The selection committee: a modest proposal

I don’t often find myself in agreement with Derek Dooley, but what can I say?  When you’re right, you’re right.  If the commissioners and presidents fail to fashion a selection committee (assuming that’s their choice, of course) the public finds credible, the new playoff format will be doomed from the start.  By itself, locking a few wise men in a room with instructions not to emerge until they have four names isn’t going to cut it.  So what will?

Well, to start with, there are certain essential truths that have to be recognized.

  1. A four-team playoff doesn’t end controversy over the postseason field.  In fact, as Tony Barnhart acknowledges, it’s likely to fan the flames:  “It won’t be less controversial. In fact, a four-team playoff will actually be MORE controversial once it is in place. Why? Because last season there was basically only one team (No. 3 Oklahoma State) that had a legitimate argument that it should have been in the top two. When we go to four teams, there will always be at least 3-4 teams that claim they were as good as No. 4. “  Worse, there will be years like 2005, when there was a clear consensus on the two best teams in the country and a four-team playoff will seem essentially superfluous – not that it’ll be cancelled – and thus, so will the fierce debate over which teams fill out the field.
  2. A selection committee is not going to be perceived as a friend of the little guy.  As Andrea Adelson puts it, “… I fear for conferences like the Big East with a selection committee in charge. How do we know that the group of men charged with making the decisions are going to choose the best four teams while turning a blind eye to conference affiliation? Particularly when a league like the Big East has such a weak national perception.”  Short answer:  we don’t.
  3. Fans ain’t stupid.  We’re really not.  As Year2 says, don’t insult our intelligence by insisting that the process will be completely unbiased.  It’s not humanly possible to do that.  When you’re being successfully mocked by Derek Dooley, that ought to give you a clue there’s a problem.

So, if those are the obstacles to creating a selection process which results are validly perceived, how can college football surmount them?  By making sure that no conflicts of interest exist – no coaches or athletic directors need apply, in other words – and that bias is minimized or “balanced”, as Year2 puts it, to the extent that the public can respect the results.

That starts with sunshine.  Whatever standards are put in place to judge the field, they need to be both public and relatively easy to understand.  For instance, if computers are to stay in the mix, whether as a standard which humans are to rely upon in making their decision or as a direct component in a formula that is used to compile team rankings, no more of this proprietary nonsense.  Either commission a stat guru to run numbers based on a formula of college football’s own devising, or use something that has no hidden features.

As for the selection committee itself, what makes me nervous is that it appears to have been the source of compromise for the commissioners in coming up with the new playoff format.  In that context, it’s a compromise in the worst sense of the word:  Slive, Delany, Scott, et al. no doubt see the possibilities of structuring such an animal in ways which benefit each the most.  Sharks don’t compromise out of a sense of the common good.  They do it because they perceive an opportunity.  How do we avoid giving them that?

Looking at Year2’s suggestions, there’s some good stuff.  I like the idea of each voter keeping a public log of games watched, so we know they’re taking their responsibilities seriously.  There are also ideas that I doubt college football is ready to adopt, like treating committee membership as a full-time paying job.  And while having weekly grillings of committee members would make for good TV, no doubt, do we really want to put that kind of power into the hands of an ESPN?  (Not that the WWL would object.)  I also wonder if a 10-person group is big enough to balance out bias, especially if each conference names individual representatives to the committee.

I’d suggest a different direction.  And if you know GTP, the home of the Mumme Poll, you know what I’m getting ready to suggest.  Quite simply, college football ought to embrace approval voting.  Instead of a small committee, put together a pool of 124 knowledgeable voters, balanced geographically and with varied backgrounds.  After the sixth week of games are in the book, have them start casting weekly public ballots of their top ten teams (not in order!) along with game logs and any commentary about how they arrived at their decisions so that they can be vetted, not by ESPN, but by a bigger world.

