Oregon (Too High Elsewhere): Speaking of Oregon, I love the backcourt, but for a team that was horrible on defense last year, the lack of clear defenders in the paint is a red flag. Perhaps the highly regarded JUCO recruits will make a difference, but the front-court weakness is what is keeping Oregon out of my Top 25.

Oklahoma  (Too High Elsewhere): Oklahoma finished 2^nd in the Big 12 last year, but the Sooners only had the  5^th best margin-of-victory in the conference, and nationally their margin-of-victory was only 33^rd. That said, Oklahoma’s potential starting lineup is impressive:

-Ryan Spangler, a dominant rebounder, who rarely touched the ball, but had an ORtg of 125

-Dante Buford, the nation’s #72 recruit and a needed piece at forward

-Jordan Woodard, who had a 28% assist rate as a freshmen, and an ORtg of 108

-Isaiah Cousins, a 40% three point shooter with an ORtg of 112

-And Buddy Hield, a high volume shooter who made 90/233 or 39% of his threes last season

On paper, that should lead to a very good offense. But my model is skeptical that it will be better than last year’s offense. Oklahoma loses a player that took 30% of the shots last year, and was extremely efficient, in Cam Clark. The most likely scenario next season is that Oklahoma will replace Clark with some of its less skilled big men. Either veteran DJ Bennett, or one of the three freshmen bigs (of which Buford is the most highly ranked) will take Clark’s minutes. The increased size will improve the defense, but it will also hurt Oklahoma’s offense. While Buford is a Top 100 recruit, based on where he is ranked, we can’t expect him to be a consistent offensive player immediately.

What could overcome the loss of Clark is if Hield, Cousins, or Woodard took a significant step forward. But what you have to remember is that none of these guys was a Top 100 recruit out of high school. Realistically, they are all playing pretty close to their ceiling. Woodard has the biggest chance to improve, as the sophomore leap is typically the biggest, but he didn’t make a ton of mistakes as a freshman, so he doesn’t have as much room to grow.

As I’ve shown on many occasions, to win at the highest levels, having elite high school talent is important. It isn’t completely necessary, but the stats show that on average, a high school recruiting rank is an important predictor of career development. And Oklahoma has only one former Top 100 recruit on its roster.

That emphasis on the potential of Top 100 recruits, particularly former Top 30 recruits, is why my model likes the next three teams that many experts are skeptical about. Essentially, if you have 7-8 players with solid recruiting backgrounds, and a coach that has been highly successful in recent seasons, my model tends to project great things. Let’s take a quick look at the elite prospects on three rosters:

Syracuse (Too High in My Projections):

#50 recruit, 6’3” PG Kaleb Joseph

#79 recruit, 6’4” SG Trevor Cooney, ORtg 122, 90/240 or 38% of threes last year

#28 recruit, 6’7” G/F Michael Gbinje, ORtg 111, passable 15.6% assist rate as backup PG

#37 recruit, 6’8” F Tyler Roberson, ORtg 89, best DR% on team last year

#24 recruit, 6’10” F Chris McCullough

#18 recruit, 6’9” C DaJuan Coleman, ORtg 109, almost always injured, but good per-minute rebounder

#21 recruit, 6’9” C Rakeem Christmas, ORtg 126 but far too passive with 11% of shots last year, best shot-blocker on team last year

Ohio St. (Too High in My Projections):

#32 recruit, 6’1” PG Shannon Scott, ORtg 101, 25.5% assist rate last year

#13 recruit, 6’5” SG D'Angelo Russell

#76 recruit, 6’3” SG Kam Williams, great shooter who red-shirted last year

#28 recruit, 6’5” SF Jae'Sean Tate

#46 recruit, 6’7” SF Sam Thompson, decent scorer, ORtg 105, but terrible rebounder

#22 recruit, 6’7” F Keita Bates-Diop

#66 recruit, 6’7” F Marc Loving, ORtg 101, aggressive scorer, possible break-out candidate as a sophomore

Temple Transfer, 6’9” F Anthony Lee, ORtg 107 on bad team, great offensive and defensive rebounder

