*Update - NCAA Tourney Predictive Model 01/30/22*

[Aike]

Here are some highlights…

New number 1 by the slightest of margins…

Cats jump a spot after a mixed week (struggle with MSU at home, pummel KU at the Phog)…

UCLA makes a big leap…


Aike’s Tourney Predictor



Hope you enjoy. Let me know what questions you have.

 

[Lempface]

So UT just lost and we skull dragged em and KU at their place and they are ahead? Back to the drawing board methinks.

 

[Aike]

So UT just lost and we skull dragged em and KU at their place and they are ahead? Back to the drawing board methinks.

There’s been a whole season played. UT remained basically level after beating a Florida team that my model likes, and narrowly losing at Texas (another team my model likes).

Besides smoking UT and KU, we also lost at Auburn and barely won at home vs. MSU. Yes, we had injury issues, but those aren’t taken into account.

Still, we improved by score and position since last week. All positive for the Cats.

 

[Lempface]

There’s been a whole season played. UT remained basically level after beating a Florida team that my model likes, and narrowly losing at Texas (another team my model likes).

Besides smoking UT and KU, we also lost at Auburn and barely won at home vs. MSU. Yes, we had injury issues, but those aren’t taken into account.

Still, we improved by score and position since last week. All positive for the Cats.

UT is 14-6, they’ve played fewer games, have two more losses, 3 fewer wins and no game close to as quality of a win as a top 5 on the road. You can’t convince me a system that ranks them that high, not just ahead of us, has any merit. They beat Arizona at home which is a good win but won by 4 compared to 18 at an away game….

 

[Odd Janitor]

If you’d told me 2 weeks ago we had a shot at a 1 seed I would’ve called you a blue glasses optimist. Now anything less than a 2 seed would surprise me.

 

[Aike]

If you’d told me 2 weeks ago we had a shot at a 1 seed I would’ve called you a blue glasses optimist. Now anything less than a 2 seed would surprise me.

I think we control our own destiny. Schedule is TOUGH, but if we conquer that schedule I don’t see 4 teams ahead of us by year end.

 

[Aike]

UT is 14-6, they’ve played fewer games, have two more losses, 3 fewer wins and no game close to as quality of a win as a top 5 on the road. You can’t convince me a system that ranks them that high, not just ahead of us, has any merit. They beat Arizona at home which is a good win but won by 4 compared to 18 at an away game….

Yeah, when you try to pick and choose a game here and there like you’re doing, you’re never going to be satisfied.

UT has played a top 10 schedule, and played pretty well against that schedule for the most part.

Are they the 5th best team in the country or 25th? I have no clue. Probably different on different nights, like every other team.

This model says that teams that play like UT plays typically do pretty well in the tournament. I wouldn’t get in a twist about it.

 

[TortElvisII]

Yeah, when you try to pick and choose a game here and there like you’re doing, you’re never going to be satisfied.

UT has played a top 10 schedule, and played pretty well against that schedule for the most part.

Are they the 5th best team in the country or 25th? I have no clue. Probably different on different nights, like every other team.

This model says that teams that play like UT plays typically do pretty well in the tournament. I wouldn’t get in a twist about it.

Just put UK at 6.0 and they will leave you alone.

 

[Lempface]

Just put UK at 6.0 and they will leave you alone.

I said it wasn’t about being ahead of us just being that high. 15-25 more like it.

 

[Aike]

Just put UK at 6.0 and they will leave you alone.

That’s funny, but I actually crunched the numbers for that one game (UT game) and our score was 5.93. That game against UT was more about what we did right than what they did wrong, in my opinion.

They turned it over too much, but we just shot the ball at a ridiculous clip. Nobody was going to beat us that day.

 

[TortElvisII]

I said it wasn’t about being ahead of us just being that high. 15-25 more like it.

UK last night was the best team in the country. UK for the first 8 minutes against Auburn was the best team in the country. UK versus Tennessee was the best team in the country. UK versus North Carolina was the best team in the country. Remember this model does not take into account injuries. Uninjured I would say we have at minimum a top 3 team.

