Dan Hanner once again is attempting to project every Division I team’s performance this season by using a lineup based approach. If you have ESPN Insider I highly recommend checking out his rankings and the methodology.
Dan was kind enough to answer a bunch of questions about teams of particular interest to Big Apple Buckets readers. What follows is the edited version of our transcript from earlier this month.
[Ed. Note: Even this edited transcript is almost 3,000 words. We’ve added sub-headings so you can find your conference(s) in the conversation, but you’ll learn a ton by reading through everything.]
Big East (plus Quinnipiac)
Big Apple Buckets: One of the first things I noticed is that St. John’s is ranked 44th overall with a variance in the rankings ranging from 21st to 76th nationally. The Red Storm are getting a lot of hype as being “the most talented team in the Big East” and are thought of as an NCAA tournament team. Why are they a little lower than expected here?
Dan Hanner: First off, I agree this is the most talented team in the Big East. They have six former Top 100 recruits and three elite JUCO recruits. No one else in the new Big East can match that. But I also think that the hype you are hearing may be due to your local area. Just as an example, College Sports Madness is doing its annual countdown and has St. John’s at #50. I’ve also seen other projections that have them outside the NCAA tournament.
BAB: You seem to be predicting that they will be squarely on the bubble. What do you think is holding them back? Sounds like it is the offense. I’m wondering if a point guard such as freshman Rysheed Jordan can make a big impact there.
DH: Yeah, the issue is the offense. Players like Chris Obekpa were very raw last year. Some of the players in the rotation may jump forward, but it seems unlikely they all will become efficient offensive players in one year. More than likely St. John’s will make the tournament as a deep, athletic, pressure defense driven team that sometimes struggles in the half-court, even with a new point guard.
BAB: I now see you have St. John’s as the 16th most improved team in Division I. It makes me realize how far that team has to go. All the “most improved” teams make a bunch of sense too. One interesting one is Quinnipiac, which you have as the 22nd most improved. What does the model like about the Bobcats?
DH: Quinnipiac is the classic case of having the right players coming back. The two least efficient starters are gone (Jamee Jackson and Dave Johnson.) And the team adds St. Francis PA’s leading scorer Umar Shannon as a fifth year transfer. Shannon’s efficiency should increase if he cuts his shot volume down a little bit. And that should happen since this is a much more talented team than the one he left. His addition should allow Quinnipiac to avoid breaking in a freshman point guard.
Northeast Conference
BAB: In the NEC I’m curious about Mount St. Mary’s and LIU Brooklyn. Both of those teams are rated considerably lower than I would’ve expected. Also, your ratings seem to really like Robert Morris. Is that related to Andy Toole’s ability to coach defense?
DH: LIU Brooklyn is largely about Boyd’s injury. I read that the coach expects him back in January, but that seems superhuman to me. I’m assuming Boyd will miss 90% of the season. If he comes back earlier, obviously they will be higher.
I’ve seen a lot of other projections that love Robert Morris so I don’t feel too awkward about that. Anthony Myers-Pate, Lucky Jones, Karvel Anderson, Mike McFadden were all efficient players last year and their return helps. But most importantly, Robert Morris is the rare NEC team to be recruiting at a higher level. Scout and Rivals loved freshmen Jeremiah Worthem and Kavon Stewart. Desjuan Newtwon was a Top 100 JUCO player. Robert Morris loses a lot, but they have higher potential pieces than some of the other NEC teams.
Mount St. Mary’s is the one that gets me. I agree with you, on paper this lineup looks pretty good. I think the impact of the star ratings helps the rankings a lot, but it really hurts a conference like the NEC that doesn’t get many recruits about the 2-star level. For 2-star recruits I use the number of offers a player receives as a proxy for that player’s potential. But Mount St. Mary’s is the kind of school that isn’t getting a lot of guys with multiple offers. Mount St. Mary’s tends to offer a very specific type of student that may only be getting one offer to their school. Thus my model thinks that most of the Mount St. Mary’s guys have low potential or have reached their full potential. But I agree with you that this projection feels a little too low.
BAB: The Robert Morris recruiting thing is particularly interesting to me. I know the Colonials – and Andy Toole specifically – are recruiting a different caliber of athlete. Out of curiosity, did that help Saint Francis Brooklyn’s projection (fourth in NEC at 9-7) too?
DH: Not yet. With St. Francis Brooklyn there are two stories. First, the defense fell off the map last year. That seems like a bit of a fluke. Granted we only have three years of data for Glenn Braica, so I don’t know that he is a good defensive coach. But regardless of the coach, when the defense falls off that much, it usually bounces back some the next year. Second, the two most efficient players on offense are back, Jalen Cannon and Ben Mockford. If they build around those two the offense can be passable. I’m still expecting Brent Jones to be a turnover machine, which is a problem, but this is really about the defense bouncing back somewhat.
