College Football Playoff Proposal
- Four Team playoff, single-elimination
- One auto-bid, three selected by Colley Matrix Rank
- Auto-bid goes to the champion of the conference of the previous season's National Champion, unless that team does not belong to a conference, in which case Colley Rank selects all four participants.
- Teams eligible for the four-team playoff are conference champions and independents. All others are ineligible.
My proposal for the four-team College Football Playoffs is to select from among eligible teams using the Colley Matrix to determine seeding. To be eligible, conference teams must win their conference. Independents are always eligible. The top four eligible teams by the Colley Matrix are selected to the field and seeded in the same order.
There is no need for a selection committee with the naturally human concerns of the distribution of the spoils. No need to worry about affiliations or realignments or the impact to recruiting and finances. We have a well-known algorithm that beautifully emulates over-all standings. It has no bias. It is not subject to kick-backs or realignments or death threats or who got snubbed last year. It is beautifully simple in the way that it emulates standings.
The only exception to eligibility is that the conference of the previous year's champion gets an automatic selection to the playoff. The Colley Matrix is used to select the other three qualifiers. As an example, if Alabama won the NCG last year, then the SEC gets the auto-bid. If an independent is the defending champion, then there is no automatic qualifier and the Colley Matrix selects all four qualifiers. The benefits of this rule are:
- provides incentive for joining a conference (as independents can never have the auto-bid),
- the auto-bid conference is identified as of "championship-caliber" and so can tolerate multiple in-conference losses without fear of falling too far in the rankings because those losses came against championship-caliber teams, and
- new champions will have successfully unseated the reigning championship-caliber conference in a playoff format.
Why the Colley Matrix?
We all say we want it "proven on the field" but it seems to me that at every step of the way, someone is trying to justify overriding on-field results - eyeball tests, margin of victory, nasty losses, struggles at home vs inferior opponents. If you support selection criteria that considers these other factors, you are minimizing the only thing that matters in playoffs and conference standings: the win.
The Colley Matrix is a well-known, easily repeatable (and, so, not subject to tampering or bias) computer ranking algorithm that considers only a team's record and the strength of its opponents. And the strength of opponents is measured only by their own record. It is the only BCS computer ranking system who's algorithm is completely transparent. In a nutshell, it makes the assumption that all teams start out equal strength. Then it applies the team's record against the strengths of its opponents to recalculate that team's own strength. The process iterates until it converges to an ordered rank of team strengths.
In effect, this process is an extrapolation of "standings" applied across the country as if every team were in the same conference. If you apply the algorithm only to teams in a single conference, the results would match conference standings exactly. We can use this algorithm to determine who belongs in the four-team playoff. The beauty of it is in its simplicity. Just like a playoff win, it doesn't care if a team won ugly or if it dominated.
Colley in the Conference Championships
Now, I can take this thought a step further and advocate the use of the Colley Matrix for selecting conference championship matchups. In the case of a complete round-robin conference, the Colley Matrix can be used to resolve ties. In the case of a conference with divisions in which entire conference record counts towards the standings and some teams play the best of the other division while others play the worst, Colley is more fair than standings. It considers the strengths of the cross-divisional games. This eliminates a Conference Championship Game matching the winner of a strong division with a much less qualified other-division winner.
I would even go so far as to say there is no need for divisional alignments (a new concept in college football anyway). Conferences would simply create the schedule without worry of maintaining traditional rivalries while playing everyone in the division. At the end of the season, use the top two conference teams as determined by the Colley Matrix to set the conference championship game participants.
Why the Conference Champion Requirement?
This method supports the view that the conference tournament IS the first round of the playoffs. College football does not have the luxury of games to support a pre-season tournament, followed by a double-round-robin conference tournament, followed by a single-elim conference tournament, followed by a 64-team invitational tournament.
Yet, conference tournaments are the closest we have to a complete tournament of all 120-odd teams in the FBS, with the conferences serving as the brackets. Yes, some brackets are stronger than others, but they ARE brackets and surviving is a great testament to the winner's fortitude. The path is clear to all teams: winning your conference is the only way in. There are no style points. There are no second chances. At some point, we must demand that merit override the "eyeball test." A team may look dominant to a selection committee, but the best team doesn't always win. With this rule, if they don't win their conference, then they don't qualify.
And if the Colley Matrix is used to select conference championship game participants, conferences are assured that their best teams are represented for eligibility, meaning that there is no fear that a weak division champ steals eligibility with a miracle game over the conference favorite.