 

[Aike]

UK last night was the best team in the country. UK for the first 8 minutes against Auburn was the best team in the country. UK versus Tennessee was the best team in the country. UK versus North Carolina was the best team in the country. Remember this model does not take into account injuries. Uninjured I would say we have at minimum a top 3 team.

I think you’re right. For sure. We are going to be a tough out if we can keep our key people on the floor.

 

[cat888]

Staying down in the rating to around 10 helps to keep the pressure off going into the tournament.

 

[mash_24]

Unfortunately having Tennessee where you do kinda invalidates your entire model. You have Iowa who is 4-5 in their conference at #10. You also have a 10-10 Oklahoma State above Michigan State. Might be time for a tweak.

 

[Jonathan40330]

Wisconsin is 50th? ********. I'm not even a fan, but that's ridiculous.

 

[mhroe1984]

I think we control our own destiny. Schedule is TOUGH, but if we conquer that schedule I don’t see 4 teams ahead of us by year end.

Imagine winning out from here to selection Sunday. We'd be 30-4 with an SEC Tournament title. No way we wouldn't be a 1 seed with that record.

 

[Lempface]

Wisconsin is 50th? ********. I'm not even a fan, but that's ridiculous.

16-3 11th ranked Wisconsin behind 11-8 Memphis. 16 spots behind Memphis.

 

[Aike]

Unfortunately having Tennessee where you do kinda invalidates your entire model. You have Iowa who is 4-5 in their conference at #10. You also have a 10-10 Oklahoma State above Michigan State. Might be time for a tweak.

That isn’t really how math works. The model predicts that Tennessee would win X number of games in the tournament because teams with similar profiles performed to a certain level in previous tournaments.

Will the model be right? Certainly not. No model ever is. But this model has performed well in the past.

It’s a tool. Use it along with Kenpom, Sagarin, your own eye test…or don’t use it.

OK ST. beat Texas and won at Baylor. They’ve played a brutal schedule. These are some of the reasons the model likes them.

That being said, they likely don’t make the tournament due to record. At that point, it becomes moot. You won’t have to worry about picking them to win or lose if they aren’t playing.

The model itself doesn’t factor in won/loss record, but there’s no need. The selection committee takes care of that element (an element that’s much less predictive than many others).

 

[Aike]

Imagine winning out from here to selection Sunday. We'd be 30-4 with an SEC Tournament title. No way we wouldn't be a 1 seed with that record.

That would be a phenomenal result considering our schedule.

 

[Aike]

Wisconsin is 50th? ********. I'm not even a fan, but that's ridiculous.

They’ve clearly had better than the 50th best season. They have a good resume. It just happens that the way that they play doesn’t typically produce strong tournament results.

Keep winning, and resume should get them a good seed. Then maybe they win a game or two anyway. But teams like Wisconsin are a reason this model tends to do well in picking the tournament. If there were no interesting results, we would all just pick based on the AP Poll.

 

[fisherscat]

If you’d told me 2 weeks ago we had a shot at a 1 seed I would’ve called you a blue glasses optimist. Now anything less than a 2 seed would surprise me.

A 2 would be huge. Think I’d be more concerned about playing the potential other 2 seeds than the 1 seeds.

 

[Aike]

A 2 would be huge. Think I’d be more concerned about playing the potential other 2 seeds than the 1 seeds.

The one line feels a little soft to me. Gonzaga is good but we’ve seen that they’re beatable. We had Auburn down at their place before TyTy went out, and it’s hard to forget that.

Baylor might be really good if healthy, but they aren’t. Houston has dealt with injuries. Arizona’s schedule isn’t great. It isn’t hard to see us moving up to a one with continued wins.

 

[UKUGA]

They’ve clearly had better than the 50th best season. They have a good resume. It just happens that the way that they play doesn’t typically produce strong tournament results.

Keep winning, and resume should get them a good seed. Then maybe they win a game or two anyway. But teams like Wisconsin are a reason this model tends to do well in picking the tournament. If there were no interesting results, we would all just pick based on the AP Poll.

Kind of reminds me of USC early . . .

 

[UKUGA]

Aike, what you clearly don’t understand, is that you need to tweak your model every week so that it satisfies everyone.

