What should the BCS formula be?

Discussion in 'The Tiger's Den' started by Indiana Tiger, Mar 22, 2005.

  1. Indiana Tiger

    Indiana Tiger Founding Member

    Joined:
    Jan 13, 2005
    Messages:
    509
    Likes Received:
    26
    Since I posted this, I thought I would add my comments. First, because this is about what can be done in the near term, suggestions like a playoff or eliminating/forcing conference championship games or the BCS conferences giving up their automatic bids aren't very productive. However, I'll offer an opinion on them in a separate post. I'll start by critiquing the CFN guys and then follow up with my own suggestions. To keep the posts somewhat moderate in length, I'll post them individually over time.

    Fuitak:
    I tend to agree with him regarding the expert panel. As someone else said, this is not baskeball. In general no one cares about who was the 66th team, but they really care about who was not #2 or the 2nd at large team. I also think they need as he says a Coalition of the Willing to make some tough decisions, but probably not the same ones that he is thinking about. While I would propose an alternative to the current models, I vehemently disagree that they should be done away with. A mixture of models and subjectivity is needed. I don't have a problem with the concept of SOS considering home vs road, but I would want to know how it would be determined before doing it.

    I like the idea of creating a new poll, but it needs more than 16 imo, way more. As far as it being full-time, watching every single game, and getting each ranking perfect, well there is no such thing as perfect and it's completely unnecessary. They really only have to come up with their best result (not the right result, because NOBODY knows what's right) for the pre-bowl poll. Everything else is preamble to provide focus for that decision. Except for #1, the final poll is just not that important. As Tommy Lee Jones said to Harrison Ford in The Fugitive, "I don't care." ... whether someone thinks it would be a better result that a team would be ranked #3 or #5 in the final poll. While it is important to ego and braggadocio, it has no consequential importance that I can see.

    Consider last year. The key decisions were 1) Au vs Ok; 2) Tx vs Cal; and 3) Should Utah be top 6. I don't know how in good conscience someone could say that these decisions didn't get adequate attention. They were the focus of fans and voters for weeks. Of course people could disagree with the results, but I saw no evidence that the voters, as a whole, didn't try to produce the best vote they could. This is the same process that happens every year as there are only a few key decisions to be made. This doesn't mean that no attention is paid to the others, but simply that the natural debate that transpires rightly focuses the attention on importance.

    In general, I've never liked the coaches' poll due to the inherent conflict of interest, but I think the AP has done a pretty good job historically. Yes, there have been some goofy results, but I don't know how you could ever design out goofy results. It's an opinion poll, and opinions must be given a wide latitude before declaring them unreasonable, or somebody is just playing games. I think that an opinion poll should play an important role in the process, but any poll will have foibles.
     
  2. cajdav1

    cajdav1 Soldiers are real hero's

    Joined:
    Dec 26, 2003
    Messages:
    7,493
    Likes Received:
    1,331
    Not sure if I agree with the entire post Indy but it is a very well thought out good opinion by you. you should write these kind of things for a living. :thumb:

    I don't think we will ever have a playoff unless something incredibly dramatic happens to change the mind of the presidents so I don't waste too much time thinking about alternatives.
     
  3. Indiana Tiger

    Indiana Tiger Founding Member

    Joined:
    Jan 13, 2005
    Messages:
    509
    Likes Received:
    26
    Zemek:
    I find the idea that SOS should be adjusted based on hopes and dreams to be completely absurd. We want to know what a team accomplished, not what they hoped to accomplish. Objective data should not be adjusted subjectively. This is the domain of the subjective component. If voters feel like team A is better than team B in spite of playing lesser schedule, then that is where it needs to be reflected.

    I would propose a new model to replace all the computer models. If MOV was included either directly or indirectly as an outcome of the process that created that model, I would accept it. Otherwise, I'm somewhat lukewarm about it. I'm more about dominance than MOV. IMO, MOV favors offensive teams over defensive teams. Defensive teams can be dominant without huge margins; they just win. This is another one that should be left to the subjective component. It seems like that there needs to be too many adjustments to make it fair (e.g. how did you score, when did you score, who did you score against, etc.)
     
  4. Indiana Tiger

    Indiana Tiger Founding Member

    Joined:
    Jan 13, 2005
    Messages:
    509
    Likes Received:
    26
    Thanks for the compliment. For the record, other than a 4 team or a +1 scenario, I'm not a playoff guy. It's not that I don't think a playoff wouldn't be a hoot. It's just that I don't think that the reality of greed would allow a design that would minimize the impact on the regular season. I think the impact of a large playoff (8 teams or more) would be huge, and it's a price I don't want to pay.

    BTW, I'm glad you figured out that the reason we don't have a playoff is not because the greatest minds in the country can't figure out how to fill out a playoff bracket. Playoff advocates might make more progress if they figured out what the real issues most likely are and propose realistic workable solutions.
     
  5. MarineTiger

    MarineTiger Founding Member

    Joined:
    Aug 30, 2004
    Messages:
    4,703
    Likes Received:
    4
    8-Team playoff with the BCS (with both human polls representing 75%, imo) deciding who goes.
     