I’m not suggesting it’s a perfect solution in and of itself, of course.  But a pool that large in an approval vote setting makes it difficult for a single individual to manipulate the results and making the process open allows the press and the general public to watch for evidence of larger scale collusion or for individuals trying to pursue dubious agendas of their own.  What I can say from our own experience operating the Mumme Poll for several seasons now is that approval voting is a useful tool in combating bias.  And I’m sure there are smarter people than I am who can come up with ways to tweak the process to further strengthen its validity.

College football is unique.  There’s no reason for it to follow another sport’s postseason selection process, even if it’s one with which there’s great familiarity by those who will fashion it.  Do better than that, folks.  We’d appreciate it.


Filed under BCS/Playoffs, Mumme Poll

47 responses to “The selection committee: a modest proposal

  1. Senator, as an MP participant last year, I thought it worked very well. Approval voting is definitely the way to go. I still think there is going to be controversy about #4 just as there is with last team in and first team out during March Madness.

    • I probably should have said this specifically in my post, but that’s one of the best features of approval voting in this context. No individual voter has to choose between a number four and a number five and the final results don’t have to be justified in that way, either.

      • Puffdawg

        Regardless of whether the final results have to be justified… they’ll have to be justified, which is impossible. What you’ll see happen is the fans and coaches (and ADs and presidents of schools left out for that matter) will complain not that the 5th or 9th or 17th teams have a legitimate argument to win the national title, but that they have a legitimate argument to be in the tournament relative to the 4th, 8th, or 16th teams.

        • Beer Money

          Precisely. What is always fascinating to me is the gnashing of teeth that goes on right after all 60-hwoever many teams it is now (does it really matter?) are announced and Dick Vitale and ESPN come out with pitchforks and torches because some mid-major who finished second in their conference or 13th place Big east team was left out of the Dance. What the hell does it matter? None of these teams will ever win the tournament. As a matter of fact, you can usually narrow it down to about 4-8 teams who have a legitimate shot anyway…sound familiar?

          What is even more fascinating is that this talk dies almost immediately once the show is over never to be reopened again.

          • Spence

            Outrage sells. Look at talk radio. Most of the outrage is entirely fabricated, worthless, and unproductive. But we’ve managed to convince ourselves there is something wrong with college football, and that has correlated with the rise of 24 hour sports coverage.

            • Elkon had a great post on Atlanta sports talk radio awhile back and how it never talks about anything substantively, but instead devolves into simplistic moral outrage.

        • I would argue that it’s not so much a question of justifying the final results as clearly understanding how they’re reached. Sort of like how Richt describes things – “whatever the rules are… just tell me what they are, and we are going to play.”

          • Mayor of Dawgtown

            The problem is when the rules appear to be fluid and changing, often in mid-stream. Part of it is the doctrine of unintended consequences. Last year 2 SEC teams (for the first time) played in the BCSNCG and the SEC got both ends of the BCSNCG money. On top of that a team that didn’t win its conference (indeed didn’t even win its DIVISION) won the BCSNCG and got the crystal football. How high is the likelihood of that happening again? i don’t know but Delaney and others are going to be damn sure it DOESN’T happen again, so they will impose a way to avoid it, probably a stacked deck selection committee.

  2. Dawg93

    I don’t get the Big East complaint there – they have a “weak national perception” because they’re, ahem, weak. West Virginia is the only team that has flirted with a national title (that RichRod team that was upset by Pittsburgh the last game of the year). Outside of that, they haven’t done squat on a national level. And now they’re gone too. So why should any give a rip about one of their teams being left out??

  3. Spence

    Thank God we are finally making college football a watchable enjoyable sport. It sucked before.

