The Last Word on a College Football Playoff System, Part I: Potential Systems Explained

Our long national nightmare is almost over. The annual BCS meetings are this week, and the assembled commissioners are almost certain to institute some sort of playoff system. By the end of this week, we may be saying goodbye to the debate over a college football playoff, or at least a particular stage of it.

Everybody knows one thing: they hate the BCS. What people can’t seem to agree on is how to fix it. One thing the BCS has brought us is a system that seemingly every year finds a new way to screw things up, and that means part of the problem with settling on a playoff format is that each year seems to support a different system, which often wouldn’t work so well in another year. Fortunately, that also means it’s also given us a lot of scenarios to examine and test the equally numerous playoff proposals. In honor of the end of this era, that’s what we’ll be doing this week, determining how they would have gotten rid of controversy (or not) and on the other hand, still maintained the sanctity of the regular season. Are BCS proponents correct that no system is perfect and so none can sufficiently improve on the BCS? (Note that I originally wrote much of this in 2009 as part of my “Evolving Take on the Debate on a College Football Playoff”, and as such most of the proposed systems are from before then with some having broken links, and you should probably read
yourself for many of my rulings to make sense.) Here are the various playoff systems supported in various places around the Internet and their backers:

The Big 5

  • Actual BCS system. The system we all know and loathe.
  • Plus-One with traditional bowls. What the BCS honchos are calling the “original ‘plus-one'”. Effectively, goes back to the old bowl system, but adds an additional postseason “week” after the bowls, selecting the two best teams in the country. Plus-ones are embraced primarily by people who don’t really support a playoff but aren’t satisfied with the result the BCS provides (or, reportedly, the BCS honchos when they’re not worrying about public opinion). This one is proposed by Brian Sakowski and… that’s it (although Frank the Tank flirted with it once). No one else needs to be told what a disaster this one would be, and not just because it would reduce the bowls to play-in games. It’s almost untestable because there are almost as many “traditional bowl lineups” as people proposing them (the Big 12 is almost as young as the BCS itself) and the ripple effects on the rest of the bowl system are almost unknowable. But plus-one after the current bowl system, as supported by no one and tested in place of this by Ed Gunther, is even worse, because it creates the most skewed bracket ever, pitting 1 vs. 2 in the first round, rendering anticlimactic at best and utterly ruining at worst classic BCS outcomes like Texas-USC in 2005. People who have proposed variants of this – including Jon Miller of – often have specific years in mind, like 2003, when they wanted LSU to take on USC for the national championship.
  • Plus-One top 4 bracketed. Also known as the “four-team event” considered by the BCS honchos. A simple four-team playoff. Basically further mangles the bowl lineup by plucking two more teams into semifinal bowls, and turns out to produce first-round rematches often if left unchecked – often enough Gunther considers it inferior to plus-one after the current bowl system, obvious nonsense. As tested below, if 1/4 and 2/3 matchups would produce rematches, they are switched to 1/3 and 2/4. This system and variants of it (often giving home-field advantage in the first round) are backed by Richard Cirminiello of CFN, Tim Layden in SI in 2001, Frank the Tank, and Ben Prather, proprietor of SBNation’s former “BCS Evolution” blog. Jerry Hinnen has the semifinals the week before students’ finals week instead of being part of the bowls, allowing semifinal losers to join a pool with the teams at 5-10 in the other BCS bowls, and seems to be confused because he wants to include “the teams that need to be in” and includes the mid-majors that won BCS bowls, but those teams placed 8th at best in the BCS before the bowls, so the only thing I can think of is he intended to give auto bids to teams that go undefeated, but that includes some teams that didn’t “need” to be in like 2007 Hawaii and 2008 Boise State… Mark Schlabach and (if they were to firmly support a playoff) EDSBS want a plus-one but don’t give details, so I classify them here.
  • 8-team playoff, BCS champion auto bids. The two eight-team systems are often backed by people who want to use the BCS bowls as quarterfinals, ignoring how that would reduce the bowls to play-in games, or people who aren’t satisfied with a plus-one but aren’t ready to embrace 11/5. This playoff allocates six spots to the champions of BCS conferences, leaving two at larges. I could call it a “6/2” system, mirroring my “11/5” terminology. Fundamentally the system proposed by the Mountain West in 2009, and originally backed by Matt Hinton (now Yahoo’s “Dr. Saturday”) to the extent he supports any specific proposal, Pat Forde, this AP simulated system, BCS Watch to the extent he supports any playoff, James Irvine, Vincent Ellerby, and Frank the Tank‘s earliest proposal (the last two attempting to maintain traditional bowl assignments in the quarterfinals). Many proposals of this system replace one of the at-larges with an auto bid for a single non-BCS team, as in the system used in College Football News’ simulated playoff and backed by their Pete Fiutak, and also proposed by (which also takes away the Big East’s BCS auto bid). (CFN teamed up with WhatIfSports for December Madness after the 2007 season.)
  • 8-team playoff, no auto bid. Just the best eight teams in the country. Supported by Gene Wojcechowski, Matt Starnes, and Stephen Carradini. President Obama famously proposed an 8-team playoff during the campaign but didn’t specify how teams would be selected (though his original comments to Chris Berman suggested the no-auto-bid approach), putting him in the company of CFN’s Matthew Zemek, Paola Boivin, and USA Today‘s editorial board.
  • 11/5 system. Also known as “16-team playoff, auto bids to all conference champions.” Supported by Yahoo Sports’ Dan Wetzel, Sloppy Joe of College Football Cafeteria, CBS’ J. Darin Darst, Fox’s Peter Schrager,, Eric B. Shaw, Mark Blankenbaker,, and yours truly, as well as all right-thinking people who pondered the college football playoff debate, at least before realignment set in (Wetzel, the most prominent voice for this model, explains his newfound misgivings about it and conversion to a BCS-auto-bid 8-teamer towards the end of this article) – yet seemingly treated as a strawman and untested by Gunther (on the grounds that one of the above four will happen before we get this). For the past few seasons, has held its “December Madness” tournament under 11/5 rules and simulating the results with its game simulation technology; I did the same with my own bracket for several years, using the same site.

