actuarygambler

I said winning out was a big IF

So much for winning out :-\

 

 

Hi there, I was at this site a while back, wanted to come back and share this work: http://actuarygambler.blogspot.com/2014/11/college-football-math-destiny.html

 

I built a model based on what we've seen from the CFP committee, how they've ranked teams, how they've talked through their reasoning for ranking teams, and how they've responded to onfield events. So far it's fitting the first 3 weeks fairly well.

 

Based on that model, it looks like IF (that's a big if) Nebraska can win out the rest of their games, they'll have a really good chance to make the CFP. Over 80%. It's a big if, but it starts this weekend with your toughest remaining regular season game @Wisconsin. Then finishes with maybe a B1G title game? Good luck!

Yet, earlier in the week you posted this entry: http://actuarygambler.blogspot.com/p/blog-page_22.html

(as instructed, click on the gray box and type Nebraska)

which says Nebraska's chances of winning out are 22.9%, and their chances of making the CFP is 8.9%. 8.9/22.9 is nowhere close to 85%. That doesn't give me much faith in whatever system or methodology you are using.

 

 

8.9% chance of making the playoffs based on where things currently stand today, and perceived ability to win out.

85% chance of making the playoffs if they actually do win out.

 

 

Hi there, I was at this site a while back, wanted to come back and share this work: http://actuarygambler.blogspot.com/2014/11/college-football-math-destiny.html

 

I built a model based on what we've seen from the CFP committee, how they've ranked teams, how they've talked through their reasoning for ranking teams, and how they've responded to onfield events. So far it's fitting the first 3 weeks fairly well.

 

Based on that model, it looks like IF (that's a big if) Nebraska can win out the rest of their games, they'll have a really good chance to make the CFP. Over 80%. It's a big if, but it starts this weekend with your toughest remaining regular season game @Wisconsin. Then finishes with maybe a B1G title game? Good luck!

Yet, earlier in the week you posted this entry: http://actuarygambler.blogspot.com/p/blog-page_22.html

(as instructed, click on the gray box and type Nebraska)

which says Nebraska's chances of winning out are 22.9%, and their chances of making the CFP is 8.9%. 8.9/22.9 is nowhere close to 85%. That doesn't give me much faith in whatever system or methodology you are using.

 

Thanks, it's a good observation. The 22.9% number you're citing is Nebraska's chance to win every remaining regular season game - it doesn't include the B1G conf championship game, which is definitely something I should be clearer about. But they're around 50/50 to win the b1g title game as a 1 loss team; if you add that into the calculations you're doing it adds up.

Hi there, I was at this site a while back, wanted to come back and share this work: http://actuarygambler.blogspot.com/2014/11/college-football-math-destiny.html

I built a model based on what we've seen from the CFP committee, how they've ranked teams, how they've talked through their reasoning for ranking teams, and how they've responded to onfield events. So far it's fitting the first 3 weeks fairly well.

Based on that model, it looks like IF (that's a big if) Nebraska can win out the rest of their games, they'll have a really good chance to make the CFP. Over 80%. It's a big if, but it starts this weekend with your toughest remaining regular season game @Wisconsin. Then finishes with maybe a B1G title game? Good luck!

Why pick one score when you can pick them all!

thanks

I don't understand why I need to look at a model to determine who will cover a spread. I watch college football every week and have a pretty good idea how good teams are and what factors come into play when playing an opponent.

Because humans have cognitive biases in what they observe, remember, think. Numbers don't. I thikn probably the best modeling system would combine numbers with human observation.

 

I'm well aware of Massey, thaaaaank you very much. But it's a composite of all models, polls, rankings, no matter how outlandish. I only look at the ones that aren't complete nonsense. There's a reason the BCS computers were a joke, and there's a reason why Sagarin is revered for its accuracy with respect to the Vegas lines. Models are like anything else, what it spits out depends on what you put in and what you tell it to do.

 

I noticed that too. LV opening lines are waiting for Sagarin Predictor column. Usually very accurate. However, dismiss and ignore September Sag's ratings/predictors. BTW, Sag's said MSU favored by 7.5 points against our Huskers ...... BEFORE Vegas opening lines.

 

Jeff Sagarin is the best by far in mid-October thru the end (IMO)

 

Sports rating systems are generally of great interest to gamblers. Gamblers use Sagarin's ratings as a source of "Power/Predictor Rankings," traditionally used as a way to determine the spread between two teams. - wiki source

 

I found a huge database of rating systems, vegas lines, and final scores and did a massive study (becuase I wanted to bet based on sagarin if it were that simple!) and found that his lines, while good, weren't better at predicting game outcomes than the vegas lines were.

 

 

 

 

Glad you found my model, happy to answer any questions. Looks like Nebraska has been hovering just outside the top 25, a decisive win this weekend against Sparty would definitely put them over the top.

 

 

 

Welcome to the board.

 

Can you expand on this a little better?

 

Not to sound rude but that is general thinking and kind of obvious.

 

Sure, but it wasn't really a modeling comment, more of just a person comment i.e. look how your rank has been hovering just outside the top 25, you know what I bet would fix that..

Anyway, a decisive win over sparty looks like it would bump nebraska up from 31 to 21. A close win to 26.

MSU 35

NEB 28

Rush 155

Pass 222

 

What is the statistical model you are looking at? Serious question.

 

 

Well, Sagarin has us at 27/28 (predictor is the only one that matters)

Football Outsiders F/+ at 31

Football Outsiders FEI at 29

Football Outsiders S&P+ at 35

This model at 31

 

so on so forth

 

should note that sagarin goes untethered after Week 5, so look to see if there's any marked changes after next week.

 

Glad you found my model, happy to answer any questions. Looks like Nebraska has been hovering just outside the top 25, a decisive win this weekend against Sparty would definitely put them over the top.