Model Update Heading to Arkansas

[Cats192]

Maybe I should build a model that only counts impressive wins, and ignores home losses to Georgia and getting drubbed by Bama? For instance.

I was trying to have a discussion, but if you're going to smart off over it, I just won't enter the "My Model" threads anymore.

They're dramatically higher in the NET and the Kenpom. They're VERY clearly a tournament team at this point. Yet you don't have them in the top 68. It seems like that doesn't line up. Georgia loss or not.

 

[Aike]

As much as anybody else...

Well, I think we’re still a notch below a few teams, but we have all the pieces.

 

[Aike]

I was trying to have a discussion, but if you're going to smart off over it, I just won't enter the "My Model" threads anymore.

They're dramatically higher in the NET and the Kenpom. They're VERY clearly a tournament team at this point. Yet you don't have them in the top 68. It seems like that doesn't line up. Georgia loss or not.

I was also trying to have a discussion. Indicating that there are reasons they don’t perform as well in my modeling. One is the ugly way they play. Another is that they also have played poorly at times.

But sure, feel free to ignore if it bothers you that my model doesn’t match Kenpom or Net for every team. Honestly, all the models out there should always come up with the same answers.

 

[Aike]

60 years of watching SEC basketball played in cow barns with SEC officials has shown me that those teams will win against other SEC teams in those cow barns with SEC officials, {or just anyone they found on the street and put a stiped shirt on), and then fizzle out as soon as they go play in a real basketball arena with real officials.

Everyone knows that the SEC doesn't care one iota about basketball and that includes to SEC offices. If they wanted it different they would change it.

We have had some hiccups recently but traditionally once UK gets away from SEC teams and SEC officials they thrive. It's like watching completely different games.

Most of the SEC just gears everything to winning those games in their barns. UK simply has different goals than the rest of the SEC. And they hate us for it.

Lot of truth to this.

 

[DreadLox]

Vegas runs on predictive models.

Predictive model + house %.

 

[Cats192]

I was also trying to have a discussion. Indicating that there are reasons they don’t perform as well in my modeling. One is the ugly way they play. Another is that they also have played poorly at times.

But sure, feel free to ignore if it bothers you that my model doesn’t match Kenpom or Net for every team. Honestly, all the models out there should always come up with the same answers.

Again, you're being sarcastic. I feel like you're trying to avoid an honest discussion.

I didn't say "All models should be the same." But 75 vs 39 and 45 is quite the outlier. RPI has them at 13. ESPN's BPI has them the lowest at 58. You're a whole 17 spots worse than that.

So forgive me for offering honest feedback that you seem to be the lowest of any model on that team. And perhaps that should raise some questions.

Anywhose: You're not interested in that discussion. You seem to think the model is great. You seem to be implying that it's ridiculous to question it. So I won't discuss it again.

 

[Montana81]

Again, you're being sarcastic. I feel like you're trying to avoid an honest discussion.

I didn't say "All models should be the same." But 75 vs 39 and 45 is quite the outlier. RPI has them at 13. ESPN's BPI has them the lowest at 58. You're a whole 17 spots worse than that.

So forgive me for offering honest feedback that you seem to be the lowest of any model on that team. And perhaps that should raise some questions.

Anywhose: You're not interested in that discussion. You seem to think the model is great. You seem to be implying that it's ridiculous to question it. So I won't discuss it again.

You’re ignoring what he’s saying or just not understanding.

There are teams that play in a way that yield more success in the ncaa tourney format than they do in the regular season format. It’s absolutely possible and makes sense for a model attempting to track what’s different about those teams to vary, potentially widly, from models attempting to track overall performance.

And one thing his model DOES agree on to a degree with South Carolina and the other computer rankings, is that they might not be as good as their record indicates.

They’re now 18-3, which is more or less tied for the 8th best record in college basketball as there are only 7 teams with less than 2 losses. None of the rankings have them anywhere close to the top 10.

 

[Cats192]

You’re ignoring what he’s saying or just not understanding.

There are teams that play in a way that yield more success in the ncaa tourney format than they do in the regular season format. It’s absolutely possible and makes sense for a model attempting to track what’s different about those teams to vary, potentially widly, from models attempting to track overall performance.

And one thing his model DOES agree on to a degree with South Carolina and the other computer rankings, is that they might not be as good as their record indicates.

They’re now 18-3, which is more or less tied for the 8th best record in college basketball as there are only 7 teams with less than 2 losses. None of the rankings have them anywhere close to the top 10.

Well there’s not much to understand, since both replies were nothing but sarcasm.

South Carolina is not the 75th best team in the country. And while I appreciate the sentiment that “we can look at certain factors to predict things,” only 1 model has them that low.

I have eyes. I watch a lot of college hoops. SC isn’t a top 10 team. They’re certainly not the 75th best team. I think the models that have them in the 30’s or 40’s are much closer to their actual ranking.

 

[Montana81]

Well there’s not much to understand, since both replies were nothing but sarcasm.

South Carolina is not the 75th best team in the country. And while I appreciate the sentiment that “we can look at certain factors to predict things,” only 1 model has them that low.

I have eyes. I watch a lot of college hoops. SC isn’t a top 10 team. They’re certainly not the 75th best team. I think the models that have them in the 30’s or 40’s are much closer to their actual ranking.

I agree they’re not the 75th best team. But that’s not what his model is ranking.

 

[Aike]

Again, you're being sarcastic. I feel like you're trying to avoid an honest discussion.

I didn't say "All models should be the same." But 75 vs 39 and 45 is quite the outlier. RPI has them at 13. ESPN's BPI has them the lowest at 58. You're a whole 17 spots worse than that.

So forgive me for offering honest feedback that you seem to be the lowest of any model on that team. And perhaps that should raise some questions.

Anywhose: You're not interested in that discussion. You seem to think the model is great. You seem to be implying that it's ridiculous to question it. So I won't discuss it again.

👋

 

[kywildcats4life]

I was trying to have a discussion, but if you're going to smart off over it, I just won't enter the "My Model" threads anymore.

They're dramatically higher in the NET and the Kenpom. They're VERY clearly a tournament team at this point. Yet you don't have them in the top 68. It seems like that doesn't line up. Georgia loss or not.

He doesn't take disagreement well. Feelings get hurt. 😭

 

[Aike]

He doesn't take disagreement well. Feelings get hurt. 😭

🤡

 

[kywildcats4life]

🤡

Definitely you.

 

[Aike]

Been a while since I ignored someone but sometimes you just have to.

 

[kywildcats4life]

Is FL in your top 100?