Proposals That Don’t Fit In the Big 5

  • “Flex” playoff system, Zane version. Basically, a system that tries to always be the best system, sometimes being a single national championship game, sometimes being Plus-One top 4 bracketed. Here’s the simplest I can explain Billy Zane’s idea: If there are two undefeated teams ranked 1 and 2 and no one else above a certain threshold, or two or fewer teams with one or fewer losses in the top four, they play a single national championship game. (If an undefeated team plays a 2-loss team in this manner, the national championship game is not necessarily for the national championship.) If there is one undefeated team at #1 and two 1-loss teams (or a 1-loss team and an unbeaten that’s not #2) in the top four, the undefeated team gets a bye into the national championship game and the two one-loss teams play each other. Otherwise, it is Plus-One top 4 bracketed, except any unbeaten over said threshold automatically gets in (top 5 as proposed in the demonstration, but top 6 as I analyze it below, for reasons that will become clear), bumping out teams with losses if needed.
  • “Flex” playoff system, Prather version. Link is the same as Prather’s support for Plus-One top 4 bracketed, which seems to be more recently supported, an admission of the complexity of this plan. Similar to the Zane version, except the size of the gaps in the BCS standings (equivalent to a 1.5-spot average difference in the polls) is used to select the qualifying teams, and the result could be as large as an eight-team playoff, larger if there are more undefeated teams than that.
  • Plus-One top 4 bracketed, preserve Rose Bowl. This is a compromise between “Plus-One with traditional bowls” and “Plus-One top 4 bracketed” by preserving the only traditional bowl matchup worth preserving, similar to Frank the Tank‘s “semi-seeded plus-one”, and a reformation of the Big Ten’s “Four-Team Plus” model under consideration (essentially a bracketed plus-one that ignores the Big Ten and Pac-12 champions for the semifinals only, considered by Frank the Tank here). Under this model, even if the Pac-10 or Big 10 champion is in the top four, and even if the other one isn’t, they will play each other in the Rose Bowl no matter what. Curiously, this often duplicates the results of plus-one after the current bowl system. For example, in 2003, USC would have been force-seeded to the Rose Bowl, leaving LSU to play Oklahoma. Under this system, non-Rose bowls are listed as the higher-seed’s tie-in bowl, unless the higher seed is the Big East champion or (somehow) an at-large.
  • Plus-One top 4 bracketed, conference champions only. An attempt to avert situations where a normal bracketed plus-one renders conference championships meaningless. Apparently this one was considered by the suits in the room but ultimately rejected.
  • 6-team playoff.
    Brian of MGOBlog wants no autobid, home field, 3 seed picks whether they face 5 or 6, 1 seed picks which first-round winner they face. Ryan Murphy gives auto bids to the BCS conferences that aren’t the Big East; that version will only be assessed for years after the ACC raided the Big East.
  • 8-team playoff, conference champions qualify based on ranking.
    Frank Xei (at least I’m assuming that’s his name) proposes putting all conference champions in the top 12 of the BCS standings in the tournament, then the remaining spots go to the top non-conference champions. As you’ll see, this will sometimes exclude BCS conference champions, which is “some” times too many for the BCS gatekeepers. Ditto for Sloppy Joe‘s more mature playoff proposal, which takes all conference champions in the top 14 and Notre Dame if they crack the top 8. After originally proposing the original BCS conference champion auto bid system, Ty Duffy subsequently suggested a system with the top six highest-ranked conference champions, regardless of conference, selected.
