≡ Menu

I’ve got a real post ready for tomorrow, so Thursday is Rant Day this week. There’s been a lot of talk this week about how this would be a great year to implement the four-team playoff that’s coming to college football for the 2014 season. I suspect part of that reason is because people are recognizing that Florida boasts a more impressive resume than either Alabama or Georgia, and that Oregon is an elite team that lost an overtime game to a division rival. Therefore, a four-team playoff featuring Notre Dame, the eventual SEC Champion, Florida, and Oregon would be great.

On one hand, of course it would be great. I’d love to watch Notre Dame face Oregon and perhaps see Alabama and Florida face off (no, you couldn’t make me watch Florida-Georgia II). But that’s only because college football is great. In fact, this year — like almost every year — stands as a good example of why the four-team playoff system is doomed to create controversy and do little to increase fairness. Let’s start with…

Winning your division = Bad

We saw this last year, when Alabama ended the year as BCS #2 and therefore had locked up its spot in the national championship game, while LSU still had to go play in the SEC Championship Game. In this particular case, I don’t feel too bad about the fact that Florida would be given a free pass to the four-team playoff while Georgia has to go play Alabama, but only because Florida has faced a much tougher schedule.

But what if instead of what actually happened — Florida beating South Carolina, South Carolina beating Georgia, and Georgia beating Florida — the order was reversed, and Georgia beat South Carolina, South Carolina beat Florida, and Florida beat Georgia? In that case, Florida would likely be #2 or #3 in the BCS, while Georgia would sit pretty at 4. And Florida would go play Alabama for the right to win the SEC… with the loser being left out of the playoffs. That’s patently unfair. Being the third best team in your conference could be preferable to being the second best team. This is a lock to happen at some point during a four-team playoff. This is also going to be a bigger problem generally as conferences get bigger, because conference schedules will become unbalanced. The Big 12 has a round robin where everyone plays everyone, but in a 14-team conference, you can easily see a better team end up with a worse record than an inferior team due purely to scheduling.


It is being conveniently forgotten that Stanford actually won the Pac-12 North. In fact, we have an exact reversal of what happened last year, which many screamed was unfair to Oregon. Lest you forget….

In 2011, both Oregon and Stanford went 8-1 in the Pac-12. Oregon beat Stanford head-to-head in Palo Alto (but lost to USC), so they won the division and then the conference. Stanford had a soft nonconference schedule while Oregon traveled to Dallas to play and lose to LSU in the season opener. So Oregon was 11-2 (but Pac-12 champs) while Stanford was 11-1. After the regular season, Stanford was 4th and Oregon was 5th in the AP, Harris, and Coaches Polls, and also in the BCS. This struck many as unfair, because the Ducks were essentially the better team but had a brutally hard non-conference game, a de facto road game against the #1 team in the country.

Fast forward to 2012. Oregon and Stanford both went 8-1 in the Pac-12. This time, Stanford beat Oregon head-to-head — in Eugene — so they won the division and perhaps the conference (we’ll find out on Saturday.) Oregon had a soft nonconference schedule while Stanford went to South Bend to play the #1 team in the country. Stanford will finish 11-2 if they beat UCLA, while Oregon finished 11-1.

Should Oregon go to the hypothetical four-team playoff instead of Stanford? If not, why not?

Three 11-1 teams in one division – AKA the 2008 Big 12 South

In 2008, Texas, Texas Tech, and Oklahoma all went 10-0 against the rest of college football and 1-1 in their three-team round robin. The same scenario nearly presented itself this year in the SEC East: in fact, had South Carolina defeated LSU, all three teams in the SEC East would have been 11-1 with wins over each other (it was this precise fact that caused Georgia to vault Florida, as without USC in the mix, the Bulldogs held the tiebreaker over Florida). How do you leave one or two of these teams out?

If the 2012 Gators were more like the 2011 Gators, the problem would have simply emerged in the SEC West, instead, as Alabama, Texas A&M, and LSU would have all been 11-1 with wins only over each other. This scenario is a lock to emerge at some point, and will cause significant controversy.

Oregon-Kansas State

No formal criteria exist to guide the pollsters, and no one knows what factors a selection committee will choose to weigh most. But why would Oregon get a spot over Kansas State? There are compelling reasons to argue for both.

If you would choose Oregon over Kansas State, chances are you heavily support one of these factors:

— The “eye” test; Oregon looks like a better team
— Favoring the “best” team over the “most deserving” team; Oregon ranks 1st in the SRS and 2nd in Jeff Sagarin’s Pure Predictor Ratings
— The best “worst game” – the Ducks lost in overtime to a very good team
— Pre-season polls/inertia: Oregon started the year 5th, Kansas State 22nd

Those are all legitimate ways to break a tie among one-loss teams. But so are these, which all favor Kansas State:

— Strength of schedule: Kansas State has the harder SOS in both the SRS and Sagarin’s ratings, and that’s before including the Texas game
— Conference champs: KSU will win the Big 12 if the Wildcats defeat Texas; Oregon won’t even win their division
— Best win: Kansas State won in Norman against a very good Oklahoma team; Oregon’s best win was against…. Oregon State

Oregon is the winner of the beauty pageant, but Kansas State’s resume is at least as good as Oregon’s. Perhaps you want to drop Kansas State for its awful showing against Baylor, but I don’t think the criteria of “best ‘worst loss'” is going to be consistently applied. The simple fact is if Kansas State was left out in favor of Oregon, Wildcats fans would be correct in feeling screwed. Had Florida State defeated Florida this weekend, they would probably have vaulted both Oregon and Kansas State, but would that make any more sense? How would we pick one out of FSU (conference champs but weakest schedule), KSU (conference champs but worst loss), and Oregon (best team but not even division champs)?

