Friend of Boneyard Doug Gottlieb says KenPom is BS...

[JCisthegoat]

So you guys believe that UConn is now worse than those teams. And even the addition of Reed and Mullins to the lineup would be insufficient to make up the difference with Arizona?

Based on UConn's loss of zero games since any of those games?

Ok.

Edit to add: All because UConn doesn't look pretty when opponents hold them motionless on offense in whatever sport the Big East play under the guise of it being basketball?

Worse than what teams exactly?

 

[JCisthegoat]

It is useful for betting. Not sure how effective it is for ranking teams or tournament selection, and Ken Pimeroy would probably agree with that assessment.

But it’s useful for betting because it’s possible to have predictive models that accurately assess a game.

 

[willie99]

I hate to break this to you, but just be cause someone calls something an opinion doesn’t s..top it from being stupid.

Yes, your strawman was pure brilliance, not sure how I missed it

 

[7774]

Well, I’ve noticed it with us over the past few years And even Hurley I think is on record talking about how winning isn’t enough - it’s the total # of points. That was interesting comment.
But you are right in the sense even ChatGPT threw its hands up with no date to prove/dissprove. This may forever stay in the “think” category until ChatGPT gets a bit smarter

You ever think it’s weird that you have to consult Chat GPT to try to find an example of what you are saying is true, while I can find an example of what I’m saying is true just by using a game I watched this week? You’re saying this is happening all the time, it should be easy to come up with 1 example.

 

[UCan3]

You ever think it’s weird that you have to consult Chat GPT to try to find an example of what you are saying is true, while I can find an example of what I’m saying is true just by using a game I watched this week? You’re saying this is happening all the time, it should be easy to come up with 1 example.

Time to switch the decaf
You pick one game out of thousands and say look it didn't happen here. So you want me to go and find a game and say look this team ran up the score? it happens all the time! Xavier the other night we were up by 30 and Braylon was still out there almost to very end of game.

But Where does that get us?


I was interested if there was any facts and figures out there because you did pose a good question.

Apparently, the data is not easy to find to support your position or mine.

 

[7774]

Time to switch the decaf
You pick one game out of thousands and say look it didn't happen here. So you want me to go and find a game and say look this team ran up the score? it happens all the time! Xavier the other night we were up by 30 and Braylon was still out there almost to very end of game.

But Where does that get us?


I was interested if there was any facts and figures out there because you did pose a good question.

Apparently, the data is not easy to find to support your position or mine.

Yes of course I’m looking for specific examples of something you claim happens all the time. Because I never really see it happen. Not sure why that’s odd to you. We didn’t run up the score on Xavier. We were ahead by 33 with 15 minutes left and won by 32. So we played them even the last 15 minutes. Furphy played 4 minutes. Millender, Koroma, and Paunovic all got in the game. Playing well and winning by a lot over a team that doesn’t play well isn’t running up the score.

I also just looked and Braylon came out of the game with over 5 minutes left so you’re not even correct about that.

 

[nelsonmuntz]

It's very useful for rankings based on how good teams will be moving forward, if that is what you're are trying to measure (but that is only a piece of tournament selection).

It is useful for measuring which coaches and conferences are managing games to the metrics by running up the score.

 

[7774]

It is useful for measuring which coaches and conferences are managing games to the metrics by running up the score.

And yet you aren’t able to provide an actual game example where that happened.

 

[nelsonmuntz]

And yet you aren’t able to provide an actual game example where that happened.

This year's SEC has the 5th highest KenPom rating OF ALL TIME. Does a 2026 SEC which has a losing record against the other majors on the actual basketball court really feel like the 5th best conference of all time?

 

[7774]

This year's SEC has the 5th highest KenPom rating OF ALL TIME. Does a 2026 SEC which has a losing record against the other majors on the actual basketball court really feel like the 5th best conference of all time?

Wow with something that egregious it seems like it would be very easy to find some specific game examples of teams gaming the system, yet you’re never able to do that.

And it’s been explained to you multiple times that you can’t look across years, so I’m certainly not going to explain it again since you don’t care.

 

[nelsonmuntz]

Wow with something that egregious it seems like it would be very easy to find some specific game examples of teams gaming the system, yet you’re never able to do that.

And it’s been explained to you multiple times that you can’t look across years, so I’m certainly not going to explain it again since you don’t care.

7774: KenPom is a very accurate analytical tool and I will attack anyone who disagrees with me on this.

Also 7774: Ignore obvious factual inconsistencies that show that KenPom's, or any efficiency ratings, actual usefulness is very limited.

 

[7774]

7774: KenPom is a very accurate analytical tool and I will attack anyone who disagrees with me on this.

Also 7774: Ignore obvious factual inconsistencies that show that KenPom's, or any efficiency ratings, actual usefulness is very limited.

nelsonmuntz: someone asking me for an example is the same as attacking me.

 

[nelsonmuntz]

nelsonmuntz: someone asking me for an example is attacking me.

I gave you THE example of how KenPom's model has all kinds of problems, and explained why any model that is so easily gamed by those it claims to measure has serious limitations.

 

[7774]

I gave you THE example of how KenPom's model has all kinds of problems, and explained why any model that is so easily gamed by those it claims to measure has serious limitations.

How is the model being gamed and what are some examples of specific games where this was happening?