Duke isn’t a perfect team. Their defensive rebounding remains suspect. On Saturday Night, it felt like Louisville could take as many bad threes as they wanted because the Cardinals always got the rebound. Ryan Kelly usually takes the blame. He wasn’t a great defensive rebounder last year, and that trend continues into this season. But the moment you start to question one of the Blue Devils, he finds a way to surprise you. Just when you thought that Ryan Kelly lacked strength on the interior, there he was at the end of the Louisville game getting a key block that led to a run-out and basket on the other end. And that four point swing was critical in a five point win.

Meanwhile, Kelly and Mason Plumlee were vital in beating VCU’s press, and that allowed Duke to escape the VCU game with a 12.3% turnover rate. That was the best job a team did holding onto the ball against VCU since the 2011 NCAA tournament. Meanwhile Plumlee and Kelly also matched the physicality in the Battle 4 Atlantis tournament. While the refs called plenty of perimeter whistles in the Bahamas, they seemed to swallow their whistles on the inside. And Duke managed to avoid being overwhelmed by the physical strength of players like Minnesota’s Trevor Mbakwe.

Duke may not be #1 in the polls, but in terms of accomplishments, no one has more quality wins than the Blue Devils at this point. They’ve beaten two preseason Top-5 teams and two more probable tournament teams. And as we near the end of November, we are once again reminded that no one has their team ready to play at the start of the season better than Mike Krzyzewski.

Year after year his teams win these preseason tournaments, and they do it by simply making fewer early season mistakes than their counterparts. With Duke nursing a three-point lead in the final minute of the Louisville game, the shot-clock was winding down. Duke was going to have to settle for a tough shot against a Louisville defense that had been shutting them down all day. But then Russ Smith made a foolish gamble. He dove for the ball instead of playing straight up, and Quinn Cook spun into the lane for a beautiful floater that sealed the victory. Duke made the right play, Louisville didn’t, and the Blue Devils won again.

That doesn’t necessarily mean that Duke will be the best team in March. Kentucky is still very young. Gorgui Dieng was out of the lineup for Louisville. And as the analysis below will show, Duke’s early season success doesn’t always translate into March success. But give Mike Krzyzewski credit for dominating the early season once again.

Other Impressive Teams

- Is the key to the early season simply good guard play? Illinois is off to a 7-0 start, and the Fighting Illini made at least ten threes in the first six games of the season. They broke the streak of outside shooting against Gardner Webb (and almost lost the game), but Illinois’ guards have looked fabulous so far this season.

Meanwhile, Cincinnati has rode its guards and looked incredibly impressive in its 6-0 start. In almost every game so far guards Sean Kilpatrick, Cashmere Wright, and JaQuon Parker have scored in double figures. In the preseason I wondered if the inside play of Cincinnati’s forwards would drag the offense down enough to stop the Bearcats from becoming an elite team. And the truth is Justin Jackson has continued to struggle with turnovers. That’s a three year trend that may not be correctable at this point. But Mick Cronin hasn’t felt obligated to give Jackson major minutes. Instead, he has been giving plenty of time to JUCO transfers Titus Rubles and David Nyarsuk, and the early returns are good. Not only have those players been able to finish the lay-ups and dunks that have come their way, their defensive intensity has allowed the Bearcats to reach even higher levels than last season. Cincinnati’s two point defense is near the tops in the nation right now. It helps that Cincinnati has played four of the worst teams in D1 in the early season. But Cincinnati also stifled Iowa St. and Oregon in winning the Global Sports Classic in Las Vegas, so there is some reason to think their defense is for real. I still think the question marks in the post could hurt Illinois and Cincinnati in conference play, but the early returns are fantastic.

Of course, perhaps we should continue to be reminded that the season is still early and we still have a small sample of games. No one looked more impressive than Baylor to start the season. The development of Cory Jefferson into a dominant player and the impressive debut of Top 10 freshmen Isaiah Austin had us all raising our expectations for Scott Drew’s club. But with Brady Heslip out of the lineup after having his appendix removed, Baylor couldn’t produce any points against the College of Charleston and the Bears fell at home on Saturday. Pierre Jackson and Heslip are outstanding guards, but the dropoff to Baylor’s backup guards Gary Franklin and AJ Walton is huge. Franklin and Walton have experience, but they just aren’t good offensive players.

Bullets

- Stanford’s Chasson Randle’s would like to erase the Minnesota game from his memory bank. Randle was 3-for-19 from the floor which was bad enough. But he may have committed the worst play of the season when he fouled Minnesota’s Andre Hollins at the end of the game. Hollins took a desperation half-court heave in a tie game, and Randle inexplicably fouled him. Hollins calmly sunk three free throws to give Minnesota the win. Hollins scored 41 the previous day against Memphis, and Stanford held him in check for most of the game, but Randle’s foolish foul allowed Hollins to win it at the end. Despite the dumb play, I love the postgame comments from Stanford coach Johnny Dawkins who tried to deflect some of the blame from Randle. Dawkins noted that one play doesn’t decide a game, and that Stanford made other mistakes that contributed to the loss.