  6. Indiana Tiger

    Indiana Tiger Founding Member

    Joined:
    Jan 13, 2005
    Messages:
    509
    Likes Received:
    26
    Harris:

    I agree wholeheartedly that there must be subjective analysis. I agree that the voters should be qualified, interested and conflict free. I don't agree that these people need to be full time, and I think you need whole lot more of them than proposed here. I want a broad based opinion poll. Members should be spread out geographically and cover both BCS and non-BCS teams. I also want diverse opinion on how to rank teams. Who really knows the best approach?
    Yes, models should be used. I really don't know what he is ranting about with his tweaking comment. No model ever used would rank 0 and 11 Twinkie State 13th, so I don't know what he is bitching about. It is completely wrong headed to impose the same algorithms on all the models. There would be no point to it. Averaging the same answer results in the same answer. You want models to emphasize different points of view as to the best way to rank teams, because no one knows what the best way is.
    This is another nonsense bitch, because this is happening today. SOS is considered by the voters and is a primary driver in the models. The voters will look at OOC schedule, and some of the models will partially determine their SOS by looking at the results between conferences. Finally, I thought AU didn't play in the championship game.

    This said, I would add a explicit SOS component. I wouldn't add it because SOS isn't considered enough; no one knows how much it matters now, so they can't tell you if it is too much or too little. I would add it because the networked nature of SOS gives fans an understandable reason to care about games that don't directly invole their teams. I thought it was great that the Boise St/Hawaii game mattered to us in 2003. As it always was, it should be scaled small so that it is essentially a tie-breaker when other components equal out.
    Should have been in place after Nebraska in 2001. cajdav1 asked what if the #2 team is from the same conference and that there are no other obvious alternatives. I would say it was a bad year for that team to finish # 2 in its conference. To me the regular season is the tournament and this team just didn't make it through the brackets.

    While I would like every conference to have a championship, I don't think they should be forced to have or not have one. Conference championship games exist only because they make money. Conferences that made this business decison must accept the risks along with the rewards that come with it. If the game is played the game must count just like I'm sure everyone would have insisted that USC's game vs VT should have counted last year if they had lost it. However, while losses do put a team at a disadvantage, wins give them an advantage against teams without them. LSU would never had played in the BCS CG in 2003 if there wasn't another game against a high quality opponent. It cuts both ways, and the conferences without these could also claim an unlevel playing field. I don't think conferences should be forced one way or the other. Let them make the business decision for themselves.

    While they would never agree to it and I doubt that the general public would accept it, I do think that non CG conferences should accept some risk in the years that they cannot produce a non ambigous champion (e.g. two teams tie and head to head doesn't resolve anything) and one of the teams is a candidate for the championship game. In this case they shouldn't be able to just put up the higher ranked team as the conference champion. If they won't have a playoff, then the champion should be chosen by a method that places the higher ranked team at risk of losing its spot. Maybe the winner could be chosen by lottery, with each team getting a certain number of chances based on their respective rankings.
    I agree with him that it's not going to happen.
    Forgetting about the Big East, the BCS conferences are extremely unlikely to forgo the automatic bids and simply go with a straight ranking. It's about the money. As it is they rig the money payouts so that all the BCS confs make about the same even when more than one team gets a bid. First they limit the # of teams to 2. Then the 2nd team only gets about 4.5M with the balance divided among all the conferences. They want to keep as much money as possible within the conferences and minimize the differences in payouts.

    If they had gone with a straight BCS ranking, in a couple of years the Big 10 wouldn't have had a team and in 2001, the Big 12 would have had 3 teams and the SEC would have had 2 teams, but the conference champion (LSU) would have been excluded. The latter case would devalue the conf CG and would probably be a bad business decision. I'm sure there are other examples that would probably piss people off and lead to claims to unfair treatment etc. After all the bitching and moaning, why would one start treating the BCS as science?
     
  7. Indiana Tiger

    Indiana Tiger Founding Member

    Joined:
    Jan 13, 2005
    Messages:
    509
    Likes Received:
    26
    What would I do?

    Short term I probably would keep the structure close to what was done last year. We would have to find a replacement for the AP and I would add back in the SOS for the reasons I stated before. I would also add the conference champion requirement. On another board, a seemingly knowlegeable poster reported that the BCS was looking at the Football writers as an alternative poll. I don't have any problem with this, but I do like the idea of starting from scratch with a poll. However, I think it has to be broader that what was suggested in the articles.

    I would like to drop the coaches' poll, but until an alternative could be developed that you could have some confidence in, they'll probably have to stay. I also think a fan poll would be a nice addition, if someone with adequate resources would step forward to do it right. Someone has to identify, vet, and select voters, provide the necessary tools and coordination for this to have a chance at gaining credibility.

    Longer term I would like to replace the computer models. My ideal would be to have select committee of recognized knowledgeable football men (BCS and nonBCS guys) work with a modeler to develop a single comprehensive ranking model. While I would hope that they would consider all phases of the game, I don't really care how simple or complex it is or what's in it or not. The only condition I would place on it is that the model must be completely transparent. Everyone needs to know what matters and what doesn't.
     
  8. tirk

    tirk im the lyrical jessie james

    Joined:
    Feb 4, 2004
    Messages:
    47,369
    Likes Received:
    21,536
    excellent analysis throughout, indi. I'll reply in full when I can give you one-tenth the amount of thought you obviously have on this subject.
     

Share This Page