First KenPom of the Year

[Upstate]

http://kenpom.com/index.php

SU clocks in at 11.

 

[HoustonCuse]

We play five of the teams in the top 15. And North Carolina isn't one of them.

 

[cuseattle]

I would like to be the first person of the year to say I never agree with KenPom. Too much stats, not enough eye test.

 

[jordoo]

So in all the games we haven't yet played we have performed as the 11th best team?

 

[CaptainJ]

So in all the games we haven't yet played we have performed as the 11th best team?

Maybe it was based on the Midnight Madness type of events.

 

[jordoo]

Maybe it was based on the Midnight Madness type of events.

LOL, I think its baseless at this point other than to create a baseline to input numbers later.

 

[Cusefan0307]

I believe it's based on returning players historical level of improvements plus historical production of recruits. I'm not positive 100%, but thats how I gather he gets to these rankings. Also for anyone complaining about Kenpom rankings in preseason, it's probably more accurate than the human polls and always has been.

 

[CusefanATL]

just out of curiosity, how can 1 use data driven rankings with no relevant data?

 

[CaptainJ]

KenPom definitely has it's advantages, but if Nerlens Noel had been injured last year in the preseason, would the ranking have changed? I find it interesting either way and thanked the OP.

 

[jordoo]

I believe it's based on returning players historical level of improvements plus historical production of recruits. I'm not positive 100%, but thats how I gather he gets to these rankings. Also for anyone complaining about Kenpom rankings in preseason, it's probably more accurate than the human polls and always has been.

I agree knowing his system somewhat he has to have a formula that he has tweaked over the years to come up with these ratings. I'm sure it includes returning players, recruiting class rankings, recent performance and a bunch of other stuff. Still its pretty meaningless like the polls are until the teams start playing. That said I wouldn't be shocked at all if his is more accurate that the polls from year to year.

 

[Cusefannotindc]

just out of curiosity, how can 1 use data driven rankings with no relevant data?

He had an explanation of it a few years ago. It relies on the previous two years Offensive and Defensive Efficiencies because they are very good predictors of the next years OE and DE. It also makes adjustments for expected improvements in returning players (and additional court time) and for new players, I believe, based on their recruiting rankings.

 

[pfister1]

He had an explanation of it a few years ago. It relies on the previous two years Offensive and Defensive Efficiencies because they are very good predictors of the next years OE and DE. It also makes adjustments for expected improvements in returning players (and additional court time) and for new players, I believe, based on their recruiting rankings.

Sounds a lot like how I put together my eye test based rankings.

 

[CusefanATL]

Sounds a lot like how I put together my eye test based rankings.

but you can use your eye test to guess how much the freshmen class will impact a team.

 

[CaptainJ]

I wouldn't be against pfister or any of us doing individual and composite pre-season rankings to see how they turn out vs KenPom. Heck, that might make a cool contest, in my opinion.

 

[VaBeachOrFan]

Useless to even debate pre-season rankings for SU this year more than most, given that we're missing two forwards, a point guard and last year's back-up center. That leaves only Cooney and Silent G as "proven" players on which to base knowledgeable estimates. Sure, CMac and KJ are high-ranking frosh, and JB thinks highly of Roberson, but nobody's seen how the pieces fit and what the team's MO will be this year. I can only hope it'll be a high-speed, high-wire aerial offense with only a dribble or two between passes -- but that's just my fantasy. All TBD, but the waiting is agonizing.-VBOF

 

[BuffOrange]

Lol kenpom. Eye test FTW all the way. My ability to compare Illinois St to Alabama based on when they're arbitrarily on TV and when I happen to be watching, is impeccable.

 

[moqui]

if his rankings hold close to form, SU is going to have a very good non-conference strength of schedule. 5 of the first 10 opponents are in the top 50 and two of those are true road games (19 Texas or 32 Iowa at MSG, at 15 Michigan, 39 St. John's & 49 La. Tech at home then at 9 Villanova). That should more than compensate for the two true dogs on the sked - 322 Kennesaw st. & 318 Cornell . . . especially if they are all wins