That’s how math works, right?

 

[Aike]

Aike, what you clearly don’t understand, is that you need to tweak your model every week so that it satisfies everyone.

That’s how math works, right?

Not a bad idea, but I might need to start a go fund me. Donate $500 and tell me which team you want in each spot.

 

[Montana81]

I see memphis has plummeted.

I expect tennessee to start falling soon as well. They’ve played one of the tougher schedules in the country to be fair to them but i don’t see them making the s16 or better which is what the model is suggesting. Rick Barnes historically is not a very good march coach either.

Looking at their numbers…i’m guessing your model likes teams that get a lot of steals. That’s their best stat. They’re 10th in the country in that category.

 

[Lempface]

Aike, what you clearly don’t understand, is that you need to tweak your model every week so that it satisfies everyone.

That’s how math works, right?

This is insane. So just make a model that spits out a random order and tell people who don't like it tough? A model needs to be somewhat accurate to be of value and when people point out the issues with a model it isn't plainly invalid.

 

[UKUGA]

This is insane. So just make a model that spits out a random order and tell people who don't like it tough? A model needs to be somewhat accurate to be of value and when people point out the issues with a model it isn't plainly invalid.

“somewhat accurate”

lempface sez its wrong, so you a dumdum @Aike

 

[Aike]

I see memphis has plummeted.

I expect tennessee to start falling soon as well. They’ve played one of the tougher schedules in the country to be fair to them but i don’t see them making the s16 or better which is what the model is suggesting. Rick Barnes historically is not a very good march coach either.

Looking at their numbers…i’m guessing your model likes teams that get a lot of steals. That’s their best stat. They’re 10th in the country in that category.

Teams that are generally disruptive tend to do well in my model. But they also need to be successful on the other end of the floor.

 

[Lempface]

“somewhat accurate”

lempface sez its wrong, so you a dumdum @Aike

Did you look at it? Lmfao.

 

[UKUGA]

Did you look at it? Lmfao.

No, I just read your posts.

 

[Aike]

This is insane. So just make a model that spits out a random order and tell people who don't like it tough? A model needs to be somewhat accurate to be of value and when people point out the issues with a model it isn't plainly invalid.

Has the season ended? Has the tournament been doing played?

Kind of hard to judge a model that predicts tournament success before the tournament is played.

Been running this for a few years now and it’s done a really good job. Might do well this year, or it might not. Every season is different.

But I mean, there’s nothing insane or random about this. It’s a math problem. Can’t for the life of me understand why anybody gets triggered by the answer it spits out.

 

[UKUGA]

I give @Aike a ton of credit for continuing to post his model in this cesspool of morons.

 

[UKUGA]

Has the season ended? Has the tournament been doing played?

Kind of hard to judge a model that predicts tournament success before the tournament is played.

Been running this for a few years now and it’s done a really good job. Might do well this year, or it might not. Every season is different.

But I mean, there’s nothing insane or random about this. It’s a math problem. Can’t for the life of me understand why anybody gets triggered by the answer it spits out.

People get triggered by things that make them feel inadequate.

 

[Aike]

I give @Aike a ton of credit for continuing to post his model in this cesspool of morons.

I would never say that, but I thanks for the credit.

 

[mash_24]

Amazing how defensive computer guys get about their “models”. Just keep moving the teams up that win in the tournament like Kenpom does, then you’ll always be right. Except that’s not really predictive is it?

 

[Lempface]

People get triggered by things that make them feel inadequate.

Triggered? It's a discussion board, I'm discussing how inaccurate this model is currently ranking teams.

 

[Aike]

Triggered? It's a discussion board, I'm discussing how inaccurate this model is currently ranking teams.

Since it isn’t designed to rank the teams, I guess I won’t worry too much about that.

 

[UKUGA]

Triggered? It's a discussion board, I'm discussing how inaccurate this model is currently ranking teams.

How do you know it’s inaccurate? Have you seen the results of the NCAA tournament?

 

[Lempface]

Since it isn’t designed to rank the teams, I guess I won’t worry too much about that.

So those numbers by the names?