  4. I hate the idea of a committee. If you have a big sample size, as we do with the Harris Poll, then the various idiosyncrasies of the voters ought to cancel one another out. (Our political system is founded on this principle.) If you have a small sample size, as we will with a ten-person committee, then one person’s idiosyncrasy is going to have far more weight. Do we really want Bobby Bowden, who was addled for the last five years of his coaching career such that he didn’t know the names of his own players or whether they had been kicked out of the game, as a major player? Or Gene Stallings, a run the ball and play defense advocate, trying to decide between a team that plays his favored way and some new-fangled spread team whose offense he doesn’t understand? And then you get into personality factors, such as that there might be an alpha in the committee who gets the others going his way. (This happens with juries all the time.)

    Plus, the outrage is going to be far more intense after committee decisions because fans will be able to focus their anger. It’s harder to get angry at hundreds of pollsters and eight computer rankings.

    I like the idea of Vegas making the decisions, although it occurred to me this morning that they will have a massive conflict of interest because they will want public teams in the game to increase betting. If Michigan and Boise State are fighting for the last spot, whom do you think Caesar’s would like in the playoff?

  5. kdsdawg

    ah, the Big East is 16-18 in bowl games against the ACC all time and 5-5 in the last ten. Why does the ACC have a better rep than the Big East?

  6. DawgPhan

    any time you use math to determine who is best there are going to be some issues with it. I run the National BBQ Rankings and use a formula to rank teams from around the country in doing so teams always complain that the competition in their state is tougher than another state and that their win should be worth more than another team’s win.

    BTW GA thinks that they are the SEC of BBQ. Thinks that Cali/Out West is the Big East. Iowa is probably the Big 10, Texas is Texas, and New England is the Pac-12.

    • paul

      Okay, maybe not just Georgia but Georgia, Alabama, Mississippi, Tennessee and Kansas city. That would be the SEC of BBQ.

  7. Cosmic Dawg

    I don’t know why we can’t just have 7 conference champs and one AP ranked indie play an eight game tourney and everybody else plays 10 days of bowl games between the regs season and the tourney. If you don’t win your conf, you lost the “first round” of the “playoffs” – this is the most tedious conversation.

    That would also cause strong teams to join weak conferences, evening out $ disparity, sos questions, selection bias, etc

  8. Cojones

    Looks like an 8-team playoff is becoming a solution (ducks again).

    • Dog in Fla

      Miss Manners feels that it may be polite for the eight-teamers to wait at least until the vase-breaking ceremony in DC for the four-team playoff next week before asking for the eight-team playoff

  9. AusDawg85

    My goodness…the TRUTH about GTP at last! It’s all about getting the Mumme Poll adopted by CFB and your reaping the licensing rights.

    Well played, Senator….well played indeed!

    • Licensing rights?

      Let me just say for the record – I may not be cheap, but I can be had.😉

      • Saint Johns Dawg

        Senator, if you think about it, the “hidden computer” issue you bring up is exacerbated by terrific research and commentary by folks like you. The WWL will certainly want their “selection show” to be exclusive to incur maximum ratings and ad revenue.

        If the computer selection data was public (including how it’s derived), then nobody except the fans of the top 4 teams would have a need to watch the “selection show” … the games would be determined and commented on long before.

        Not saying that’s a bad thing (love me some First Amendment all day, every day) … just saying money is the root of that evil, too.

        • I don’t think anybody is suggesting that a computer model would be the exclusive gauge of entry into the playoffs. If you want the drama of a selection show, just time the release of the final vote for that.

  10. Cojones

    Senator, how would “knowledgeable” be determined. Many would say that the people who play or write about the sport for a living would have the most knowledgeable voters. How do we get below that level to arrive at 124 members?

    While not wanting to come off as a sycophant, how do we find 124 people who can perform beyond their prejudices and biases? There is only one of you. Are the others recommended by someone who comports themselves as you do? Even Barnhart doesn’t make it. Hey! Wait a minute! Are you proposing that bloggers like you, who touch the CFB pulse each day, should make up the greater part of the committee? Now that would put daily bloggers in touch with the decision-making; make the process more than transparent; not involve very many who have other fish to fry and or, at the least, dilute them out of making the needle move in one direction or the other as one voter; make the blogosphere the overseer of CFB.