#50 recruit, 6’11” C Amir Williams, ORtg 110, great offensive rebounder

UCLA (Too High in My Projections):

Unranked, 6’3” PG Bryce Alford, ORtg 110, 19% assist rate last year as backup PG

#53 recruit, 6’4” SG Norman Powell, ORtg 120, high steal rate last year

#19 recruit, 6’5” SF Isaac Hamilton, red-shirted after he could not get out of NLI

#12 recruit, 6’8”  F Kevon Looney

#36 recruit, 6’8” F Thomas Welsh

#69 recruit, 6’11” C Jonah Bolden

#24 recruit, 6’9” C Tony Parker, ORtg 108, strong rebounder last year

That’s a ton of players that were elite prospects coming out of high school. Given how well Jim Boeheim’s defense has performed in recent seasons, Syracuse should be ranked, no matter how much talent they lose. And Thad Matta might be the most under-rated coach in the nation. In the 12 years, Ken Pomeroy has been keeping track of the stats, Matta’s teams have never finished worse than 33^rd in margin-of-victory. Steve Alford has also done a brilliant job developing players over the last several years. For example, Kyle Anderson became a star under Alford, after struggling to find a college role under Ben Howland. Talent + Great Coaching = Teams that should be ranked.

But a word of caution is warranted. This version of my model is based on the mean projection for every player. Later this summer, when teams fully fill out their depth chart, I will run my full simulation and project scenarios where players are allowed to have good or bad seasons. When I allow for the possibility that one or more of these recruits are busts, teams like Syracuse and UCLA are going to have a very worrisome downside. For example, if Syracuse PG Kaleb Joseph isn’t ready to play major minutes, Syracuse’s season could be a disaster.  The whole point of the above exercise is simply to point out that as much as these teams lose, their primary rotation has the athletes to be competitive with anyone.

Of course, you might also ask why teams like Stanford, UNLV, and Memphis are not ranked because they are also filled with elite recruits. Personally, I’m particularly high on Stanford based on their returning players and recruiting class. But my statistical model basically says this: If you look at the last several years on a per-possession basis, Jim Boeheim, Thad Matta, and Steve Alford, have been great at molding their players into efficient contributors. Meanwhile, despite occasional flashes of brilliance, Johnny Dawkins, Dave Rice, and Josh Pastner have been a lot less effective. Even in making the Sweet Sixteen last season, Dawkins team was only 36^th in the nation in margin-of-victory.

Someone has to fill out the Top 25, and arguments can be made against all these teams. But that is always the case outside the Top 10-12 teams. For a description of the headings in the table, and a deeper run-down on some of the teams, click here.

Moving Up Since Early April

Team – Players That I Did Not Expect to Stay: Analysis

Kentucky – Aaron Harrison, Andrew Harrison, Willie Cauley-Stein:  Kentucky now has eight players who were former Top 20 recruits out of high school. And the other two, Willie Cauley-Stein, and Tyler Ulis, might end up being the most important players on the team next year. Cauley-Stein is an elite rim-protector. And Ulis is a true PG who will be a better defensive match-up against speedy perimeter players, as well as a true facilitator. With experience to go along with all its talent, Kentucky is the clear favorite. 

Louisville - Montrezl Harrell:  Russ Smith and Luke Hancock were high volume efficient shooters who will be hard to replace. But keep in mind how many minutes Stephen Van Treese played last year, and he basically never shot. In net, you have a team with elite high school recruits at every position, depth at every position, and thanks to the return of Montrezl Harrell, you have a clear offensive star. The offense should be fine. As I've said before, the thing that will be harder to replace is Russ Smith's ability to create steals. 

Dropping

Team – Players That I Did Not Expect to Leave: Analysis

Arizona – Nick Johnson (pro): Nick Johnson is not a consensus first round pick, so in my first set of predictions, I assumed he would return to school. But even if the NBA doesn't value him as a first round pick, that was how he played. Johnson was the most efficient of Arizona's starters, and he used the highest shot volume last year. Arizona has a stacked recruiting class, and has talent across the board. But the model no longer predicts the same big offensive jump now that the top offensive producer from last year is gone. The defense should still be elite, but keep in mind that Johnson was also better at grabbing steals and blocking shots than Gabe York and Elliott Pitts. His loss does hurt. 