MAAC
BAB: In the MAAC the top of your projected standings (Manhattan, Iona, Fairfield, Canisius) makes sense to me. I can understand the arguments, especially going with a lineup-based model, of everything I see through Rider’s 10-10 mark. I also appreciate Marist only being 7-13. Does that have to do with the GIANT hole the team has at PG? The Red Foxes have two all-league players at center and wing, but they have so many other weaknesses that everyone seems to overlook.
DH: The best thing Marist has going for them is that they won’t have to break in a lot of new players. They should have a very veteran lineup. My problem is that there aren’t a lot of reasons to expect their offense to jump forward at this point. We have a lot of data on Chavaughn Lewis at this point. He’s a high volume scorer, but year-after-year he takes too many bad shots. Meanwhile most of their players were sophomores or juniors last year. They didn’t have a lot of freshmen who you would expect to make great strides. The model basically feels like Marist was close to their ceiling last year.
BAB: Siena at 7-13 seems really low to me, not record wise but standings wise as your model has them 10th. I would think Niagara (which lost a TON including probably three of its top four players) and St. Peter’s would definitely be below the Saints.
DH: Niagara does lose a ton, but the players they bring back (like Antoine Mason) were efficient.
Meanwhile, St. Peter’s three most efficient (only efficient) players Desi Washington, Chris Burke, and Markese Tucker are back. And St. Peter’s adds transfers Jamel Fields and Marvin Dominique. Those guys weren’t particularly highly rated or particularly efficient at their last school, but players that sit out a year via transfer do tend to do pretty well when they join the lineup. That’s a core they can build around. If you think they should be 6-12 instead of 7-13, I’m not going to argue with that, but they will be better than last year.
BAB: Reasonable. I’ve talked myself into 9-11 for Siena thanks to Jimmy Patsos’ ability to coach offense in the MAAC, but maybe I’m overrating that skill.
DH: First, I want to give myself a pat on the back for last year’s Siena prediction. My model had a lot of errors, but that was one thing it got right. I took a lot of flack for ranking them so low, and they have a very passionate fan base that deserves better. But as I said last year, their lineup looked exceptionally weak on paper.
We agree Siena will be better with their new head coach. The question is how much of an impact Pastos can make in year one. I’ve studied coaching changes extensively and I can tell you these things are across the board. Some coaches make an impact in year one, sometimes it doesn’t happen until year two, sometimes until year three. If you think they will be more improved in year one, I’m not going to argue with it. That is very plausible.
Ivy League
BAB: The Ivy League actually had the most surprises for me. Part of it has to be because of all the parity below Harvard. I was shocked to see Princeton second. Mostly because I think the consensus is that Pennsylvania is the team that can go 10-4 and at least keep things interesting in the Ivy race for a little while. I believe Denton Koon is going to have a monster season and I also like Hans Brase to breakout, but is there something else I’m missing?
DH: Look; everyone loves Penn because they bring back 100% of their roster. That’s hard to do. But that’s a mixed bag. Penn had the 243rd margin-of-victory in the nation last year. Harvard is a borderline Top 25 team. Are all those Penn players going to miraculously start playing at a much higher level just because they are all back?
Losing Ian Hummer is just a real bummer for Princeton. He was high volume and efficient. When players like that leave, other players have to shoot more, and they usually suffer a little because of it. I do love Koon though. I love Brase a little less (based on his past efficiency and 2-star rating), but if you’ve seen reasons to believe he will break out, you almost certainly have more information than my model.
Someone will probably have a better season (just by luck) and make the race close, but I wouldn’t necessarily peg any of the teams to do it.
BAB: What I find most interesting about Penn is that they don’t seem to have the upside of some of the other Ivies either. For instance the “Best Case” for Princeton and Yale was 93rd and 96th respectively, whereas Penn’s is 120th. I guess that’s because all the minutes are returning so we know what this team is essentially.
DH: Exactly, that’s what the simulation adds. In some ways it is counterintuitive because we usually have really high expectations when teams return a lot of minutes. Penn will be better, but expecting them to be great with the same lineup is asking too much.
BAB: What about Dartmouth? The Big Green’s projection seems rather low to me at 5-9. They have a young core that with another season of development could be a surprise contender or at least 7-7?
DH: Dartmouth has a young core and I do expect big improvements for players like Malik Gill this year. The sophomore leap will help this team. But they did have the worst margin-of-victory in the league last year (mostly based on non-conference play) and overachieved to finish 5-9. This is one of those cases where I agree Dartmouth should be better, but their conference record might not show it.
