A different look at the BCS

[Buck 'em up]

I have posted this in other places but have never received any feedback. I hope in a place like this these rankings of mine will at least garner some discussion. Since I am sure most people on this site are from Ohio or at least are familiar with the way Ohio determines high school state playoff teams you will understand these.

The basis for these rankings is very simple. Points are awarded for victories based on Football Bowl Sub-division BCS Conference teams (5 points), FBS Non-BCS teams (3), Football Championship Sub-division (1) and any other level (.5). Each team also get secondary points based on the same scale from each defeated opponent. The total points is then divided by the amount of games played producing an average, then the number of losses is subtracted from the average and that number is used to rank the teams. This method is based on the rankings used by the Ohio High School Athletic Association to determine the teams that qualify for the playoffs.

An example --

Ohio State gets 3 primary points for beating Navy. Ohio State then gets secondary points for Navy's wins over La. Tech (3 points), Western Kentucky (3), Air Force (3), Rice (3), SMU (3), Wake Forest (5), Notre Dame (5) and Delaware (1) for a total of 26. That is 29 points for the win over Navy. This is done for every defeated opponent. Ohio States total primary points comes to 44 and their secondary points are 232 for a total of 276. That number is then divided by the number of games played (12) for an average of 23.0000. Then the two losses are subtracted for a final total of 21.0000.

Top 25

1. Alabama 11-0 25.5909
2. Florida 11-0 24.1364
3. Texas 11-0 24.0909
4. Cincinnati 10-0 21.9500
5. Georgia Tech 10-1 21.3636
6. Ohio State 10-2 21.0000
7. TCU 11-0 20.3182
8. Oregon 9-2 20.1818
9. Iowa 10-2 19.0000
10. Pittsburgh 9-1 19.0000
11. Boise State 11-0 16.5909
12. Oklahoma State 9-2 16.5455
13. USC 7-3 16.3000
14. Penn State 10-2 15.9167
15. Miami (Fla.) 8-3 15.6364
16. Virginia Tech 8-3 15.1818
17. Clemson 8-3 14.5455
18. North Carolina 8-3 14.2727
19. Houston 9-2 13.3636
20. California 8-3 12.9545
21. Wisconsin 8-3 12.9091
22. Oregon State 8-3 12.6818
23. Nebraska 8-3 12.3636
24. Arizona 6-4 11.5000
25. Stanford 7-4 11.2727

The same method is used to rank the conference except losses aren't subtracted or everyone would be negative.

PAC 10 59-48 12.8879
Big 10 75-54 12.7791
SEC 83-50 12.5414
ACC 73-59 12.2235
Big East 53-29 11.9817
Big 12 78-56 11.9254
MWC 53-47 7.2500
C USA 60-70 6.4846
WAC 48-50 5.9898
MAC 62-82 4.9618
Sunbelt 38-58 4.2031

I know it a geeky kind of thing to do, but I love college football. Thoughts, concerns, questions?????

 

[Buck 'em up]

If I knew how to post and Excel sheet/table the numbers would be easier to read. But I think you can make out what it says.

 

[SCBuck13]

Not a bad idea.

There's no way this system would be implimated, but thats cool to see.

Nice work.

 

[Buck 'em up]

Not a bad idea.

There's no way this system would be implimated, buts that cool to see.

Nice work.

Thanks. I think there is ONE major reason this would never be allowed by the NCAA... There is a strength of schedule -- to some extent -- aspect to it.

 

[LordJeffBuck]

The basis for these rankings is very simple. Points are awarded for victories based on Football Bowl Sub-division BCS Conference teams (5 points), FBS Non-BCS teams (3), Football Championship Sub-division (1) and any other level (.5). Each team also get secondary points based on the same scale from each defeated opponent. The total points is then divided by the amount of games played producing an average, then the number of losses is subtracted from the average and that number is used to rank the teams. This method is based on the rankings used by the Ohio High School Athletic Association to determine the teams that qualify for the playoffs.

This approach works only if every team plays the same number of games. The problem arises because teams who play conference title games will have higher averages, and the losers will have an extra loss subtracted from their totals.

For example, if Ohio State (44 primary points, 232 secondary points, 276 total points, 23.00 unadjusted points per game, 21.00 points per game after losses) played and beat Iowa (44 secondary points) in a conference championship game, then Ohio State would end the season with 49 primary points, 276 secondary points, and 325 total points, for an unadujsted average of 25.00 per game, and a net after losses of 23.00 points per game. Thus, based on beating an extra opponent with a high secondary points total, Ohio State would jump from #6 overall to #4 overall, passing Cincinnati and Georgia Tech in the process. On the other hand, Iowa would lose a point, and would fall from #9 to #10 overall (falling behind Pitt).

So, good idea, but you've got to find some way to factor out the conference championship games - good luck!

 

[GoBucks1014]

This approach works only if every team plays the same number of games. The problem arises because teams who play conference title games will have higher averages, and the losers will have an extra loss subtracted from their totals.

For example, if Ohio State (44 primary points, 232 secondary points, 276 total points, 23.00 unadjusted points per game, 21.00 points per game after losses) played and beat Iowa (44 secondary points) in a conference championship game, then Ohio State would end the season with 49 primary points, 276 secondary points, and 325 total points, for an unadujsted average of 25.00 per game, and a net after losses of 23.00 points per game. Thus, based on beating an extra opponent with a high secondary points total, Ohio State would jump from #6 overall to #4 overall, passing Cincinnati and Georgia Tech in the process. On the other hand, Iowa would lose a point, and would fall from #9 to #10 overall (falling behind Pitt).

So, good idea, but you've got to find some way to factor out the conference championship games - good luck!

What a novel idea, the NCAA dictates the number of games its member schools play. Seems to me a great example of why you bypass the end of season money grab, er CCG presented by..., and move to a true post-season. Of course this would require conferences to align their number of schools to provide a regular season champion.

I know the NCAA will never do this, but I'm just sayin'.

 

[Buck 'em up]

This approach works only if every team plays the same number of games. The problem arises because teams who play conference title games will have higher averages, and the losers will have an extra loss subtracted from their totals.

For example, if Ohio State (44 primary points, 232 secondary points, 276 total points, 23.00 unadjusted points per game, 21.00 points per game after losses) played and beat Iowa (44 secondary points) in a conference championship game, then Ohio State would end the season with 49 primary points, 276 secondary points, and 325 total points, for an unadujsted average of 25.00 per game, and a net after losses of 23.00 points per game. Thus, based on beating an extra opponent with a high secondary points total, Ohio State would jump from #6 overall to #4 overall, passing Cincinnati and Georgia Tech in the process. On the other hand, Iowa would lose a point, and would fall from #9 to #10 overall (falling behind Pitt).

So, good idea, but you've got to find some way to factor out the conference championship games - good luck!

I thought about this and may have a solution. Since the most games any team can play -- I believe -- is 14, and the majority of teams play 12, teams that play 12 games will get 100 percent of their primary points, 13 games would get 93 percent and 14 games would get 86 percent. In order to ensure that teams only playing 12 don't receive and additional bonus for not playing additional games, the secondary points that each team gets from a defeated opponent would be reduced in the same fashion. They would only get the same percentage of points from the defeated team. At game 13 for OSU in your scenario, OSU would get 45.57 of their 49 primary points and USC and Purdue would also get 45.57 of the 49 points. That also means OSU would get 81.84 points (40.92 from each win over Iowa) for a total of 45.75 primary points and 269.84 secondary. At this point it is hard to tell if it would work because Cincinnati has only played 10 games as opposed to the mythical 13 OSU would play. And Cincinnati, at the very least, has two more games to accumulate points. I will input the numbers as the season goes on and see how it works out.

Thanks for the input.

 

[JDMBuckeye]

Take a look at this:

USATODAY.com

Part of this system is used for the BCS rankings. It explains it on the website - factors is SOS, Win Margin, and Record. Its a useful source of information. Let me know if you think it is accurate - i agree completely and refer to it as opposed to the AP poll or the BCS standings in which human opinion is used as a bias.

 

[JXC]

Take a look at this:

USATODAY.com

Part of this system is used for the BCS rankings. It explains it on the website - factors is SOS, Win Margin, and Record. Its a useful source of information. Let me know if you think it is accurate - i agree completely and refer to it as opposed to the AP poll or the BCS standings in which human opinion is used as a bias.

I think we've all heard of the sagarin ratings.