    Man, wouldn’t that be something?

    • In a country with a population of more than 300 million, I would suggest it shouldn’t be that hard to find 124 people who know a satisfactory amount about the game of football.

      And, again, I don’t think the goal should be to find 124 perfectly unbiased souls. It should be to devise a system that minimizes the committee members’ prejudices.

  11. Nate Dawg

    I know this is strayin’ a bit – but I’ve never been impressed with the ‘O5 Trojans or ’05 Longhorns. As memory serves, Oklahoma was down that year and despite all the highlights of R Bush you see, USC was one bad pass away from losing that damn Fresno St game (that I wanted them to loose oh so badly…). Therefore, who did they play to get there? In this instance, I think it woulda been cool to see a Va Tech beat one of them in a playoff game or, dare I say, an SEC Champ UGA team take it to one of them that year (although I’m not in love with this whole playoff thing…we still get to play awbarn, right?). Maybe we could of avoided all those camera shots of V Young and the confetti…uugh…

    • For one, I don’t think there was really a compelling case for any other participant to be in the conversation in 2005 (didn’t VaTech lose to FSU in the ACC title game?).

      Additionally, just doing a little research – Texas and USC were 3-1 in total offense, 1-2 in offensive points per game, 6-40 in total defense, and 4-27 in defensive points per game. Okay, defense was not so hot for Southern Cal there but weirdly enough Texas tied Georgia at #4 overall…sigh. Sagarin also rated the schedules for Texas and USC as 13 and 8, respectively. It’s hard to argue that those weren’t the two best teams in the country out there despite the near-hiccup against Fresno. If I remember correctly, Texas straight-up obliterated everyone on their schedule except for Ohio State (played at the horseshoe) and in the national championship game.

      Sorry to stray with you there, but I thoroughly enjoyed that 2005 Texas team with VY.

      • Nate Dawg

        That’s fine, I was just throwing Va Tech out there on the hypothetical-they did loose to FSU in Champ game, you’re right. And all those numbers & stats that they achieved – I’m just not impressed with the competition. You mean THE ohio state, the SEC whippin’ boy?? Fresno? I’m sure you’re right about Sagarin, but the Nate Dawg rankings were not near as high for both those schedules, and that’s the one I’m really worried about, ha!

        • Mayor of Dawgtown

          If you look hard at several BCS National Champions and the teams they faced in the BCSNCG they come up wanting. The year THE Ohio State University beat Miami is an example. Tennessee sucked and won the crystal football anyway. The year Auburn (not that I have any love for them) went undefeated and got locked out the War Tigers were better than the 2 teams that played in the BCSNCG. And in retrospect can anybody seriously argue that Ohio State was better than Georgia in 2007 after seeing the ass-whipping LSU put on ’em? If you base it on who really had the best team that year’s BCSNCG should really have been between UGA and LSU, but that couldn’t happen ’cause Ohio State only had 1 loss and UGA and LSU both had 2 losses. Of course, as we now know, tOSU hadn’t really played anybody and was a paper tiger.

  12. Always Someone Else's Fault

    All data is good data. The fault always lies in the interpretation.

    So… why not make sure we have as much data as possible?

    Data points which should absolutely be in the mix:
    Mumme Poll
    Football Outsiders F/+ rankings
    AP poll
    Computer rankings which are willing to have their methodologies vetted
    The Legends Poll, or something along those lines

    I am sure others here have good suggestions, but I am betting that more data points will show clear trends towards consensus. And if the multiple points of view have no consensus, then the panel can point to that and say, “Look, everyone had a different 4, and we went with Clemson because they won their conference” or “we went with Boise St because they were undefeated.” It won’t stop the bickering, but it will be demonstrably true that no real basis for consensus decision was really possible, based on team performance.

    The only way this thing fails is if it starts picking teams well outside the consensus. And the more data points the system embraces, the less likely that happens.