BAB: Then there’s Columbia. I know the Lions got really unlucky last season, but if you offered fans 6-8 right now, I bet most Columbia fans jump at the offer. What’s boosting their ranking after losing Brian Barbour and Mark Cisco who were two huge pieces of the rotation?
DH: Not enough people have mentioned the return of Columbia’s Meiko Lyles. Last year I expected Columbia to be competitive for the Ivy League title, but what I didn’t know when I ran my projections is that Lyles, the team’s best outside shooter, was going to miss the entire season due to personal reasons. And even though Columbia had the fourth best margin-of-victory in the Ivy League last year, they collapsed to a last place finish. Unfortunately, Columbia does lose two of its best players, Barbour and Cisco. So the team would have been a lot better off with Lyles on the roster last year. But Lyles is back and that’s one of the reasons I expect Columbia to finish in the middle of the Ivy League and not the cellar.
Horizon League
BAB: A random aside from my former Chicago days and the Horizon League. I was really surprised by Milwaukee’s 7-9 record. What’s driving that?
DH: If you look at Rob Jeter in his career, he has been a very successful coach. So there are two schools of thought here. One, Jeter lost it last year when the team wasn’t winning and he can’t get it back. Another school of thought is that last year was a fluke. His player development and defense were far below their historical levels. The model thinks that last year was a fluke and the team will bounce back, but I agree with you that this projection may be a little too optimistic given the personnel.
Patriot League
BAB: Another random thing I noticed is that Army is projected to go 8-10 in the Patriot League, which I think would be a big bummer for Knights fans.
DH: Just as Princeton will suffer without Hummer, Army will struggle without Ella Ellis. Ellis was an efficient high volume scorer and those guys are just hard to replace. Everyone else has to shoot more and that usually leads to lower efficiency. It’s funny, when a player like Otto Porter leaves, everyone says, “Oh yeah, Georgetown is going to be worse.” But when players like Hummer and Ellis leave, you don’t hear the same commentary. But Hummer and Ellis were just as important to their team as Otto Porter.
(For further analysis into Hanner’s Patriot League projections, go here)
Colonial Athletic Association (and the Templon Rule)
BAB: Tell me about your Towson prediction. A lot of people think they are the CAA favorite.
DH: Let me stop you here. You caused me to add a new line of programming code to my process. For simplicity, let’s call it the Templon rule.
In almost every projection model I have ever run, I’ve noticed that it is important to control for the coach’s ability to develop players. Certain coaches are just particularly good at turning less heralded talent into efficient basketball players. (Think Bo Ryan at Wisconsin). Thus I use the historic player data to determine a player development factor for each coach.
But one dilemma is what to do about coaches with little or no historic data. In last year’s rankings, I just assumed that if a coach had one or two years of data, that whatever he did in those first years would predict his true ability. But since then I’ve studied at the data and concluded that it is hard for coaches to have big improvements multiple years in a row. Thus for new coaches with 0-2 years of data, if I just used their own data, the expectations were often unrealistic.
That led me to make a change this summer. For new coaches, I now set their player development factor based in part on the historic performance of the coach that preceded them.
But when you asked me about the Towson prediction in September, that really struck a chord. Evaluating Towson head coach Pat Skerry based on what happened under Pat Kennedy seemed unusually pessimistic. Thus I’ve changed the formula again. I now evaluate new coaches based in part on their own expectations, and in part based on what you would expect for an average coach in the conference. Call this the Templon rule. My prediction for Towson moved from 164th to 137th thanks to this change.
I still have Towson as the 3rd best team in the CAA, behind Drexel and Charleston, but I now have them in a tighter race for the title.
BAB: I get why Drexel is highly rated thanks to the return of Chris Fouch, but it’s interesting to see the Dragons so far above the rest of the conference.
DH: Drexel was a really big beneficiary of the simulation model. This is about Frantz Massenat and Damion Lee as much as much as Fouch. When you have two efficient high volume scorers, you should be able to find role players to compliment them. For some reason Drexel couldn’t do that last year, but it shouldn’t be that impossible.
America East
BAB: Finally, one team that I see dropped a lot, but is still incredibly competitive in America East is Stony Brook. Is that related to the graduation of Tommy Brenton or are there other factors at play?
DH: For Stony Brook, last year’s defense looks like a bit of a fluke. They’ve never had a defense better than 90th and suddenly they were 34th in the nation. I think we need to realize that college basketball is all about small sample sizes and that teams can get in a groove defensively for a limited number of games. On a personnel level, arguably the loss of undersized rebounding monster Brenton will set the defense back a little. The model likes Stony Brook to be competitive for the league title without winning it hands down.
Thanks again to Dan for taking us through his model and a all of the thought process that went into it. I hope people learned a lot from this Q&A. You can check out all of the rankings on ESPN Insider. For more on his approach you can read last year’s Q&A.
NYC Buckets Archive 