KSUFans Archives

Sports => Frank Martin's OOD sponsored by the "Angriest Fans in America" => Topic started by: QuinnMac on January 21, 2010, 04:31:36 PM

Title: I don't understand KenPom
Post by: QuinnMac on January 21, 2010, 04:31:36 PM
Why is Wisconsin @ 6 (15-4)
and
K-State is @ # 8 (16-2)

http://kenpom.com/rate.php (http://kenpom.com/rate.php)

 :confused:
Title: Re: I don't understand KenPom
Post by: mocat on January 21, 2010, 04:34:06 PM
Why is Wisconsin @ 6 (15-4)
and
K-State is @ # 8 (16-2)

http://kenpom.com/rate.php (http://kenpom.com/rate.php)

 :confused:


kenpom lists PROJECTED finishes and RPI rankings, not current ones
Title: Re: I don't understand KenPom
Post by: Pett on January 21, 2010, 04:34:48 PM
Their wins over Duke & Purdue probably put them ahead of us (gave both teams their first loss). But seriously, these clowns have been one upping us since the '08 NCAA tourney. :curse:
Title: Re: I don't understand KenPom
Post by: Skycat on January 21, 2010, 04:38:01 PM
Here you go. (http://kenpom.com/blog/index.php/weblog/ratings_explanation/)
Title: Re: I don't understand KenPom
Post by: QuinnMac on January 21, 2010, 04:55:30 PM
Here you go. (http://kenpom.com/blog/index.php/weblog/ratings_explanation/)
I am not reading all that crap.
3-4 sentences and I will gladly oblige.
Otherwise GTFOOMF, and explain how Wiscky is better than K-State.
Title: Re: I don't understand KenPom
Post by: Skycat on January 21, 2010, 04:58:31 PM
Here you go. (http://kenpom.com/blog/index.php/weblog/ratings_explanation/)
I am not reading all that crap.
3-4 sentences and I will gladly oblige.
Otherwise GTFOOMF, and explain how Wiscky is better than K-State.

mocat is right.  KenPom does predictive rankings, not results based rankings.  He computes offensive and defensive efficiencies for each team, adjusts them based on schedule strength, and then cranks them through a formula.

His numbers indicate that Wisconsin would beat K-State on a neutral floor more often than not.
Title: Re: I don't understand KenPom
Post by: QuinnMac on January 21, 2010, 05:01:23 PM
Here you go. (http://kenpom.com/blog/index.php/weblog/ratings_explanation/)
I am not reading all that crap.
3-4 sentences and I will gladly oblige.
Otherwise GTFOOMF, and explain how Wiscky is better than K-State.

mocat is right.  KenPom does predictive rankings, not results based rankings.  He computes offensive and defensive efficiencies for each team, adjusts them based on schedule strength, and then cranks them through a formula.

His numbers indicate that Wisconsin would beat K-State on a neutral floor more often than not.
Thanks for the explanation, srsly.

But I don't believe his calculations are correct, just imo.

Thanks again. :thumbsup:
Title: Re: I don't understand KenPom
Post by: michigancat on January 21, 2010, 05:12:10 PM
Here you go. (http://kenpom.com/blog/index.php/weblog/ratings_explanation/)
I am not reading all that crap.
3-4 sentences and I will gladly oblige.
Otherwise GTFOOMF, and explain how Wiscky is better than K-State.

They are better than KSU on a points per possession basis when strength of schedule is considered.

They ARE based entirely on results.
Title: Re: I don't understand KenPom
Post by: QuinnMac on January 21, 2010, 05:19:59 PM
Here you go. (http://kenpom.com/blog/index.php/weblog/ratings_explanation/)
I am not reading all that crap.
3-4 sentences and I will gladly oblige.
Otherwise GTFOOMF, and explain how Wiscky is better than K-State.

They are better than KSU on a points per possession basis when strength of schedule is considered.

They ARE based entirely on results.
well, so you are saying KenPom kinda sucks then.
Title: Re: I don't understand KenPom
Post by: Skycat on January 21, 2010, 05:22:03 PM
Here you go. (http://kenpom.com/blog/index.php/weblog/ratings_explanation/)
I am not reading all that crap.
3-4 sentences and I will gladly oblige.
Otherwise GTFOOMF, and explain how Wiscky is better than K-State.

They are better than KSU on a points per possession basis when strength of schedule is considered.

They ARE based entirely on results.

meh, they aren't based on win-loss results, and I'm pretty sure you knew what I meant
Title: Re: I don't understand KenPom
Post by: michigancat on January 21, 2010, 07:26:36 PM
Here you go. (http://kenpom.com/blog/index.php/weblog/ratings_explanation/)
I am not reading all that crap.
3-4 sentences and I will gladly oblige.
Otherwise GTFOOMF, and explain how Wiscky is better than K-State.

They are better than KSU on a points per possession basis when strength of schedule is considered.

They ARE based entirely on results.
well, so you are saying KenPom kinda sucks then.

no.
Title: Re: I don't understand KenPom
Post by: JTKSU on January 21, 2010, 08:12:00 PM
What's the deal with it projecting us to lose 3 more games, yet ending up 24-6 (11-5)?
Title: Re: I don't understand KenPom
Post by: michigancat on January 21, 2010, 08:16:05 PM
What's the deal with it projecting us to lose 3 more games, yet ending up 24-6 (11-5)?

It's a sum of the percentages.
Title: Re: I don't understand KenPom
Post by: JTKSU on January 21, 2010, 08:18:48 PM
What's the deal with it projecting us to lose 3 more games, yet ending up 24-6 (11-5)?

It's a sum of the percentages.

The percentage of that making any sense is <1.
Title: Re: I don't understand KenPom
Post by: Chingon on January 21, 2010, 10:08:18 PM
the final record is based on total probabilities not indv ones.
Title: Re: I don't understand KenPom
Post by: ednksu on January 21, 2010, 10:14:23 PM
he uses result to predict rankings.  The idea is that as the season draws to a close with more and more statistical info his final predicted results will be the same as the real world final results. 
Title: Re: I don't understand KenPom
Post by: Litterbox on January 22, 2010, 02:58:47 AM
So, he's like a meterologist predicting a tornado?  There's what he predicts and then there's what happens. amirite?
Title: Re: I don't understand KenPom
Post by: mcmwcat on January 22, 2010, 01:22:30 PM
if i flip a lopsided coin that has 51% chance to land on heads i will predict it will land on heads. 

i then decide to flip the coin 100 times.  before the 1st flip i'm not going to predict that all 100 flips will result in a heads.  i will predict that 51 times out of 100 that coin will come up heads