  • 8-team playoff, conference champions qualify based on number of wins. As suggested here, conference champions would be required to have nine wins over FBS teams, but would go in automatically regardless of ranking if they had that many.
  • 8-team playoff, conference champions only. Proposed by Arizona State president Michael Crow. Chris Suellentrop‘s system would also include BCS auto bids, as would Jason Nafziger.
  • 10-team playoff. Used by ESPN c. 2009 (no auto bid) and by College Football Campus (BCS champion auto bids) for their simulated playoffs, and suggested by (current BCS automatic qualification requirements for auto bids), Dr. Saturday (BCS champion auto bids, two-team-per-conference rule remains in effect), and Brandon E. Kennedy (BCS champion auto bids), mostly because it’s a nice round number we’re all familiar with, despite the fact it’s arbitrary (our counting system would make more sense if we had six fingers on each hand) and produces an ugly bracket. One advantage of the format is that it’s possible to give automatic byes to the BCS conference champions, no more, no less.
  • 12-team playoff.
    Ryan West backs giving the BCS conference champions auto bids. (West also proposes holding each round at a BCS bowl site, which’ll never happen.) does not, and includes a 16-team NIT-like tournament, while thinking “Bowl Tournament” isn’t an oxymoron from the bowls’ perspective. Jonathan West (who I don’t believe is related to Ryan) also backs an auto-bid-free format. The Enhanced Bowl Season gives auto bids to the BCS conference champions and one non-BCS champion, plus any other non-BCS champion in the top 12 of the rankings.
  • 16-team playoff, only BCS conferences get auto bids. As proposed by ESPN blogger Ghostsof1948, Russ Thorson, and Jeb Puryear.
  • 16-team playoff, auto bids for champions of top 8 conferences and BCS conference runner ups. This is what Bill Hahn‘s bizarre proposal amounts to. Even more bizarre, seeding is random, with the conference champions randomly seeded 1-8 and the runner ups randomly seeded on the opposite side of the draw as their respective champions.
  • 16-team playoff, auto bids for qualifying conference champions. This is the Mountain West’s 2011 proposal, with the qualification being that the team must be in the top 30 teams in the country.
  • 16-team playoff, no auto bid. The basis for SI’s simulated playoff, and also proposed by Sam Matta, Chad Crabtree, and “Tommy” on The Right Sphere. John (?) Houlgate would pick the 16th team randomly from the teams ranked 16-25, obviously untestable.
  • 20-team playoff. Tyler West essentially adds four play-in games to the 11/5 system. He also suggests making every conference get a championship game, but the system is testable without it. He also has an odd rule that teams in position to get an at-large before playing the conference championship can’t be left out in favor of a team from the same conference, which seems to have the effect of reducing the incentive to win the conference championship.
  • 24-team playoff. Simple: there are over 100 teams in college football, and over 300 teams in college basketball. There are 11 conferences in college football, and 31 conferences in college basketball. There are 12 games in a college football season, and 30-some games in a college basketball season. Therefore a college football season should have roughly a third of the teams in March Madness, and that adds up to 22-23. 20-team proposals aside, 24 is the nearest number that produces the neatest-looking bracket. Whether it preserves the sanctity of the regular season is another matter… I want to keep my level of work sane, the BCS standings a useful baseline, and the regular season with a modicum of meaning, so this is the largest playoff I’m willing to consider.