    • Always Someone Else's Fault

      This is why I am firmly against a football version of the RPI (Barnhart is so wrong on this one). A single metric like that encourages everyone to start gaming the system – which is exactly what happens with CBB. A multiplicity of data points, each of which values different variables, all of which are known on the front end (SOS, performance over and under expected value, eye test, whatever), makes such gaming nearly impossible.

      The BCS required such a formula in order to eliminate the human element in the final decision making process. A Football RPI just repeats that mistake if you make it the sole reference point. On the other hand, if Sagarin, the RPI people, etc. want to produce a ranking product and are willing to publish their methodology, then absolutely let them be part of the mix being evaluated by the panel.

      • Always Someone Else's Fault

        And finally (should have thought all this out beforehand), the panel is free to weigh all of these data points as they see fit on a year to year basis. If bloggers have detected a serious flaw or bias in a particular data point, then that bias will be known to the decision-makers. It’s a fluid decision-making process, but it’s flexible.

        This statement will be the death of a 4 team playoff system, making Cojones and ML leap for joy: “Well, we recognize that just about every other ranking system out there had Georgia #4, but we weren’t allowed to consider those. We’re only allowed to consider x, y, and z.”

        • Always Someone Else's Fault

          In other words, I’m way less concerned about the size and constituency of the jury than the jury instructions.

  13. FortWorthDawg

    The bowl selection committee proccess will simply open up controversy for further request for growth of the playoff as year progress. Look what’s happen with basketball. If they couldn’t satifay everyone with 64 teams and had to expand, then how does the FBS think they can get by with just 4.

    I suggest something moderately simple. Keep the bowl system intact! Leave out the championship game and invite ALL the teams to the current bowl system games you currently have in place. You have 4 MAJOR bowls now. Take the winners of the 4 major bowls and send them to the 4 team playoff. I would hope someone would point out the downside to doing a 4-team playoff this way. You would still have your conference tie-ins to the bowl games AND at large births. That way EVERYONE has an EQUAL shot to make it into the playoff thru the bowl games and it doesn’t take away from the regular season.

    Any takers?

    • Mayor of Dawgtown

      That’s an 8 team playoff by another name.

      • Cojones

        It’s similar to what I think is inevitable. Except 8 teams should be borne out of the top 8 bowls. Two games from there is a true NC.

        And there is no jumping for joy. I propose this approach to shortcut the process, not because it’s the sharpest tool in our toolbox. Any proposal is shot full of problems, but when you manufacture one to start with (slip slope) it’s not fair to the discussion. It’s (8 team proposal) out there to be shot down like 16 teams, 4 teams, 2 teams or plus one. In the process you arrive at the best model that we are capable of advocating and that satisfies best a fanbase that wants a playoff.

        I have no earthly idea which one is most correct, just trying to help discussion. Some of you that I agree with most on this blog, but don’t on this proposal, wouldn’t want me to advocate something that didn’t appear best to me, would you? I would then have to call myself No Cojones.

        • Always Someone Else's Fault

          CFB is about the last thing in our culture that (A) matters and (B) most people can argue about without turning into total jerks. All opinions welcome. All opinions fair game. Long live CFB.

        • Cosmic Dawg

          The one thing I don’t think you can do for the playoffs is lead it up to polling or voting or whatever. You really would need the conference champions to be the 8-game playoff major bowl seeds – otherwise, why not just keep doing what we’re doing?!?

          Some of the conferences are weak, true, but that would encourage somewhat strong teams in super competitive conferences to move to those weaker conferences in the hopes of getting a bowl bid, and so maybe you’d see more parity between the conferences?

          • Cojones

            Not unless there is perceived parity in the polls. How many SEC teams do you think would be available in the top 16 teams,seeded and paired in bowl games? I would bet money on at least 3 SEC teams in the 8 winners to vie for the NC. I think that 8 validated teams works most in favor of the SEC.

  14. JasonC

    I wonder if Seth E or another reporter could question Slive about the mumme poll at media days.