If we were actually the top offense and 35th best defense in the country right now, stating four freshmen and a senior with two sophomores coming off the bench, I'd be pretty thrilled with that. We're only going to improve as the season progresses.
Kenpom's pre-season rankings are up: https://kenpom.com/
Duke checks in at #6 overall, with the projected #1 offense and #35 defense.
Not too surprising that we would be ranked so "low" since we finished #14 and #17 overall the past two seasons. We also have a lot to prove defensively.
If we were actually the top offense and 35th best defense in the country right now, stating four freshmen and a senior with two sophomores coming off the bench, I'd be pretty thrilled with that. We're only going to improve as the season progresses.
Not sure evidence supports that.
We had one year in the one and done era in which the defense came together down the stretch. Unfortunately we don't have daily data for the past 7 years, but duke has typically started pretty highly, and often has dropped significantly as the year goes on.
Now, (again from memory), we usually pull that number back somewhat from the low points of the year, but I don't believe it ever really reaches the preseason expectation.
1200. DDMF.
How does Ken even do rankings for offense and defense before a single game has been played?
"We are not provided with wisdom, we must discover it for ourselves, after a journey through the wilderness which no one else can take for us, an effort which no one can spare us, for our wisdom is the point of view from which we come at last to regard the world." --M. Proust
It is basically meaningless. But he does it based on a combination of returning player contributions, incoming freshmen (based on historical production of freshmen), and the program’s recent performance.
But this year’s team really doesn’t work for any type of preseason statistical analysis.
His "method" actually works only after a sizable number of games have been played -- as in, like, January. But he has subscribers who are customers, and customers want estimates for the entire season. So, as I understand it (I heven't pored through the details), he has come up with a preseason measure based on last season, returning players, new players, and a large dose of heuristics.* Then he gradually reduces the "preseason estimate" as he gets more data from 2017-2018 games. By early-to-mid-January, his data-driven estimates are 100 percent.
Kindly,
Sage Grouse
*Judgment
Sage Grouse
---------------------------------------
'When I got on the bus for my first road game at Duke, I saw that every player was carrying textbooks or laptops. I coached in the SEC for 25 years, and I had never seen that before, not even once.' - David Cutcliffe to Duke alumni in Washington, DC, June 2013
I know the injury vs. OAD debate has been litigated a number of times, but I think this is a generalization that requires a little bit of nuance. KenPom does not factor injuries into his numbers. Duke has had poor (relatively speaking) defensive numbers in four out of the last seven seasons so 35 sounds about right to me in a vacuum. However, Duke also had major injury issues in three of those four bad seasons.
It's not surprising Duke's overall rating would more often decrease because that is the more likely for a team ranked near the top. Last year, Kenpom started Duke number one, but all that meant is that Duke had, per his numbers, the best percentage chance of being number one in the rankings. I'd bet that the actual chance Duke had of being number one was really, really small in isolation. (Kentucky 2x, Louisville 2x, and Duke 3x are the only teams in the KenPom era with multiple number one rankings, so Duke is 3/17 over that span, good for 18% which actually sounds pretty high to me) So any team ranked at the top has a MUCH better chance of falling lower than remaining on top.
Then, of course, injuries have also plagued Duke teams. Duke didn't have any major players miss significant time save for Carlos Boozer from 1999-2010, which was pretty phenomenal. Yes, some guys had some ailments, obviously, but there were no extended absences, especially from the teams' stars. That trend reversed in a big way starting with 2011 with Kyrie, 2012 and 13 with Ryan, 2016 with Amile, and 2017 with basically everyone. I definitely think there is strong evidence that veteran teams were able to handle big losses better than younger teams. 2011 was capable of winning the title with or without Kyrie. 2013 hung in there during Ryan's absence and was also a legitimate contender at the end of the season. Meanwhile, 2012 fell apart after Ryan's injury. 2016 played well on offense but never got it together on defense after Amile's injury. Meanwhile, despite the disappointing finish, the veteran-laden rotation for the 2017 team actually did pretty well considering all the injuries and drama, recovering to win the ACCT.
The team that relied on OAD's the most was 2015, which also happens to be the team that managed to improve the most on D by the end of the season. And I think the 2017 team was hurt by NOT being able to rely more on a healthy and freaky-athletic Harry Giles . . . so one could actually argue that 2017 ended up worse because we weren't able to rely on a OAD star more.
This isn't to say that there isn't good evidence that relying on OAD's makes for a wider variance in results. I would argue that the reasons for this have less to do with having any particular OAD on the roster and more to do with the overall higher turnover in personnel between years. That is definitely a concern for this coming season. But I think there is reason to be optimistic that if Duke stays healthy, Duke will be good, even with OAD talent.
Who needs a moral victory when you can have a real one?
Do exhibition games count in the Kenpom stats
Okay, I know that there are much better statistical measures of a season that RPI.
But the RPI remains the selection committee's favorite tool, so we need to pay attention to it.
I also know that it's too early to pay much attention to RPI rankings (or anything else, such as Ken Pom). Still, just for fun, I checked out the RPI as of today:
https://www.teamrankings.com/ncb/rpi/
Duke is 19 today ... and it's not likely to get a lot better until ACC play starts,
Duke's upcoming opponents are not all that -- Portland State is 339, Butler is 106, Texas is 98, Florida is No. 7 in the nation in the AP and No. 248 in RPI. Gonzaga is 148. Indiana is 203.
I take those rankings of evidence as to how the early season RPI can be warped. Still, it will be interesting to watch the rankings evolve as we get deeper into the season.
should this be wrapped up in a dork polls thread?
In kenpom news, arizona's demise bumped us up.
1200. DDMF.
I expect our strength of schedule at the end of the season to be pretty bad, and in turn our RPI. Weak OOC schedule outside of MSU, and a favorable conference schedule. Which is good in the sense that we’ll take fewer losses, but bad in terms of computer numbers.
RPI Forecast uses Sagarin to project the season, and Duke ends up as #3 RPI and #17 SOS. Not bad at all. We have enough tough games on the schedule. As mentioned elsewhere, I don't think the ACC will suffer through a down year.
Interesting tidbits to me...
KP lists Grayson Allen with a sensational oRTG of 139.7, good for 17th in the nation among all players, not just qualifiers. What is of further interest to me is that KP still lists G as a "Role Player" at 16-20% possession usage. For reference, Trae Young is #1 in the country in oRTG at 130.3 for all players with at least 28% usage. Not coincidentally, KP lists Trae as his early frontrunner for NPOY.
Pivoting for a second to Trae Young, he has been exceptional this year, averaging 28.8 and 8.8 assists, dropping 20+ in 5 of his 6 games. But I wonder if Oklahoma will keep pace in the Big 12 enough to keep Trae up near the top of the NPOY race. Similar issue to Luke Maye, sort of.
On this young season, my NPOY shortlist is Bagley, Devonte Graham, Colson, Bridges on Villanova, and probably this Murphy guy on Minnesota, all of which are listed on KP's listing. If Trae Young and Luke Maye can keep it up, then they'll be in the mix - I just don't think they will, at least at this level. The other guys I think will.
- Chillin