Fixing College Football’s Postseason

There’s no good reason the following should be posted here. But what are blogs for if not to share ill-informed opinions and half-baked ideas? And education is one of the stated themes of this blog and this post is about college football, so here goes…

This past Wednesday officials with college football’s Bowl Championship Series (BCS) system (plus the athletic director for independent Notre Dame) met at O’Hare Airport outside Chicago to continue discussions on developing a playoff system that would produce a true national champion (rather than rely so heavily on computer rankings and polls to pick a top two). A four-team model seems likely to prevail, but the question will ultimately be decided by university presidents, perhaps at a June 26th meeting in Washington.

Realizing that it’s late in the game, permit me to present a rather different model that occurred to me while my friend Sam and I were talking about how much we love the Euro 2012 soccer championship.

(And I know that I just lost 95% of any college football fans who just happened to stumble across this and were already wondering what a Pietist Schoolman was and why he/it thought he/it had anything to say about their favorite sport. For the rest of you…)

The best international soccer championships bring together relatively large numbers of teams: 16 of the 53 European national teams are competing in Euro 2012, with that number due to rise to 24 in four years; the World Cup is a bit more exclusive, with only 15% or so of FIFA members making it every four years, but that tournament still includes 32 sides. Rather than following a March Madness-style one-and-done system, teams are initially put into groups for round-robin play: four in each group in the Euros and World Cup, with each side playing games against each of the other three in its pool. The top two teams from each group move on to a knock-out tournament, with the 1st place finisher in each group playing a 2nd place team from another group. This format continues through quarterfinals, to semifinals, and concludes with a final between the two teams left standing.

Alabama-Tennessee, 2011
Would the University of Alabama (here about to host Tennessee last October) have ended up national champs in a playoff system starting with group play? – Creative Commons (André Natta)

So… While there are those who have proposed an eight-team college football playoff system, I don’t think I’ve seen anyone suggest that those semi- or quarter-finalists be chosen not simply by BCS-like rankings, but using something like soccer’s group play….

Just for simplicity’s sake, let’s start by saying that the sixteen highest rated teams (regardless of conference) in the penultimate BCS ranking (i.e., the one that seeds the bowls) make it to group play. In 2011, that would have meant (in order of finish) LSU, Alabama, Oklahoma State, Stanford, Oregon, Arkansas, Boise State, Kansas State, South Carolina, Wisconsin, Virginia Tech, Baylor, Michigan, Oklahoma, Clemson, and Georgia, with Michigan State and TCU just missing the cut.

Now, rather than randomly picking groups of four, the World Cup has something of a seeding process meant to avoid putting (in a worst case scenario) the four best teams in the world all in the same group. Names of countries go in “pots,” with the highest ranked teams in their own pot and then the others clustered geographically.

That would look something like this, in our college football version:

  1. Alabama, LSU, Oklahoma State, Stanford
  2. Boise State, Michigan, Oregon, Wisconsin [a “Northern” pot]
  3. Clemson, Georgia, South Carolina, Virginia Tech [“Eastern”]
  4. Arkansas, Baylor, Kansas State, Oklahoma [“Southern”]

But that doesn’t produce an especially equal division of power: Pot 2 would have three of the BCS top 10, while Pot 3 would have no team higher than #9, plus the last two to make the tournament.

Better, we’d do like the World Cup draw, which puts together diverse geographic regions in order to produce relative equity: Europe has enough teams to merit its own pot, but then (at least in 2010) North and Central America were in with Asia and Oceania, and African entrants were drawn from the same pot as South American ones. So that version would leave us with a slightly different make-up for pots 2-4:

  1. Alabama, LSU, Oklahoma State, Stanford
  2. Clemson, Michigan, Virginia Tech, Wisconsin [ACC, Big 10]
  3. Arkansas, Boise State, Georgia, South Carolina [SEC, Mountain West]
  4. Baylor, Kansas State, Oklahoma, Oregon [Big 12, Pac 12]

You could make a case that Oregon and Boise State should change places, but it’s close enough to let us conduct a highly unsophisticated yet randomized drawing that produces the following groups for our alternate universe finish to the 2011 college football season:

Group A Group B Group C Group D
Boise State Oklahoma Alabama Arkansas
Clemson Oklahoma State Baylor Oregon
Kansas State South Carolina Georgia Stanford
LSU Virginia Tech Michigan Wisconsin

And now we’re on to something! Stanford and Oklahoma each getting chances to avenge embarrassing conference losses (to Oregon and Oklahoma State, respectively)! A ‘Bama-UGA game that never happened in the SEC! Robert Griffin III vs. Denard Robinson! And Boise State in a universe where a last-second missed field goal didn’t deny them a chance to play a team like LSU!

Andrew Luck, 2010
Stanford QB Andrew Luck after a November 2010 game – Creative Commons (John Martinez Pavigla)

I’ll let readers imagine what the knock-out rounds would look like. (Have the winner of Group A play the runner-up in Group B and the winner of C play the D runner-up in one half of the bracket, then B1 vs. A2 and D1 vs. C2 in the other half.) Going purely by BCS rankings, you’d expect these quarterfinals: LSU vs. South Carolina, Alabama vs. Oregon, Oklahoma State vs. Boise State, and Stanford vs. Baylor (in which Andrew Luck and RG3 would have had the chance to quarterback against each other before being selected 1-2 in the NFL Draft).