Proposals Too Radically Different To Be Tested (Or Ever Pass)

  • 16-team playoff, only winners of conference championship games get auto bids. As proposed by Bruce Leban. Why is it untestable? Because conferences would race to adopt championship games somehow, someway, forcing realignment and turning this into a variant of one of the below.
  • 4-team playoff in place of conference championships. This is essentially what Mark Cuban has proposed, and it’s untestable for the opposite reason: it effectively forces conferences with championships to get rid of them.
  • Gunther Modified Season. Teams play only 8-9 regular season games, most of them conference games. Then a selection committee divides teams into two groups, with the top 32 teams in Group A and everyone else in Group B. Group B teams play each other the first week of November, ideally teams within 10 ranking spots of them with home field based on attendance rankings, Group A teams the second week with home field to higher seed. Then everyone is reseeded and Group A is condensed to the top 16 and everyone plays the third week, and then Group A is condensed to the top 8 and everyone plays the fourth week, and then Group A is condensed one more time to the top 4 and intra-conference matches are allowed there, though not rematches. Finally bowl bids are divvied out and #1 and #2 play in the championship game. The “playoff” in Group A is not a traditional single-elimination tournament, as teams can lose early on and remain in Group A, but conversely a team can win and still not stay alive; one commenter proposed fixing this by first having the field cut from 28 to 18, with 14 winners and 4 losers, then to 12 similarly, then 8, then cut in half each subsequent round; I would go 24 to 16 to 12 to 8. Gunther claims the basic framework is “workable and realistic” as opposed to some other plans both above and below, but no one will accept such a major change in a million years – it changes too much tradition, like November rivalry games, and it’s too complicated for Americans – even though it’s more conservative than its cousin…
  • Johnson Swiss System/32-cum-8-team playoff. The Swiss system is the system used in chess tournaments, and it basically boils down to always facing someone with the same record as you. If you win your first game, you face someone else who won their first game. If you lose that game, you face someone who’s also 1-1, and so on. Ben Johnson’s postseason (leaving his conspiratorial thinking aside) would work like that. It would realign all the BCS conferences, as well as a conference taking the best teams from the Mountain West and WAC, so that each conference had a uniform number of teams, and would have the first eight games all be conference games. The top four teams from each conference would play a mini-bracket leading to a conference championship, with the losers of the conference semifinals playing in a conference third-place game (if they haven’t played already in Johnson’s current proposal). The conference champions, as well as the winner of a game between the MAC and C-USA champions, would form an eight-team bracket, but as originally proposed, losers of each game would face losers of a comparable game, so quarterfinal losers would face other quarterfinal losers, semifinal losers would meet in a third-place bowl, and the winners of quarterfinal loser games would meet in a third bowl, the losers in a fourth. As originally proposed, a similar process would be followed for the losers of the conference championship game and conference third-place game participants. The conference 5-8 teams (3-4 for MAC and C-USA) would enter a similar “Holiday” bracket, the conference 9-12 teams (5-6 for MAC and C-USA) would enter another similar “NIT” bracket, and the remaining 24 teams (would these all be MAC, C-USA, and Sun Belt teams?) would enter some sort of “Sportsman’s” bracket. As presently proposed, at least last I checked, Johnson seems to have taken a cue from Gunther’s system, as teams that weren’t in the main championship bracket would play primarily regional games, an attempt to mollify people who would protest that people couldn’t possibly move from game to game, although the Swiss system seems to still be in effect. In any case no one will agree to the major restructuring of college football conferences required by the Johnson system, or the loss of control over non-conference schedules, or the destruction of interconference rivalries, or the movement of “rivalry week” to week 8 at the latest, or how, comments that this system gives every team a shot at the championship notwithstanding, Sun Belt teams are supposed to ever have a shot at the championship (or in the case of the original proposal, even the non-Sportsman brackets)…
  • Realignment. Hunter Ansley of proposes realigning FBS into eight 12-team conferences, divided into two leagues, which would play a 16-team playoff, with conference championship games, two league championship rounds, and a game between the champions of each league, with bowls serving as rounds of the tournament. Two words: Pipe dream. The proprietor of the Get the Picture blog seems to want a playoff comprised solely of all the conference champions but would rejigger the conferences to create a competitive balance, or the appearance of one. At the height of the first round of conference realignment in 2010, several people entertained dreams of the power teams forming four 16-team superconferences that could then set up a de facto four-team playoff. Yours truly proposed blowing up the current conference system in favor of instituting a system of promotion and relegation last year, and weaker forms of pro/rel have also been proposed. Most realignment plans just shuffle teams around into new alignments without regard for whether the conferences or teams would agree or even the impact on other sports. The way realignment has actually played out, especially the first round in 2010 when the Pac-10 very nearly became the Pac-16, underscores the inherent unpredictability of the enterprise.

Tune in next time, when we take a look at how each of these systems would have done each year of the BCS era.

Leave a Comment