Do any of these models publish their assumptions and limitations?

Sign up, and you'll be able to ignore users whose posts you don't want to see. Sign up

Report problem with this ad

Jan 3, 2025
2:47:03pm

kcca All-American

Do any of these models publish their assumptions and limitations?

Report problem with this ad

kcca

Bio page: kcca
Joined: Aug 3, 2015
Last login: Jan 5, 2025
Total posts: 3,626 (9 FO)

Report problem with this ad

Messages

Author

Time

4 loss Bama's ranking, per the analytics

Indy Coug Loquacious Lummox

Jan 3, 2:02pm

Analytic models give way too much credit for losses to good teams. Its a loss. Not all losses are equal, but still an L

Crazy Joe Davola

Jan 3, 2:04pm

Analytics are primarily driven by scoring margins, adjusted for the quality of the opponent

Indy Coug

Jan 3, 2:05pm

Sounds like you prefer resume metrics, such as SOR, SP+ resume, or K Ford Most Deserving

TCuz

Jan 3, 2:10pm

Losses just need to be weighted heavily against a team. A 4 loss Bama team is not a top 10 team (top 5 in some of those)

Crazy Joe Davola

Jan 3, 2:15pm

The purpose of FPI is as a predictive tool. Maybe that’s the disconnect

TCuz

Jan 3, 2:20pm

There should be a difference, then, between ranking a team and predicting

Carnac

Jan 3, 2:26pm

But nobody is voting Alabama as a top 5 team, so I’m confused about what we are even mad about

TCuz

Jan 3, 2:31pm

Just realized they played Michigan in their final game 2 years in a row.

JAGA97

Jan 3, 2:04pm

That dang eye test

GFCougar

Jan 3, 2:05pm

This is the exact opposite of the eye test.

TCuz

Jan 3, 2:06pm

Right. This shows that the eye test does have some merit as opposed to the

GFCougar

Jan 3, 2:41pm

Hey, good losses matter!

TrueBlue67

Jan 3, 2:09pm

How does that make sense with losses and getting blown out in some games?

Tuskaloosa Cougar

Jan 3, 2:11pm

3 of the 4 losses were by a combined 18 points so only one was a blowout

Indy Coug

Jan 3, 2:14pm

Garbage data in... Garbage data out. The core metrics are bad.

BYUtka

Jan 3, 2:14pm

If the garbage beats the Vegas spread more often than not, how garbage-y is it

Indy Coug

Jan 3, 2:16pm

The SEC is getting filthy rich using it.

BYUtka

Jan 3, 2:41pm

And you could too. It's up to you if you want to be rich or be wrong.

Indy Coug

Jan 3, 2:43pm

They cheat with recruiting rankings. When someone gets an offer from an SEC school their star ranking goes up. These

HarlemCoug

Jan 3, 2:15pm

We should just stop playing games and rank based on measures.

Carnac

Jan 3, 2:12pm

How do you do that when everyone plays a vastly different schedule?

PHXSUNS31

Jan 3, 2:14pm

The measures come from playing the games.

Indy Coug

Jan 3, 2:14pm

Weird then. Four losses. Where do those play into the analytics?

Carnac

Jan 3, 2:16pm

It turns out when you play a weak schedule you don’t lose as much

Odysseus

Jan 3, 2:34pm

Explain Oklahoma.

Carnac

Jan 3, 2:37pm

Teams lose. Do you question all data and metrics when the Dodgers drop a series to the Rockies?

TCuz

Jan 3, 2:39pm

Should the games be played?

Carnac

Jan 3, 2:46pm

Of course. Games matter. That’s why Alabama was excluded despite these metrics.

TCuz

Jan 3, 2:48pm

A talented team in a tough environment played in the middle of an SEC schedule

Odysseus

Jan 3, 2:53pm

The system is so rigged that even when they lose they win. This is data that needs national attention. When you post it

HarlemCoug

Jan 3, 2:12pm

People should focus more on who teams beat than who they lost to

Odysseus

Jan 3, 2:31pm

Why? Both are relevant.

LifeLongLakerFan

Jan 3, 2:42pm

Of course they’re relevant. But we have to make judgments about who is better

Odysseus

Jan 3, 2:54pm

What about teams that you made score a lot of points to beat you? Think 🌹

GFCougar

Jan 3, 2:53pm

Do any of these models publish their assumptions and limitations?

kcca

Jan 3, 2:47pm

We should just do away with all ranking systems that don’t have Bama number 1.

JohnnyComeLately

Jan 3, 3:06pm

What did the analytics say about an Alabama-Michigan matchup?

Doogal McCougar

Jan 3, 3:32pm

RE: What did the analytics say about an Alabama-Michigan matchup?

Indy Coug

Jan 3, 3:42pm

I’m not saying they should be top 5… but come bball season we all say good

Briguy

Jan 3, 4:32pm

something seems very circular about all this

Bob Vance

Jan 3, 4:54pm

yup - and they'll start off next year in a good place too because they had another good recruiting class and because

HarlemCoug

Jan 3, 5:40pm

Report problem with this ad

Posting on CougarBoard

In order to post, you will need to either sign up or log in.

Report problem with this ad