Connecticut -DeAndre Daniels (pro): Daniels was a unique forward who could spread the floor, but who also had an uncanny ability to block shots. Lineups with Phil Nolan and Amida Brimah might be dominant defensively, but they won't have nearly the same great spacing on offense.

Minor note: Virginia lost Teven Jones to transfer. His playing time fell off last year, so this was not a huge surprise, but he did play 500 minutes in his career, and he was expected to be a backup guard next season. His departure makes a miniscule change in Virginia’s projection. This caused Virginia to fall below Villanova and Wichita St.

Dropping Out of My Top 25

Michigan – Mitch McGary (pro), Glenn Robinson (pro), Jon Horford (transfer): I thought Michigan could be a Top 10 team even without Nik Stauskas, but after losing basically the entire front-court, Michigan will have to deal with somewhat of a rebuilding process, at least early in the season. You are going to start hearing a lot about red-shirt freshman forward Mark Donnal in the coming weeks. Donnal was not a consensus Top 100 recruit, but he was a 4-star prospect, and he is the type of player that should have an impact offensively. But keep in mind that Donnal did not burn his red-shirt when Mitch McGary went down last year. The fact that he didn't says we shouldn't expect Donnal to be a superstar. If Michigan adds a JUCO forward, that would likely move the Wolverines back into my Top 25, but right now the front-court depth is lacking and Michigan sits at 26^th in my projections. 

Colorado – Spencer Dinwiddie (pro): Because of the sophomore leap, I expect a number of Colorado’s players to get better this year. But there is no way they can be ranked without Spencer Dinwiddie. Colorado’s margin-of-victory fell to 77th late in the season with Dinwiddie out. Saying that Askia Booker's experience can make up for Dinwiddie’s departure is a joke. Dinwiddie had an ORtg of 129 last year, while Booker's ORtg was 99. Top 100 recruit and PG Dominique Collier will help a lot, but given where he is ranked, he is probably about a year away from dominating at the college level.

Maryland – Nick Faust (transfer), Shaq Cleare (transfer), Roddy Peters (transfer): Maryland was my sleeper team, but they fall a little bit due to these departures. Maryland was still incredibly unlucky last season. By just winning a few more close games, they should be in the NCAA tournament hunt. And they bring in a great recruiting class. But without Faust, they no longer project as a Top 25 team.

Baylor – Isiah Austin (pro): I thought after Austin’s stats fell off in every area and after he was no longer projected as a first round pick that Austin might spend another year in college. I was wrong. I don't hate the Bears roster. Kenny Chery, Royce O'Neale, and Rico Gathers are all talented players. But they are not quite a Top 25 squad anymore.

Utah – Princeton Onwas (transfer):  Onwas departure knocks Utah out of my Top 25, but they are close, and this should still be the year that Utah returns to the NCAA tournament. (Utah also added an international center, but he hasn’t been fully scouted by the US scouting services, and it is hard to project a big impact for him this year.)

Moving into My Top 25

Iowa St.: Bryce Dejean-Jones committed to Iowa St. this month, and his addition moves the Cyclones from 27^th to 17^th. Dejean-Jones was a high volume scorer and efficient passer for UNLV.

Oklahoma: While Je’lon Hornbeak wasn’t terrible by any means, he was the least efficient player on Oklahoma last year, and swapping him for a JUCO PG didn’t drop the Sooners. Oklahoma basically moved into the Top 25 because some of the above teams fell out.

Michigan St., San Diego St., Pittsburgh: Gary Harris’ departure was not a surprise, so these teams did not add or lose any key pieces since my initial rankings. But with the above teams losing key pieces, the Spartans, Aztecs, and Panthers moved into the Top 25.

Next week: I will have a few more comments on the defensive projection for Duke and Kansas.