But as keen-eyed college presidents will have noted, this means that the two finalists would play six games (three round-robin, plus quarters, semis, and finals) on top of a regular season schedule already at 12-13 games, and these being student-athletes, we can’t take away from their December studies, or risk injury, or anything like that. We’re not the NFL, after all. So…

Let’s add a scheduling revolution and mandate that the college football schedule begins on Labor Day weekend with all teams going straight into conference play.

While we’d lose the occasional non-conference bowl preview, we’d also say good riddance to predetermined blowouts pitting top tier teams against those a division or two below them (scheduled primarily as glorified, record-padding practices for the marquee squads and as money-makers for the Louisiana-Lafayettes of the world). And don’t you think some network would pay crazy money to broadcast Michigan-Ohio State on an NFL-less Labor Day?

This would also have the added virtue of forcing Notre Dame to join the Big 10 or another conference.

So you’d have, roughly, an eight-game schedule in September and October, with conference records then being the sole determinant of who moves on to the sixteen-team tournament. If conferences wanted to keep their lucrative championship games, that’s fine: the stakes would be high for weaker conferences with only one spot in the tournament; for those at or near the caliber of the SEC, it would just affect seeding.

Sure, it would take much haggling to determine how many spots go to each conference, but (a) the soccer world manages to figure this out on a continental scale, and (b) we’re moving towards having only a handful of super-conferences anyway.

For the most part, we’ll just say that this model of the tournament will have the same conference representation as the first version and give five spots to the SEC, four to the Big 12, two each to the Big 10 and Pac 12-2, and one to the Mountain West, but those spots will be allocated purely on the basis of conference record. The top seeds (in their own, separate pot for the draw) would be the champions of the four conferences rated highest by Jeff Sagarin’s computer. For 2011, that was the Big 12, SEC, Big 10, and the Pac 12. And then the Big East — #5 in Sagarin’s ratings — would get one of the two spots previously given to the ACC — #7, behind the independents.

This model produces the following changes:

  • The SEC would have the same five teams, but Alabama would be replaced in Pot One by Big 10 champion Wisconsin.
  • The team the Badgers narrowly defeated in their conference’s inaugural championship game, Michigan State (7-1 in conference), would replace its arch-rival, Michigan (6-2), in the tournament.
  • It would still be Stanford and Oregon representing the Pac 12, but the Ducks would replace Stanford in Pot One (fittingly, having blown them out in the regular season).
  • Boise State (6-1) would not be rewarded for scheduling a tough non-conference schedule, since that missed FG did mean that TCU (7-0) won the Mountain West.
  • Virginia Tech’s (7-1) loss to Clemson (6-2) in the ACC conference championship game would cost it a spot in the tournament.
  • West Virginia Football Fans, 2011
    West Virginia fans in September 2011, who surely would have advocated for this playoff system as their team’s season played out – Creative Commons (mt_t_77)

    Then the Big East muddies the waters: Cincinnati, Louisville, and West Virginia tied at 5-2, and each was 1-1 against the other two. And there was no conference championship game. We’ll go with West Virginia, since it was the BCS representative for the conference.

With no changes for the Big 12 teams, that would leave these four pots:

  1. LSU, Oklahoma State, Oregon, Wisconsin
  2. Alabama, Arkansas, Georgia, South Carolina [SEC]
  3. Baylor, Kansas State, Oklahoma, TCU [Big 12, plus the Texas-based Mountain West champ for geographic proximity]
  4. Clemson, Michigan State, Stanford, West Virginia [everyone else]

And that produced the following draw:

Group A Group B Group C Group D
Baylor Kansas State Arkansas Alabama
Georgia Michigan State Oklahoma Clemson
LSU South Carolina Oregon Oklahoma State
West Virginia Wisconsin Stanford TCU

You still get the Oregon-Stanford rematch (plus one for Michigan State vs. Wisconsin), but this draw doesn’t seem quite as compelling as the first one. Though a Wisconsin-LSU quarterfinal would have been the quintessential unstoppable force (44.1 point per game Badger offense) vs. unmovable object (a Tigers defense that allowed just 10.5 PPG). And moving the Oklahoma-Oklahoma State derby (to borrow more soccer lingo) from group play to a winner-take-all round might have been worth it as well.

In any case, it would leave the finalists playing 14-16 games, finishing on New Year’s (preferably in the Rose Bowl).

September 1-3, 2012: the regular season starts on Labor Day
October 26-27: allowing for a bye week in many conferences, the regular season ends after eight games
November 2-3: conference championship weekend (or a week off for those without such games)
November 9-24: group play, with the pivotal third games happening the Friday-Saturday after Thanksgiving!
December 7-8: after a week off, we continue with quarterfinals
December 15: Semifinal Saturday, then two weeks off for exams and holidays before wrapping up with...
January 1, 2013: college football's championship game restored to its proper spot on the calendar

Lots of details still to be worked out… For example: where would the games take place? I’m guessing BCS bowl sites, to keep those folks happy (since they’re losing their January paydays), with Friday and Saturday games to avoid running into the NFL’s schedule. Though I’m not sure why schools would ever agree to give up the revenue of hosting a couple of non-conference games…

And I don’t know what happens to the other 100+ teams in what we’ll now call the Premiership (rather than the Football Bowl Subdivision) of college football. Maybe a second tournament… Relegation and promotion?

3 thoughts on “Fixing College Football’s Postseason

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.