Vegas odds to win the superbowl

[CharlotteBeer.com]

Da Bears have the easiest schedule on paper, a consistently strong defense and running game, even a recent SB appearance. Now, for the first time in over a decade, they have a proven and charismatic QB that they believe in. It pains me to say it, but in 2009, I feel they honestly have a better shot than we do.

They do indeed have a good shot, but I wouldn't rule out Green Bay, either. They're making some nice moves on defense and I think Aaron Rodgers will continue to improve. If I had to guess today, I'd say the Bears win the division with Green Bay nipping at their heels.

Minnesota has been the trendy pick since Peterson's arrival into the league. Let's see if Rosenfels or Jackson can get it done.

[Happy Panther]

Most of the initial lines are based on realist expectations, however, the "fan favorite" teams like Dallas are always going to be skewed. They know that people will still line up around the corner to bet on Dallas, so they push the odds as slim as they can get away with. Those guys aren't stupid, they know Dallas doesn't have the 2nd best chance in football to win the super bowl.

I really don't think the initial guys are skewed but you could be right.

[R0CKnR0LLA]

I disagree on the Bears for many reasons that I don't feel like going into Either way that's only 1-2 teams, I think we are just about where we should be.

One thing for anyone whose going to bet on these, consider taking the team to win their conference championship instead. You have to figure, you have no idea who the other team that would be in the super bowl, or how well they'd be playing. Maybe an oversimplification but you can say there's a 50/50 shot of winning once you get there. Generally, a 20-1 to win the super bowl, would be 10-1 just to get there. So why take an already longshot 10-1 and DOUBLE the odds of hitting it. It's like taking a 10-1 shot, winning, and then saying, OK double or nothing on a coinflip.

One thing you can do too, is to take half the money, put it on them winning their championship game, and the other half going all the way winning the super bowl.

[Happy Panther]

They move significantly many years. What do you think happened to the odds last year when Brady went down? The Panther odds went from 50-1 to less than 30-1 after the first 3 or 4 weeks. Is that not significant?

You don't need a ton of modeling to do what they did. Other than Dallas the rest of the top are teams that have make the playoffs and had good records last year or won their division. It doesn''t even get iffy until you get halfway down the list. If there are any surprises it is who fell to the bottom like Arizona and Minnesota. The reality is that no matter what they set the odds at, you are going to be right 31 out of 32 times. And if you do a good job of making sure the betting is spread on all teams you wouldn't lose money until a team wins with odds greater than 30-1. So as long as the bottom 10 teams don't win it, you are safe.

But for the record, I have heard the same thing about the modeling, I am just not convinced that it really is very meaningful or valuable.

That's different. I am talking about lines moving due to betting. I am also talking about before the season. Once games start happening the models are re-run.

I agree that you or I could have put together a list that would have reasonably predicted the odds above at least in order of favorites. The modeling firms are concerened about the minutiae. I can usually guess that the line on a panthers game will be like 4 or could be 7.5. But if the company misses a line by 3 points this is a relative failure 9barring unforeseen happenings)

I do predictive modeling for a living and the whole point is you make the best estimate with the data in front of you. It is not meant to be a predictive of the final answer when all the data is known. People who don't understand that are overly critical of modeling arguing that the modles are always wrong

Oh well

[panthers55]

That's different. I am talking about lines moving due to betting. I am also talking about before the season. Once games start happening the models are re-run.

I agree that you or I could have put together a list that would have reasonably predicted the odds above at least in order of favorites. The modeling firms are concerened about the minutiae. I can usually guess that the line on a panthers game will be like 4 or could be 7.5. But if the company misses a line by 3 points this is a relative failure 9barring unforeseen happenings)

I do predictive modeling for a living and the whole point is you make the best estimate with the data in front of you. It is not meant to be a predictive of the final answer when all the data is known. People who don't understand that are overly critical of modeling arguing that the modles are always wrong

Oh well

Predictive modeling is fine and can be very useful. Predicting games for example typically a week or so ahead of the time with most of the relevant facts in front of you is much different that predicting an outcome 6 months from now with little of the data in front of you.