If you wanted to heavy on conference winners, you could pick Notre Dame, the SEC Champion, Kansas State and Stanford, assuming the latter two win this weekend. If you wanted to pick based simply on the best teams — i.e., the Vegas test or the SRS/predictive method — you’d pick Notre Dame, Alabama, Oregon, and … maybe Kansas State, maybe Florida, maybe Georgia. If you wanted to go based on resume, it would be Notre Dame, Florida, the SEC Champion, and Kansas State. If you wanted to go popularity test, it would be Notre Dame, Florida, the SEC Champ, and Oregon.

What if Ohio State was eligible this year? They would join the SEC Champ and Notre Dame, but picking that fourth team would be tough. At that point we would start looking at Florida’s warts. Does it matter that they scored only one offensive touchdown against an FCS school and won 23-0? That against a Sun Belt team, the Gators trailed by a touchdown with under two minutes left? That they were tied with a bad Bowling Green team midway through the third quarter of a game? That at home against Missouri they were tied entering the 4th quarter?

How much does the “eye test” count here? What’s worse — losing in overtime to Stanford or winning only after trailing with 2 minutes left against Louisiana-Lafayette? What’s better, winning in College Station by 3 points or winning in Corvalis by 24? Does it matter that we already have one SEC team in the picture — should that count when deciding who gets the fourth spot?

A four team playoff is great for college football because it gives us three times as many playoff games. But it’s going to do almost nothing to eliminate the controversy and bias that’s always been a part of the sport. No real solutions here — that’s for Saturday — but just a rant that picking four teams is going to be as silly and arbitrary most years as picking two teams.

  • At some point, the debate lightens up, though. Two is generally better than four, because the “best” team can fall just outside the top two based on a close loss, schedules, etc. When I looked back at history just using SRS, though, the “best” team by that measure is usually in the Top 6. My playoff method of choice would be six teams, four conference champions or three plus Notre Dame if ranked higher than fourth champion, two at-larges, where #1 and #2 get a bye and must be conference champs (or Notre Dame/other independent if top two).

    Applying it this year, we would get:

    1. Notre Dame
    2. SEC Champion (Alabama or Georgia)
    3. Florida (at large #1)
    4. Oregon (at large #2)
    5. Kansas State (if win Big XII)
    6. Stanford (if win Pac-12)

    Waiting in the wings are others, like Oklahoma if Kansas State loses and they win, or Florida State, or the Big Ten Champion or Big East Champ.

    Last year, it would have been

    1. LSU
    2. Oklahoma State
    3. Alabama (at large)
    4. Stanford (at large)
    5. Oregon
    6. Wisconsin

    By making it six with conference champ requirements, you insure that the Top 3 teams always make it, Top 4 almost always, and that winning conference championships matter. Boise State would have missed out last year because TCU beat them, but the opportunity would be there.

    • Chase Stuart

      I agree 6 is better than 4, which is better than 2. I have a post on Saturday describing my preferred playoff format, which is basically just your idea that I stole, IIRC.

      The wrench this year would be an eligible Ohio State in a 6-team format. They would knock out Stanford, I think, which would lead to the question of why was Oregon going ahead of Stanford. Six does feel a lot better than four, though.

  • George

    I know someone else has suggested something like this at some point (it is referred to as a suggestion from elsewhere in Mathletics but slightly differently) but I think for the cleanest way of doing it to avoid bye’s etc. and so everyone would essentially have to play the same number of games (excluding conference championship games); take the top 8 (in an 8 plays 1, 7 plays 2 type of format) in effectively a quarter final, re-grade everyone after that round (if the result actually caused a change in ranking given the teams remaining) – 4 plays 1, 3 plays 2 in the semi and then the winners of those games play off for the National Title.

    Basically you’d have an incentive to take big games rather than padding your schedule out to get strength of schedule benefits in some of the computer rankings to ensure they get a top 8 spot, potentially you wouldn’t get teams resting starters at the end of the first half (as likely the more higher ranking teams would take soft games out of the schedule if they had the opportunity to) and by taking the top 8 you would leave the door partially open the team catching the wrong end of an Oregon/Stanford kind of situation. Also when has a team outside of the top 8 been talked of as a potential National Champion (admittedly in the above scenario you would have Stanford in and Texas A&M out but could you argue either should be National Champions)?

  • Andrew

    No matter how you do a playoff system, someone is going to be upset. People even get upset about the NFL playoffs because division winners from bad divisions get in. I would actually like to see an 8 or 16 team playoff system just entirely replace bowl games, where all conference champs get in and all other spots are determined strictly by record. No more of these BCS computer rankings and coaches’ polls, just a straight up did-you-win-enough-games system. No frills, no fuss. he best team probably won’t win most of the time, but it will be one of the best teams and they’ll all have a shot at it for once. And yes, this would probably necessitate some governing body that would set schedules to prevent teams from just scheduling the 12 worst teams they can think of and going to the playoffs. And, because it would be run by the NCAA, it would probably be quite corrupt. but I still think it would be better than the BCS. And that’s really all anybody can hope for.