At this point, there's not much to say about this game specifically. At this point, it's all about the big picture. In two games here, Duke chose to go defense over offense without Kelly. In some ways that really seemed to help the defense, but it's hard to tell for sure given that FSU and Va Tech aren't exactly offensive juggernauts. Duke held their own on the glass for both games, and didn't allow great shooting nights from either one, but they only marginally increased their turnover output from what they had been forcing for most of the year. The other thing is that both games were low possession games so the efficiency increase isn't as great as it might seem (In the FSU game, they still gave up about a point per possession). Against a quality offensive team, I'm not sure that this defensive strategy works as well.
The offense was a complete mess in both games. Without Kelly, Duke can't use its spacing to stretch the D. The Plumlees will come out and set screens, but end up with the ball in their hands more than 15 ft from the basket way too much. There is very little that either one will give you from that range, either dribbling, or shooting. Without the spacing, you have to rely on screens and dribbling to get open, but the book is out on Duke's two favorite plays for that, a down screen on the wing, and a hand off up high. Those two plays were shut down completely. Even on simple passes around the perimeter, only Thornton would end up open, which is exactly the way other teams wanted it. Out of 17 3's in two games, about 14 or 15 were wide open (the others were end of shot clock scenarios), but he shot under 25% overall. There's no incentive for the other team to stay honest with those numbers on wide open shots. When you're covered that tightly, you think of two ways to beat it, passing and dribbling, but Duke isn't great at either of those. Duke is 287th in the country at A/FGM, and that rate has actually been worse than average recently. Without the spacing or shooting threats, teams can pack it in, which takes away cutting to the hoop and finesse driving. Whether it's at this level or the next level, the biggest thing Rivers needs to do is add muscle to drive with power. When he get near the rim, his drives usually end with a reverse layup, an underhanded shot, or a floater. His floater isn't at the level of a guy like Nolan Smith, so defenses let him have it. On the other shots, he has a very low release point, which makes it easy to block, or at least alter without fouling, and turns it into a lower percentage shot. Duke gets plently of shots from 8 feet and in, but they are always the lowest percentage shots possible. Even after all of that, everything is still one-on-one for this team against their opponents offense. I still find it baffling that for a team who only has one guy who's great at getting his own shot that they have such a low assist rate. This team has saved themselves all year with spacing, but with Thornton in the game and no Kelly, there are only going to be two three point threats that other teams are afraid of, and Duke's offense will stuggle big time in that scenario.
What I think Duke needs to do to help alleviate this is push the pace. If they're half court offense is struggling as much as it is, the best way to curtail that is to not let the defense get set. That's a lot easier said than done, especially with the sloppy play we've seen, but I think it's the best shot they have against a big team to score points. Against a small team, many of these problems go away, because the Plumlees have proven themselves to be able to dominate small teams, but Duke will struggle with big teams. The good news as long as Duke's one seed is not a big ten team (there aren't even any currently projected big ten one seeds), they have a 50/50 shot of drawing Michigan as their 3 seed, since the first three teams from a conference have to be in separate regions. Out of the projected three seeds, the only team that is really a bad matchup for Duke is Baylor. Michigan and Marqutte are small and both would be great matchups, and Georgetown has less size than Georgetown normally has. The seven through ten seeds are much more fluid, and there are definitely some bad matchups in the group. With or without Kelly, I think that matchups will play a bigger part in Duke's fate than it has in the past. That doesn't mean they can't beat a bad matchup or lose to a good one, but I'd say that it is less likely than in other years.