Kind of like predicting what the weather will be next Monday versus predicting what the temperature will be on February 15, 2010. You can come up with all the models you want but something which rides on a number of factors 6 months from now has much less validity than predicting something a few days from now based on data which is relevant and timely.

So lines on football games done a week ahead have little in common with a prediction made 6 months from now before free agency is even completed and training camp begun.

So my issue is that predictive models regarding the Super Bowl winner are not very valid no matter who does them by the nature of what you are trying to predict. Doesn't mean they all suck or can't be useful but the further out you go and the less data you have, the less accurate the prediction. You made it sound that with all the modeling you can reasonably predict the outcome or use these numbers to determine who has the best team. I simply said it looks like a fan with some knowledge of how teams finished last year, who had gone to the Superbowl, and what they had done in the offseason could do as good a job with a lot less effort or work. Doesn't take away from what you do for a living. Is this your list by the way??

[double_blue]

18/1... not bad. i would wonder what detroit and kansas city's odds will be when the season starts...

[JJT]

Most of the initial lines are based on realist expectations, however, the "fan favorite" teams like Dallas are always going to be skewed. They know that people will still line up around the corner to bet on Dallas, so they push the odds as slim as they can get away with. Those guys aren't stupid, they know Dallas doesn't have the 2nd best chance in football to win the super bowl.

Can you lay the odds and bet against Dallas?

If so, that's a good play.

[R0CKnR0LLA]

Can you lay the odds and bet against Dallas?

If so, that's a good play.

I don't know if you can bet against them winning the super bowl. I can tell you this, last year I bet against them almost every week for individual games. Even when they win, they rarely beat the ridiculous spread they get. I think there were 4-5 weeks in a row at one point where I bet against them and won every time.

[Happy Panther]

Predictive modeling is fine and can be very useful. Predicting games for example typically a week or so ahead of the time with most of the relevant facts in front of you is much different that predicting an outcome 6 months from now with little of the data in front of you.

This is correct. the variance around the estimates is massive making the estimates practically useless as far as comparing the actual result to predicted. And you should know that going in.

Kind of like predicting what the weather will be next Monday versus predicting what the the temperature will be on February 15, 2010. You can come up with all the models you want but something which rides on a number of factors 6 months from now has much less validity than predicting something a few days from now based on data which is relevant and timely.

I think this is a specious analogy. A better would be what are the odds of it snowing that day. There is plenty of data to make a valid prediction. (The answer is 12% chance btw

So lines on football games done a week ahead have little in common with a prediction made 6 months from now before free agency is even completed and training camp begun.

They both use all the data in front of them one has a much higher variance around the mean - although the variance isa massive on both

So my issue is that predictive models regarding the Super Bowl winner are not very valid no matter who does them by the nature of what you are trying to predict.

This is true of any prediction. Try predicting which of the playoff teams will win the superbowl before the playoffs. In fact the superbowl line is considered the hardest line to set all year despite having 20+ weeks of data

Doesn't mean they all suck or can't be useful but the further out you go and the less data you have, the less accurate the prediction.

obviously

You made it sound that with all the modeling you can reasonably predict the outcome

absolutely not

or use these numbers to determine who has the best team.

Absolutely yes. Before the season I would say the Cowboys have the second best team - based on oodles of data plugged into some model. This is about the best you can do. Unless you have a better model or want to argue that the data is flawed. i think you are saying it can't be used to predict who will have the best team after the season which is true.

I simply said it looks like a fan with some knowledge of how teams finished last year, who had gone to the Superbowl, and what they had done in the offseason could do as good a job with a lot less effort or work.

Absolutely disagree. But we can just be different here

Doesn't take away from what you do for a living. Is this your list by the way??

I just pulled it from some website. I can't check bodog from work.

