PDA

View Full Version : Tentative preseason playoff predictions


cdcox
06-03-2006, 03:16 PM
The software is not updated on nfl-forecast.com yet (probably won't be for another week or two), but I'm providing a preview of the 06 preseason playoff predictions as an exclusive to Chiefs Planet. This is the first time I've ever done preseason predictions. They are based on Vegas money lines on the over/under number of wins which I previeously documented here:

http://www.chiefsplanet.com/BB/showthread.php?t=141461

Screen shots posted below.

cdcox
06-03-2006, 03:17 PM
AFC and NFC playoff odds:

cdcox
06-03-2006, 03:19 PM
Division Standings:

cdcox
06-03-2006, 03:20 PM
Chiefs distribution of wins. Red = make playoffs, grey = not make playoffs.

Baconeater
06-03-2006, 04:43 PM
Chiefs distribution of wins. Red = make playoffs, grey = not make playoffs.
So there is 0 probability that we will win 15 games?

cdcox
06-03-2006, 06:12 PM
So there is 0 probability that we will win 15 games?

Questions like these are very useful to me in evaluating how realistic the my predictions are. Good question.

There have been three 15-1 teams (Bears, Vikings and Steelers) out of about 900 regular seasons in the 30 years since the NFL went to 16 games. Or about a 10% chance of getting one in any given season. When I looked at all the teams together, there was only a 5% chance of getting a 15-1 team in 2006. I noticed the Vegas projected wins tends to group teams around 8-8. This probably accounts for the underestimation of having a 15 win team in 2006.

On repeated simulations my program predicts a very slight chance of KC gettnig to 15 wins. Correcting for the effect above I'd estimate KC's actual chance of getting 15 wins to be between 0.2 and 0.5%.

banyon
06-03-2006, 06:21 PM
I'm sure I don't know nearly as much as you do, cdcox, about the art of making NFL predictions, but I have a genuine question.

Why do all of these predicted finishes look almost exactly like the way the teams actually finished last year? The only teams I see changing very much are Dallas (from 3rd to 1st), and Tampa Bay (from T-1st to 3rd). Other than that , there's a couple of cellar dwellers that switch, but little else.

But things never work out this way in the NFL and that's why it's the best sport going. Why don't take NFL predictions make any more bold guesses than they do?

Baconeater
06-03-2006, 06:35 PM
Questions like these are very useful to me in evaluating how realistic the my predictions are. Good question.

There have been three 15-1 teams (Bears, Vikings and Steelers) out of about 900 regular seasons in the 30 years since the NFL went to 16 games. Or about a 10% chance of getting one in any given season. When I looked at all the teams together, there was only a 5% chance of getting a 15-1 team in 2006. I noticed the Vegas projected wins tends to group teams around 8-8. This probably accounts for the underestimation of having a 15 win team in 2006.

On repeated simulations my program predicts a very slight chance of KC gettnig to 15 wins. Correcting for the effect above I'd estimate KC's actual chance of getting 15 wins to be between 0.2 and 0.5%.
Wow. I had no idea only 3 teams had ever gone 15-1. I figured it might have just been an immeasurable amount. Plus, if you're going off Vegas money lines, who the heck would put their money on KC to go 15-1 anyway? I know I wouldn't.

cdcox
06-03-2006, 07:43 PM
I'm sure I don't know nearly as much as you do, cdcox, about the art of making NFL predictions, but I have a genuine question.

Why do all of these predicted finishes look almost exactly like the way the teams actually finished last year? The only teams I see changing very much are Dallas (from 3rd to 1st), and Tampa Bay (from T-1st to 3rd). Other than that , there's a couple of cellar dwellers that switch, but little else.

But things never work out this way in the NFL and that's why it's the best sport going. Why don't take NFL predictions make any more bold guesses than they do?

Good question.

First, Pittsburgh is also predicted to finish first, when they were second last year. So that is 3 out of the 8 divisions were there is a change at the top of the division is predicted, along with a few other changes that aren't as significant (Arizona moving up to second in the division).

Second, lets take a close look at the five divisions where the division winner from last year has the best chance of repeating. Here are the program-predicted odds of x out of 5 of these teams repeating:

5 out of 5 repeating: 5.8%
4 out of 5 repeating: 23.0%
3 out of 5 repeating: 35.1%
less than 3 out of 5 repeating: 36.1%

So the program predicts at least one and very likely two or more of those 5 teams won't repeat as division champs. It just doesn't know which ones won't repeat.

Finally, these predictions are based on the opinions of the betting public. If you don't think enough change is projected, it is because the betting public doesn't think that much change will occur.

So in conclusion, based on my simulaitons, I believe there will be plenty of change in divison leaders by the end of the season.

Discuss Thrower
06-03-2006, 11:19 PM
My guesses:

AFC North:
Pittsburgh
Cincinnatti
Cleveland
Baltimore

AFC East:
New England
Miami [TBH, MIA winning the division is not too far out there]
Buffalo
New York JETS

AFC South:
Indianapolis
Jaguars
Tennessee
Houston

AFC West: [NOTE, I do believe that KC/DEN will have either a tied record or one that is no more than one game's difference.
Denver
Kansas City
San Diego
Oakland

Likely Wildcards: {Yes, I know there are only two WCs.]
[b]Cinci
Jaguars
Miami
Kansas City
S.D.
Tennessee

NFC North:
Chicago
Green Bay [Favre's last hurrah]
Minnesota
Detroit

NFC East:
Washington
New York GIANTS
Dallas
Philadelphia

NFC South:
Carolina
Atlanta
New Orleans [Provided Drew Bree's isn't torn to shit, The Saints will dominate any team with a weak defense.]
Tampa Bay

NFC West:
Seattle
Arizona
San Francisco
St. Louis

Wildcards:
New York/Dallas {One will eliminate the other}
Atlanta
Arizona
Green Bay


At some point, I'm gonna do my typical game-by-game horribly wrong prediction.

rad
06-03-2006, 11:25 PM
Welcome to the Planet J-Town Fan 1988 :toast:

Discuss Thrower
06-03-2006, 11:27 PM
Welcome to the Planet J-Town Fan 1988 :toast:

Oy, thanks for the cordiality. First one in a long while. ;)