Friday, September 5, 2014

What Playoff Standings Should Look Like

There will be no more BCS standings. In fact, there will be no standings with a formula that we can reliably project when the College Football Playoff era begins in the 2014 season. The 13-person selection committee will have sole discretion on which teams make the four-team playoff field.

A few months ago, we introduced a standings model for the committee, and asked for readers' suggestions and comments. We received a healthy amount of responses, most of which were very helpful. After taking much into consideration, we have revised our proposed standings for use by the playoff committee.

We're not arrogant or foolhardy enough to think the committee will necessarily adopt our formula, or admit it publicly. But we do hope that by starting this discussion, we'll move into a more transparent process where we won't be greeted with major surprises come football's version of Selection Sunday.

We more or less stuck to the main criteria, which the committee has emphasized as crucial in its selection process. But we have made major revisions to the distribution of each category:

1) AP poll (40 percent): The eyeball test has to mean something, and the AP poll is the only poll that's completely transparent, with each voter's ballot available for the public to scrutinize each week. It's also the most prestigious poll that's widely used by the media. (Increased from 20 percent)

2) Computer rankings (40 percent): Kenneth Massey compiles the median and mean rankings of each team from more than 100 computers each week. It's less biased than the human polls, and the large sample size removes undue influence by outliers. (kept at 40 percent)

3) Strength of schedule (10 percent): While there are many models to choose from, Jeff Sagarin has the most time-tested SoS formula—including results from all Division I games, FBS and FCS—that's meticulously and promptly updated each week. But since SoS is a component in every computer ranking, too much influence by the SoS would create a double-jeopardy redundancy. (Decreased from 30 percent)

4) Conference championship (10 percent): Only teams that win their conference championships will get the bonus, and it has to be a significant one. A team that fails to win its conference must be so highly-ranked in every other aspect to jump champions from other Big Five conferences. (kept at 10 percent)

With that in mind, check out what the standings would've looked like after Week 1.


Mike said...

I think the conference championship should be worth even more. The whole idea of not winning the conference means that you are not as good as another team. The playoff system is meant to find the best team among several teams that have no direct competition. I would rather have a 60/30/30 split between polls, schedule, and conference championship.

Ideally, we would have 4 major conferences, and each conference's championship game would be essentially the first round of the playoffs. I don't care how much the SEC thinks their #3 is worth the 4th seed in the playoff. If you aren't the best team in the conference, you aren't the best team in the country.

Anonymous said...

But you may still be better than the #1 team in another conference

Anonymous said...

I am sorry Mike but, you can have a conference that is really great at basketball but not so good at football. Why should their conference champion who could lose to a local high school team be in the play off? I thought the idea was to get the 4 best teams? Many times that would be 4 different conference champions. Sometimes it wouldn't. Sometimes it may have 2 teams from the SEC. It might even have a year with no teams from the SEC. That would just be the way the teams play. So stop crying about it.

Scot Dew said...

Conference Championship should be a tie-breaker.Special teams (Kicker rating 1-4, coverage teams 1-4, return teams 1-4) should be a tie breaker for seeding purposes.
We want the best 4 teams, period. SOS, week 6-end of season with each week counting as some pct and increasing until the end of season (the best coaches teams get better), with a weighted scale for the toughest game vs. weakest game. Consideration for trap games.
Point differential up to 27 points (after 28- personnel and play calling changes in the 4th quarter), Schools with Conference championships must throw out weakest opponent quality points. Teams losing their last game by more than 2 scores (that is 11 points or more)- OUT! Teams giving up over 44 points in a game get minus (x)quality points unless the team they played averaged over 50 points a game. AP poll no more than 33.3 percent. We want the committee to own their decisions, not the AP or Computers (16.7%).
By the way, ideally we would have 10 teams. 6 seeded teams and 4 teams that had a play-in game, Higher seed get's a "Home" game indoors regionally close to the school 1 v. 4, 2 v. 3. winners go bowling with the big boys. This would allow for a second non-AQ school (group of 5) to bully their way in. Or a 1 loss runner up team from a conference with no championship game (Big 12) to play their way in. Same seeding and break down for the 8 teams on January 1st. Final 4 play in January week 2 Championship played Monday night week 3 Dome, or warmer climate site.
The goal is to have the best 2 teams plying at the highest level at the end of the season play for the championship. The side effect is more great football games.

On a slightly different tangent. As these extra games create extra income for the Universities and NCAA. The players who participate in these games should receive some compensation separate from any current and payable upon graduation or 6-9 months after withdrawal/forfeiture of scholarship from their school, Transfers have to wait until they finish at the school they transfer to unless they already have earned their Baccalaureate degree before transfer.
Coaches, assistant, graduate assistants, volunteer student assistants and all employees of the program get team performance bonuses equivalent of the head coach pro-rated vs. their salary or in the case of the student assistants/trainers 40-50% of the players individual kicker bonus.

Anonymous said...

I think the Conference championship HAS TO BE LEFT OUT COMPLETELY !!!! Heres why.......
1.) Big 12 doesnt have one !
2.) The best team doesnt always win conference championship game !(see alabama vs LSU)
3.A team that is top a top 4 team may be unfairly eliminated(Example=SEC, who would probably send 2 or 3 this year)

Anonymous said...

strength of schedule 10%???? WTF? so the SEC can continue to play an FCS team in week 10, and it won't matter...BS

Anonymous said...

sos is a must and conference a must the ones that dont have better be getting one if not they will wish they will cost them in the long run they better get to work its the first year of this system you can do it or maybe you dont want to

Anonymous said...

So, looking at the rankings chart you created, you ranked 33 teams, and for some reason declared that the playoff should be the teams ranked 1, 4, 6, and 8.

It's clear that in your selection of the teams that are to be in the playoff, you made a strict only-one-team-per-conference rule. Isn't the point of a 4-team playoff to have the 4 best teams in the nation, which according to your own formula would be MS State, Alabama, Auburn, and FL State?

Anonymous said...

Double post, since I can't apparently edit anonymous posts...

Would it then not be to every team's benefit to pull a BS Notre Dame move and declare themselves "independent" (even though Notre Dame is for all intents and purposes actually in the ACC)?

The Guru said...

That's actually not the reason. We have to be mindful that those SEC teams have a guaranteed six losses among them (assuming Georgia win the SEC East) so not all of them will still have 0 or 1 loss.

I don't believe the committee will only pick conference champions, but the nature of these conference games typically only allows 1 team to be picked.

herdadreh said...

When you add in SOS as a separate factor, you double count (overweight) SOS, since it is incorporated into AP polls and computer rankings. The fact that almost everyone makes this mistake when discussing overall CFB ratings doesn't make it OK for you to make this mistake too.

The Guru said...

Herdadreh - I actually considered the SoS part very carefully so that we didn't double-dip. I feel like a 10 percent distribution for SoS is fair as the polls often times don't take that into consideration.

Anonymous said...

Is there a published procedure on the access bowl selection process? Why is Ole Miss ignored and Georgia picked up? With the projected ACC champ in the playoff, does the Orange have to take ACC#2?

The Guru said...

It's at the discretion of the committee, with certain tie-ins to be honored. Orange Bowl has to take an ACC team no matter what. I projected that Ole Miss would lose again so it's not in an NY6 bowl.