[Happy Panther]

Arizona Cardinals

23.00

Atlanta Falcons

23.00

Baltimore Ravens

21.00

Buffalo Bills

51.00

Carolina Panthers

21.00

Chicago Bears

21.00

Cincinnati Bengals

91.00

Cleveland Browns

71.00

Dallas Cowboys

15.00

Denver Broncos

46.00

Detroit Lions

151.00

Green Bay Packers

26.00

Houston Texans

36.00

Indianapolis Colts

13.00

Jacksonville Jaguars

31.00

Kansas City Chiefs

66.00

Miami Dolphins

46.00

Minnesota Vikings

19.00

New England Patriots

6.00

New Orleans Saints

21.00

New York Giants

11.00

New York Jets

36.00

Oakland Raiders

101.00

Philadelphia Eagles

17.00

Pittsburgh Steelers

10.00

San Diego Chargers

15.00

San Francisco 49ers

51.00

Seattle Seahawks

46.00

St.Louis Rams

101.00

Tampa Bay Buccaneers

41.00

Tennessee Titans

21.00

Washington Redskins

26.00

[panthers55]

Absolutely yes. Before the season I would say the Cowboys have the second best team - based on oodles of data plugged into some model. This is about the best you can do. Unless you have a better model or want to argue that the data is flawed. i think you are saying it can't be used to predict who will have the best team after the season which is true.

Best at what? The second best team from a talent point of view? Second best team from a win/loss record? Second best team playoff history? Second most likely to go deep in the playoffs?

And exactly what does their model tell you about how a team will gel and play together or whether they will play better when the playoffs begin. Or even that past performance will have any bearing on the future. That is the problem with all these models. The sum of the parts don't alway add up to the best team or performance. You can overlay all the stats and information and as long as you are describing what happened you are fine. As soon as you start predicting what will happen you are on a slippery slope.

There is a long line of teams that typically underacheive in the last decade despite having all the talent in the world. Dallas is surely one of those.

How long have people been saying Arizona is a playoff contender? How many of them even thought the Cardinals would be in playoffs?

These models are a little better than guessing but for all the dollars spent and models made, the accuracy that they will predict the winner is very low. The best thing it does is keep a lot of people working.

And lets be clear- the Vegas odds are expressly setting the odds for each team winning the Super Bowl, not who looks best on paper or who is best at this point from a talent point of view. And by your own admission predicting the winner at this point is virtually impossible. That is exactly what I said several threads ago which you keep arguing with.

So which is it? Are the odds of predicting who will win the Superbowl doing just that, or are they saying who looks best on paper right now? In either case they appear very invalid.

[Happy Panther]

Best at what? The second best team from a talent point of view? Second best team from a win/loss record? Second best team playoff history? Second most likely to go deep in the playoffs?

And exactly what does their model tell you about how a team will gel and play together or whether they will play better when the playoffs begin. Or even that past performance will have any bearing on the future. That is the problem with all these models. The sum of the parts don't alway add up to the best team or performance. You can overlay all the stats and information and as long as you are describing what happened you are fine. As soon as you start predicting what will happen you are on a slippery slope.

There is a long line of teams that typically underacheive in the last decade despite having all the talent in the world. Dallas is surely one of those.

How long have people been saying Arizona is a playoff contender? How many of them even thought the Cardinals would be in playoffs?

These models are a little better than guessing but for all the dollars spent and models made, the accuracy that they will predict the winner is very low. The best thing it does is keep a lot of people working.

You are saying all the things I already said. We agree. :hurray:

[R0CKnR0LLA]

Saying the Cowboys have the 2nd best team in the NFL makes baby jesus cry.

[R0CKnR0LLA]

I would say the Cowboys have the second best team

(referring to Carolina) I don't think they can beat top teams

(referring to Carolina) there is no way they do well against Patriots, Giants, Miami, Philly, Minnesota. And if they don't then they need to run through the division which they will not.

I think I speak for everyone, when I say, go become a Cowboys fan. You will fit right in I'm sure.