cdcox's 2004 playoff prediction thread
 

I've decided to post all of my playoff predictions in a single thread this year. That way it will be easier to track threads and always find the latest predictions.

Chiefs Playoff odds (division title odds, making playoff odds)
week 7: 1.6%, 3.5%
week 8: 7%, 14%
week 9: 3.4%, 8%
week 12: zip
week 13: technically 2 out of 1000, in reality zip.
week 15: technically 2 out of 1000, in reality zip.

AFC Division Winners
week 7: New England, Pittsburgh, Jacksonville, Denver
week 8: New England, Pittsburgh, Houston, San Diego
week 9: New England Pittsburgh, Indianapolis, Denver
week 12: New England Pittsburgh, Indianapolis, San Diego
week 13: New England, Pittsburgh, Indianapolis, San Diego
week 15: New England, Pittsburgh, Indianapolis, San Diego

NFC Division Winners
week 7: Philadelphia, Minnesota, Atlanta, St. Louis
week 8: Philadelphia, Minnesota, Atlanta, St. Louis
week 9: Philadelphia, Minnesota, Atlanta, Seattle
week 12: Philadelphia, Minnesota, Atlanta, Seattle
week 13: Philadelphia, Green Bay, Atlanta, St. Louis
week 15:Philadelphia, Minnesota, Atlanta, Seattle

An evaluation of the accuracy of the predictions is available here:

http://www.chiefsplanet.com/BB/showt...ff+predictions

After week 7:
At these very preliminary stages, I estimate that the Chiefs chances of making the playoffs are only about 3.5%. This is lower than the estimate I posted in another thread earlier today because the website I down loaded the scores from had an error. I think the revised estimate is a little low for reasons that I will explain below. In many of the cases where they made it, they did so with a 9-7 record. Projected AFC division winners right now are New England, Pittsburgh, Jacksonville, and Denver*. Strongest additional contenders are the Jets, Indy, San Diego, Baltimore, Houston and Cleveland. As you can see, things are pretty crowded in the AFC.

In the NFC, projected division winners are Philly, Minnesota, Atlanta, and St. Louis. Also in the thick of the race are the Giants, Detroit, Seattle, Arizona, Green Bay, Dallas and nearly everyone else. The quality of teams in the NFC is much worse so far. My program is predicting that 8-8 or even 7-9 might be enough to win the NFC west or a wildcard.


The attached txt file contains the detailed breakdowns. For each team, the first number is the odds of winning the division, the second number is the odds of making the playoffs, and the third number is the power rating I used to calculate the outcome of the remaining games.

*Violated the salary cap during their championship seasons

Overall, these predictions seem a little optimistic for the front-runners. This is always a problem early in the season, but is even more so this year because so far their have been relatively few games where a team of lower power rating beats a team of higher power rating. Tinlar pointed this out the other day in his power ratings. This leads my program to believe that the NFL is predictable with a high degree of accuracy and favors front-runners disproportionately. As we get a few more upsets, things will even out. Another way of handling this is to include uncertainty in the power ratings (Bayesian statistics), but that would require another revision of the program and a few hours of computer time to calculate.

Very nice work!

I'll rep you even though KC has only a 3.5 percent chance....

Rep to cdcox for the asterisk.


Mods, is there any way we can replace the Denver Smiley with an asterisk and cdcox's footnote? That's be cool.

Sure... rep me now that you have been discounted 4%.

Again I say this, we must win out this year to get the division title.

Jax. and Indy both will go around 12-4, 11-5 with there schedules. The Jets are a lock with there schedule, unless either of these 3 teams does a complete fallout, we are out of the picture in the WC.

The Indy game is a must for 2 reasons, game up on WC and the fight to finish winning out !

After Week 8 (assuming the Jets hold on):

Kansas City’s playoffs chances climbed to about 14% with the win over the Colts. Hope should be officially alive. Division favorites in the AFC are New England, Pittsburgh, Houston!, and San Diego! Other strong competitors include: the Jets, Denver*, Jacksonville, Baltimore, Indy, and KC.

In the NFC, division favorites are Philadelphia, Minnesota, Atlanta, and St. Louis. The strongest competitors among the remaining teams are the Giants, Detroit, Seattle, Green Bay and Dallas.

The attached txt file contains the detailed breakdowns. For each team, the first number is the odds of winning the division, the second number is the odds of making the playoffs, and the third number is the power rating I used to calculate the outcome of the remaining games.

*Violated the salary cap during their championship seasons

shameless pimp

Broncos

Pretty freaking hilarious, huh?

As it stands now, with Denver and SD tied at 5 wins ... and Denver having already defeated SD, how is it that SD is the AFCW favorite?

I like the work you do ... just curious.

Really,it's simple, all we need to do is win them all, the division is ours, Denver and SD will be taken care of as SD beating Denver.

Now the hard part of simple ... winning out !

Thanks for the question. These are useful because it requires me to check that things make sense, thereby identifying areas for improvement.

First of all, San Diego only has a slight edge over Denver* for the predicted division winner. 48.9% to 43.9%. You can see this in the detailed text. The difference comes down to the power ratings: San Diego has a slightly higher power rating, thus meaning that the computer predicts that it will win more games from here on out than Denver*. This has to do with the outcomes of the games so far:

San Diego:
Wins over: Houston, Tennessee, Jacksonville, Carolina, and Oakland. Losses to: Jets, Denver*, and Atlanta

Devner*:
Wins over: Kansas City, San Diego, Tampa, Carolina, and Oakland.
Losses to: Jacksonville, Cincinnatti, and Atlanta

San Diego has two wins against quality teams (Houston and Jacksonville) and all their losses were to quality teams. Denver* has two wins over quality teams (Kansas City and San Diego) but thier losses include one to a lower quality team (Cincinnatti). That is the difference in power rating, and thus the reason they are slightly favored to win the AFC west. Of course my program is completely wrong about this. KC will win the west.

*violated the salary cap during thier championship seasons.

Thanks cdcox!

:thumb: