You are the BCS Czar............

[Spence]

It was a shot, but I would guess it was in jest. No reason to get all worked up. I am from Appalachia and get those kind of comments all the time. Most of the time there is no harm meant.

[Spence]

I don't disagree with that, but if you start talking about what the definition of a legitimate reason would be, then you would start clouding the issue. What you and I would see as making no sense at all, someone else would think is perfectly logical. That has been proven out on this board with a small sampling of people. It would be like the best conference argument. You have millions of people who all strongly disagree with what the criteria would be. Add in all the PC junk about inclusion and you would have a bigger mess then you have now.

The way the BCS currently decides things isn't fair to all teams. We all can acknowledge that, it is when you start asking for a better answer, one that will solve all the problems about fairness, that you have problems.

The only way to actually know who the best teams really are is to have a playoff system with the best of 5 series between teams. That is the only way to insure that the best teams get to the final game. We all know that that isn't possible, even in the NFL.


I still think that if every team played 2 home and to away games with four of their peers (2 major and 2 mid major), it would go a long way to solving the fairness issues between all conferences. It would be a strain financially on some of the schools, I understand that, but it would solve the fairness issues. You would know how the SEC compares to the B-10, how the B-12 compares to the Big East, and how the majors compare to the mid -majors. There would be many questions with that.

I also believe it would help even out recruiting a little, because kids who went to mid major schools would have a reason to go there. They would be assured that at least four games on their schedules every year would be games allowing them to showcase their talents among good competition in front of a large audience. Kids who didn't want to red shirt at Ohio State might head to Miami instead.

Except for the travel issues and the bigger schools not being able to schedule that extra home game, I really don't see a down side to this proposal. The conference rivalries would remain intact. The traditions would remain intact, the only thing that would change is that the non conference matchups would be better. Ratings, nationally, for early games would go up.

Another good thing this proposal would do, is it would force Notre Dame and the other independents to join a conference. They would have to because the pecking order for the next years schedule would determined by where you finish in your conference from the year before.

[Spence]

I am sure there is a downside or someone smarter then I am would have thought it up before me, but I don't see a real downside other then the ones I have stated and I think those could be made up by getting more money from TV for having better matchups early in the season.

[colorado_loves_football]

The way the BCS currently decides things isn't fair to all teams. We all can acknowledge that, it is when you start asking for a better answer, one that will solve all the problems about fairness, that you have problems.

Actually, the BCS has become a lot more 'fair' in how teams are selected, believe it or not. Consider Boise St, for example. Clearly they aren't being given 'priority' in the BCS, regardless of what any one of us 'thinks' on the subject. If they make the BCS, it will be because they 'deserve' to, at least in the minds of the voters, which makes sense to me. SOS really isn't a component of the BCS anymore, although it might still play a role in how computers rank teams.

The only way to actually know who the best teams really are is to have a playoff system with the best of 5 series between teams. That is the only way to insure that the best teams get to the final game. We all know that that isn't possible, even in the NFL.

I don't see that happening, ever, myself. But, if you are talking playoff I think applying one, within the BCS format would be one possibility.

I still think that if every team played 2 home and to away games with four of their peers (2 major and 2 mid major), it would go a long way to solving the fairness issues between all conferences. It would be a strain financially on some of the schools, I understand that, but it would solve the fairness issues. You would know how the SEC compares to the B-10, how the B-12 compares to the Big East, and how the majors compare to the mid -majors. There would be many questions with that.

Competitve scheduling already exists, to some degree. I still believe conference games are more important than OOC games, but that's something we can maybe disagree on. The only way to know for sure how each conference 'sizes-up' relative to each other, would be to have a 'playoff' of conference champions, which is exactly what my proposal is.

I also believe it would help even out recruiting a little, because kids who went to mid major schools would have a reason to go there. They would be assured that at least four games on their schedules every year would be games allowing them to showcase their talents among good competition in front of a large audience. Kids who didn't want to red shirt at Ohio State might head to Miami instead.

I couldn't have said it better myself! But, I still think the 'smaller' schools would necessarily have to be ranked sufficiently high (A la Notre Dame). Top-12 is 'sufficient', which coincidentally is what the BCS uses.

Except for the travel issues and the bigger schools not being able to schedule that extra home game, I really don't see a down side to this proposal. The conference rivalries would remain intact. The traditions would remain intact, the only thing that would change is that the non conference matchups would be better. Ratings, nationally, for early games would go up.

This is where we disagree, in principle, Spence. I think schools should set up their own 'deals', without intervention. It makes for better pairings, that way. You need look no further than the Sun Belt to see what I'm talking about.

Another good thing this proposal would do, is it would force Notre Dame and the other independents to join a conference. They would have to because the pecking order for the next years schedule would determined by where you finish in your conference from the year before.

I think Notre Dame has been catered to, by the BCS. Last year, they 'cashed' in a $12 million jackpot, all for the Irish. Now, I know where the 'gold' comes into play. But, if nothing else, they have to meet a 'higher' standard, than other traditional BCS teams, progress of a sort, but I still would prefer they play within an established conference.

[Spence]

Notre Dame will be independent until they can't afford to be independent. Which means as long as NBC keeps paying the TV contract they are going to be independent. Notre Dame has a pretty good deal. They won't give it up unless they have to give it up.

[donovan]

I wold suspect also Notre Dame will never be independent. They are able to schedule games, both in football and basketball, with whomever they want, because their away games guarantee a built in crowd, many of whom are Notre Dame fans. There are other schools in a similar situation. BYU has almost filled the stands at their away games with their supporters.

Fair is an important element of life but sometimes the cause is in the definition of the problem. All a National Champion, or even a conference champion means is they were the best at the time they played in whatever system is used. Few people believe that in the long haul, St Louis was the best team in baseball. But for the system, they won and are the Champions. Remember back at the BYU Oklahoma fiasco for the football championship. Do we think that in individual games Washington State would be USC four out of seven?

So as admirable as fairness is, and it should be pursued, it will be elusive if we continue to objectify a subjective issue.

So do we like the Congrove Computer rankings...which brought most all of us to this site? I have never really understood the criteria. It always so far off the other polls. But the impressive point is it prediction percentage. Makes me wonder if Congrove is more accurate. Anyone knows how it works and why the difference?

[Spence]

The way I understand it, Congrove didn't like the polls and wanted to do it better. I don't agree with the computer's rankings sometimes, but I would bet he doesn't aways agree with them either. In any case the computer picks all the games and puts it out there for everyone to criticize. They are right almost 75% of the time. In our picks game no one has ever picked a sampling of games and picked at 75% for the season, let alone having to pick every game at that clip. So who is to say that we are right and the computer is wrong? Obviously the computer picks games better then us.

[Howdy]

The way I understand it, Congrove didn't like the polls and wanted to do it better. I don't agree with the computer's rankings sometimes, but I would bet he doesn't aways agree with them either. In any case the computer picks all the games and puts it out there for everyone to criticize. They are right almost 75% of the time. In our picks game no one has ever picked a sampling of games and picked at 75% for the season, let alone having to pick every game at that clip. So who is to say that we are right and the computer is wrong? Obviously the computer picks games better then us.

-----------------------------------------------------------------------------

The computer rankings are needed also.

People who have them have to update them each week and they do all 119 teams and they work in S O S into them.

People who vote like those voting for AP ETC only vote there top 25 .
Voting like voting some one into office (like I like him or I like this team).

[openSkies]

Spence,

This is a long thread and I didn't have time to read it all but... doesn't the BCS require the computers to use SOS now? I believe that's why everyone had to either change their formula or leave the BCS a few years back. Maybe it was quality wins and not SOS. Please, correct me if I'm wrong.

Donovan/Spence,

Congrove started the poll just to see if he could do it, really. Or at least that's what I gathered. I don't think the BCS was around back then, and I know we didn't have the internet for him to look at other computer rankings and say "Hey, I can do better". It started off as a hobby, transformed into a radio show, and then 7 years later went online in 1998. If you want the low-down on how his whole system works (minus trade secrets), I can have CFP Admin come in and explain things a bit better. I don't know the system by any stretch of the imagination. I take a look at the spreadsheet and my mind instantly shuts down, drool starts collecting at the corner of my lip, and eventually I fall over and pass out.

[CFP Admin]

The question in the poll is moot. It already exists in every human and computer ranking.

