Perhaps Virginia, not Duke, Should be the ACC Preseason Favorite

Over the past two seasons, using returning win share totals to predict the standings in the ACC has been a fairly successful endeavor, nailing the regular season champion in both cases. The only huge swings and misses came last year with Virginia Tech and Maryland, and there is a good explanation at least in the latter case.

Maryland's total did not account for two things: what Alex Len would have done in a full season and the immediate eligibility of Dez Wells, who ended up being responsible for 4.4 wins but wasn't included in the total below because his status was undetermined when the totals were rendered.

I don't think anyone could have predicted the impact losing Dorian Finney-Smith to transfer and Dorenzo Hudson to full-time drug sales would have on the Hokies. Everyone on the roster not named Erick Green took a collective step back in productivity, with Jarell Eddie and Robert Brown being particularly egregious examples. Brown was the worst player in the ACC last season by win share total, coming in at -0.8 after posting a 1.3 total the season before. Eddie slumped from 3.3 to 1.8. Some folks just aren't cut out for increased usage. Regardless, having the Hokies in the top half of the league was a clear strike against the win shares prediction method.

Overall, though, it's painted a fairly reliable picture. Here are the last two seasons with teams ranked by returning win shares with where they actually finished listed as well.

Though not perfect--win shares can't, for example, predict Lorenzo Brown's injury that likely cost N.C. State a win or two last year--analyzing win shares is useful because it can show us two things: 1) a team's level of experience, and 2) whether a team is already successful. In other words, if a team is not already getting a decent amount of wins, there aren't going to be many wins to share.

The experience factor cannot be emphasized enough. Last year I looked at the class of 2011 McDonald's all-Americans and found that they averaged just three win shares apiece. Take out the four best Burger Boy freshman campaigns--Anthony Davis (11.3 WS!!!), Cody Zeller (7.2 WS), Michael Kidd-Gilchrist (6.3 WS), and Brad Beal (6.0 WS)--and the average of everyone else was just two win shares. The 2012 class had even fewer immediate impact players, with Anthony Bennett (5.7 WS) and Marcus Smart (5.5 WS) the only two Burger Boys to best five win shares. T.J. Warren's 4.2 win share season was actually the fourth highest total in a class that averaged just 2.8 win shares per man. The best programs are the best programs because they get the best players, so it stands to reason that some of these guys were limited in what they could contribute due to playing time constraints. Nonetheless, even the most highly touted freshmen are not guaranteed to make an immediate impact.

That's good news for Virginia, the most experienced team in the ACC, and it's bad news for N.C. State, a team that lost four starters to graduation or early entry to the NBA (not to mention its sixth man to transfer). The following table shows that the Wahoos, like Miami last year, could be poised to make the media experts, who nearly unanimously picked Duke to win the conference, look bad. And, with the exception of Notre Dame, a team pegged fifth by both the media and the returning win shares measure, perhaps all the shiny new things in the ACC are ranked too high by the media. Syracuse, picked second by the media, has perhaps the best player in the league in C.J. Fair but lost three of its four top players; Michael Carter-Williams, James Southerland, and Brandon Triche combined for 17 win shares last season. Pittsburgh will be without four of its top seven contributors in 2013-14.

Duke should be buoyed by the transfer of Rodney Hood and the return of Andre Dawkins, but the team win share total above includes their totals from 2011-12. The Blue Devils will be without Plumlee, Seth Curry (5.8 WS), and Ryan Kelly, who helped account for 3.5 wins in just 23 games. Even if Jabari Parker is the rare freshman who immediately performs on a level that meets the hype, Duke seems to me more like a blue bias pick to click rather than an obvious preseason favorite.

Maryland may seem too high, but remember that the Terps add a proven transfer in former Michigan forward Evan Smotrycz, a stretch four who posted a 3.1 win share season two years ago (that figure is included in Maryland's total).

Boston College fifth? That seems like a stretch, but remember that the young Eagles ended last season on a high note, winning four straight games before a season-ending loss to eventual champion Miami in the second round of the ACC tournament. BC led that game at the half and stuck around pretty much to the final horn. Steve Donahue's squad doesn't lose a single player of consequence and adds big wing Alex Dragicevich, a 6-7 transfer from Notre Dame who posted 1.7 win shares in 2011-12.

Georgia Tech ahead of UNC? Tech's only significant loss was Udofia, and that could be addition by subtraction, especially considering that the Yellow Jackets have an upgrade waiting in the wings in Tennessee transfer Trae Golden, who was about a win better than Mfon a year ago (and rocks a #beardgang worthy face sweater). Golden has been granted a waiver by the NCAA and can play immediately. Meanwhile, the Heels lost their best player and will rely heavily on P.J. Hairston to carry the load. Assuming Hairston doesn't miss a significant amount of playing time due to his off-the-court problems, he still may be distracted enough by the hot water he's been in to falter under the pressure of being the man. And the Heels' win shares are already lower than Tech's without trying to factor in all the ways Hairston might sabotage the season. UNC will be very dependent on freshman contributions, and freshman contributions are not exactly an Ol' Roy hallmark.

After two years of challenging the blues for ACC supremacy, it is hard to see a path to the top half of the conference for Miami and N.C. State. Rion Brown's 2.1 win shares are the lowest of any team "leader" not from the enriched-uranium fueled dumpster fire at Virginia Tech. And the Canes only get to 7.2 total win shares thanks to some generous accounting. I added two for Garrius Adams from his last healthy season, which was two years ago. Donnavan Kirk also returns for Miami, where his career started in 2009, after a hiatus at DePaul. Kirk totaled 1.5 win shares last year for the Blue Demons and those numbers are included in the Canes' total.

N.C. State has a star to build around in Warren, who led the league in field goal percentage last year while scoring an efficient 12.1 points per game, though it remains to be seen if he can maintain his efficiency as the focal point of the offense (and the focal point of opposing defenses). Tyler Lewis averaged .109 win shares per 40 minutes; for comparison sake, UNC's Paige (a fellow Mickey D's all-Amercian) posted a .087/40 mark. Lewis will likely at least triple his one win share from his light usage freshman campaign with more playing time, but that alone will not be nearly enough to take the Pack back to the top. Ralston Turner's last season at LSU netted 1.9 win shares, which I included in the total on the table. Turner's experience will be useful to a young team, but he's likely not a difference maker, and that means it's up to JUCO Desmond Lee and a bunch of freshmen. And that likely means that N.C. State is going be good again...in 2014-15.