- My biggest question for Michigan was whether someone could step in and fill the spot-up shooter role (played by Stu Douglass and others in past seasons.) So far Nik Stauskas has been more than up for the challenge. I don’t think his current efficiency numbers are sustainable, but the 78th ranked player out of high school has been much better than advertised. Certainly playing alongside Trey Burke and Tim Hardaway helps, but as teammate Matt Vogrich has shown, simply playing with other good guards doesn’t guarantee you will become a dominant scorer. I don’t know how much longer John Beilein can start Vogrich ahead of Stauskas.

- Northwestern held off Illinois St. in OT (thanks to a late technical by Illinois St.’s Jackie Carmichael), and that gave the Big Ten its fifth title in the nineteen Feast Week tournaments, most of any conference. That has given the Big Ten the early lead in most ranking systems from conference RPI to Sagarin.

- Oregon got 44 points from freshmen in the team’s 83-79 win at UNLV. With those new players and Arsalan Kazemi surprisingly declared eligible, Dana Altman has some great new pieces to work with.

Early Season vs Late Season Coaches

Team X is “going to be so much better by the time conference play rolls around.” I heard this phrase a lot during Feast Week. But I hate to break it to the announcers. Even if everyone improves, some teams are going to fall behind the curve and look even worse in conference play.

Two weeks ago, I talked about how freshmen-led teams have the most potential to improve, (unless the losses pile up and they stop playing defense.) But another question I have been pondering is whether certain coaches are particularly good at developing their teams over the course of the season. The results are presented below. At some level, they make sense to me. Duke hasn’t lost an early season tournament game since 2006. And Krzyzweski has more wins in the ACC/Big Ten challenge than any other coach. Clearly, Duke is at its best in November. And I think Tom Izzo clearly has a reputation as a coach whose teams improve over the course of the season.

But I am hesitant to say that these numbers have much predictive power. By random chance, some coaches are going to have teams that get better over time, and some will have teams that get worse over time. This doesn’t guarantee that Michigan St. will be better in March or that Duke will be worse in March again this season. But whether they have predictive power or not, the past trends are worth examining.

A few notes on the tables. I calculated the average Pythagorean winning percentage for each coach for the time periods listed duplicating the formula used on kenpom.com. Then I averaged those splits over a coach’s career. At most this includes 10 years of data for each coach. This doesn’t include the current season, or the near death-penalty 2006 season at Baylor since I cannot construct splits for those seasons. In many cases, such as Bill Self’s splits, the differences are so small as to be meaningless. Bill Self is dominant in all seasons. The Pythagorean winning percentages are listed in the following tables:

Coaches that Are Great Early

Unfortunately, I think there is a negative stigma associated with the coaches on this first list. Rich Barnes teams have often looked dominant early in the year only to disappoint as the season progresses. But instead of praising Barnes for getting his team up to #1 at some point in the season, we all view him as a “choker” for not getting his team to the Final Four more often.

There also seems to be a negative stigma associated with blasting small schools early in the season. Wisconsin absolutely punishes the small schools it faces, and that raises their power numbers, even though a 40-point win against Presbyterian may not have as much predictive power as the top ranking systems give it.

But that is a very negative way to look at this table. There is real skill to having a team prepared to play in November and December. Mark Few’s teams have typically been fabulous in the Feast Week tournaments, and we respect him for that. So don’t bash Mike Krzyzewski for having his team ready to play in November.

One other trend in the table is worth noting. The best mid-majors tend to show a pattern of less impressive play in January and February. (See Dave Rose and Mark Few.) I think the issue here is that it is hard to blow out conference opponents by 20 points every night, even if they are inferior teams.

The mid-major effect partly explains Mick Cronin’s poor numbers as this includes his time at Murray St. Also remember the 2007 season where Cronin’s Cincinnati team was 9-4 at the end of December with wins against three Top 100 teams, only to go 2-15 the rest of the way. Cronin inherited a team in chaos at that point in time, and I wouldn’t necessarily say it reflects on his future expectations, but it does cause his splits to look pretty poor. I wouldn’t expect a collapse this season based on past performance.

Coaches that Dominate Mid-Season

It is no surprise Bill Self would be on this list given how his teams win the Big 12 ever year, but the truth is Kansas is usually pretty dominant in all seasons. Ben Howland on the other hand, typically needs a couple of months before his teams get on a roll. Buzz Williams numbers look a little lower than expected because this includes one season at New Orleans.

Coaches at their Best Late

I love the contrast of Roy Williams and John Calipari here. Both coaches have had their teams playing at a peak level in March, but John Calipari’s teams have relied much more heavily on in-season growth. John Calipari’s numbers are even more impressive here when you consider that this includes his time at Memphis.

Many of the coaches like Rick Pitino, Tom Izzo, and John Beilein have truly been at their best in the NCAA tournament and everyone is aware of that. But I didn’t want one tournament loss to skew how a team was playing late in the year, so I used the whole month of March instead of only the NCAA tournament. Thus Leonard Hamilton and Frank Haith also get credit as late-season coaches based on their ACC and Big 12 tournament title wins.

Follow @DanHanner