First - I have copied and pasted the basic information about the CCR Top 119 here:

Dave Congrove is a college football fanatic who devised the Congrove Computer Rankings in 1993 as a means of providing an unbiased comparison of every Division 1-A college football team.

His approach is vastly different from the numerous computer rankings available and the goals can be summarized in two words - simplicity and accountability.

Simplicity. The computer program is designed to predict who is supposed to beat who and by how much.

What actually happens when the game is played directly alters subsequent predictions. A unique "season-averaging variable" prevents wild fluctuations and renders it meaningless to "run up the score".

A strong team that is supposed to whip an obviously out-matched opponent by 40, but wins by 60, only sees a fraction of the difference added to their power rating. If they win by only 20, the failure to "cover" the other 20 points also declines their power rating only fractionally.

The rise and fall of a team's Power Ratings are weighted toward the strength of the team they played, and whether they won or lost.

Accountability. The Congrove Computer Rankings are designed to predict the outcome. Success is determined by how many games were predicted correctly.

College Football Poll.com not only publishes the rankings, but also provides the computer's pick of every game and tracks the success rate by each team and conference. Our users can plainly see where the computer is right, and where it is wrong.

Additionally, the Season Preview section features team pages that highlight the computer's predicted record of every team for every year since 1993, and compares it to their actual record.

Results. The computer has picked the winner in 74.7% of all games played since 1993 (6,644-2,245) while beating the spread in 54.6% of those games (4,487-3,727) (year-by-year record of computer).

For 2005, it projected the exact regular season record of 13 teams, came within one victory on 46 others, and within two victories on 22 more teams. That means the computer came within two wins of forecasting the records of 81 out of 119 teams.

The computer has predicted the national title match-up three times (1993, 1998 and 2005), and at least one contestant every year except 1994, 1997, and 2001. The team it picked to win the title did so three times (1993, 1998, and 1999), and lost in the title game four times (2000, 2003, 2004, 2005).


Second - Whether you agree or disagree with the results, they have their proven record of reliability. The measurement of the ratings is NOT based on what other polls and rankers do, it's based on the success of picking game-winners and spreads which automatically points to its use of schedule strength, and its absence of bias.

Third - You can read his own points-of-view on this in several articles that he has written over the years that are all archived on this site.

http://www.collegefootballpoll.com/special_features3.html (From 7/31/01)

http://www.collegefootballpoll.com/special_features4.html (From 10/19/01)

http://www.collegefootballpoll.com/special_features5.html (From 12/12/01)

http://www.collegefootballpoll.com/special_features6.html (From 12/8/03)

http://www.collegefootballpoll.com/games_preview_092005_bcs.html (From 9/20/05)

[..fanatic]

Sometimes it is not necessary to say, we are better than you. It is OK to enjoy the journey of the season and then just move on.

Amen, brother. Amen.

[..fanatic]

First it presupposes that programmers do not influence the algorithms towards their bias...probably some more than others......

People who create computer rankings are too interested in seeing them succeed. There is no way they would tinker with them. A little too much "conspiracy theory" ideology for me to agree with you on that.

[CFP Admin]

The way I understand it, Congrove didn't like the polls and wanted to do it better. I don't agree with the computer's rankings sometimes, but I would bet he doesn't aways agree with them either.

To borrow from Lynyrd Skynyrd, "You Got That Right".

[donovan]

Thanks for the explanation. You may not agree with the methodology, but you have to admire a philosophy that someone sticks with and does not shift with every wind, especially when it has proven results. I can think of all kinds of professions that do not get it right 74% of the time.

[Spence]

I know the quality win thing was taken out, but that was separate from the computers or the human polls. The computers, if I understand it correctly, have had to take the "margin of victory" component out, but can still use SOS. Even though some would consider margin of victory part of the SOS equation.

My point in all of this is that the BCS computer polls should have changed to conform to what the BCS wanted. If they thought they had a good system, they should have stuck to their guns. Congrove has done that, from what I understand, and that makes his results more credible. He puts his results out for all the world to decide what they think of it. He publishes his results and doesn't try to change the system everytime he thinks he may have gotten a bad result. Congrove stands behind his results and that is all you can really ask.