A Simple and Objective Solution to the College Football Playoff Problem

Show All Comments

gobucks5413's picture

Interesting...will never happen...but interesting. Certainly the first time I've read/heard something like this. In reality, I think we're going to keep seeing the same song and dance.

HS
wigmon's picture

I appreciate the thought and idea, but I can see every power 5 team bidding to get Rutgers, Oregon State, Kansas, etc on their non-conf schedule.  Their has to be a way to weight the value of the win other than just the conference the opponent comes from.  I think a collection of computer poll averages probably does a pretty good job of creating a value for the wins each team has.

HS
Dignan's picture

That's where the bonus points come in to play. A team would only get 2 points for beating the Rutgers of the world, but would get an additional 1 or 2 points for beating better teams. 

On the run from Johnny Law...ain’t no trip to Cleveland.

HS
BeatTTUN's picture

I didn't read any of your equations and theories Dignan, just wanted to let you know that Bottle Rocket is a good movie.

I saw Ryan Day hang 62 on Michigan...His hair was perfect.

Go Buckeyes Beat Michigan

HS
Dignan's picture

Yeah, it's been my favorite for a very long time.

On the run from Johnny Law...ain’t no trip to Cleveland.

HS
InTressITrust's picture

I like the thought but would need to crunch numbers more to get on board.  If a team (G5) wins all their games against 6-6 teams (bowl eligible) would that make them more worthy than a team that played a really tough schedule with a loss.  Not sure if this scenario could play out but would be interesting to see the numbers over the past 5 years with tweaks to them.

"I'm not going to lie. We're anxious to be a part of a matchup like that. It's two states that love the game of football." -Jim Tressel

HS
Dignan's picture

I don't think so, because the Group of 5 teams that get to 6 wins would only, presumably, have 6 points for 6 wins against other Group of 5 teams. Therefore the 12-0 Group of 5 team wouldn't get bonus points for those wins. I should have clarified by "bowl-eligible" I meant Power 5 teams with at least several wins against other Power 5 teams.

On the run from Johnny Law...ain’t no trip to Cleveland.

HS
bucknut1994's picture

You should call this the DCS and then in 4 years when OSU gets left out once or twice we can find a new system. 

#94Ways

HS
Dignan's picture

You only get left out if you don't win enough games or you play a weaker schedule than 2, 4, or 8 other teams. So a team has no one to blame but themselves and their AD for putting together a crappy schedule.

On the run from Johnny Law...ain’t no trip to Cleveland.

HS
ATHENS GA BUCKEYE's picture

Why not just do it like basketball does?  Its seems to work well for them.

First off, it's going to expand to 8 teams.  Thats happening.  Probably after next year.  ESPN owns the rights to broadcast the playoff for another 7 years but the marketing contracts all expire after next year and go back out for bid...

So at 8 teams, you have the winners of the 5 power conference championship games get an automatic bid.  If Northwestern would've beaten Ohio State or if Pitt would've beaten Clemson, they're automatically in. Then you take the 3 teams that didn't win a power 5 conference title game.  This year (prior to the playoff), that would have been Georgia, Notre Dame and UCF. 

Seed them based on who the committee determines is best.  This year would've been:

1 Bama v 8 UCF, 2 Clemson v 7 Washington, 3 Notre Dame v 6 Ohio State, 4 Oklahoma v 5 Georgia

The end.

HS
Dignan's picture

The problem is any and all subjectivity. Football teams play considerably fewer games than in any other sport, so there is more room for subjective opinion and debate to create and control the narrative. Ohio State didn't get in to the CFP because of a bad loss. That was the narrative. But nobody talked about how Bosa's departure 3 days before the game deflated the defense. Nobody talked about OSU having more wins against Power 5 teams than anybody else in the country. One aspect of the story - granted, it is a significant aspect - took over, and the debate was settled. With my system, there is no debate. Win your games, especially against good teams, and you'll get your shot at the championship. It's all settled on the field.

On the run from Johnny Law...ain’t no trip to Cleveland.

HS
allinosu's picture

The main sponsor for the playoff (Dr. Pepper) just signed a 6 year extension for the 4 team playoff  a couple of days ago which guarantees it until 2025.

HS
allinosu's picture

I can't fault the point system because it would get rid of chicken shit Saturday and forced ND into a conference but I'm fizzy on the 2-4-8 team playoff since it's tightly contracted for 4 and sites are determined years in advance that are contracted for 4 team higher than they were for above that. I don't see them ever doing a fly by night playoff.

HS
BuckeyeRealist13's picture

The committee has gotten the 4 teams correct every year. I have no issue with 4. Have to earn it, can’t expand the playoffs to let lesser deserving teams in. It’s not a participation trophy.

2x account suspension survivor 

HS
outdoorbuckeye's picture

I disagree. You are making teams "earn" it in different ways. Imo you havent earned it losing your conference. You havent earned it playing FCS teams. You definitely havent earned it playing one less game.

HS
Dignan's picture

My system is the very definition of earning it, and it removes all subjectivity. The same standard applies to every team. And it works great with just 2 teams in a single championship game, too.

On the run from Johnny Law...ain’t no trip to Cleveland.

HS
outdoorbuckeye's picture

I dont mind your system, however I would tweak it so you lose major points if you lose to an fcs team. Heck I'd have it -1 for even scheduling an fcs team.

HS
BuckeyeRealist13's picture

Your system is fair, that doesn’t make it right. Washington, didn’t deserve to play for a championship, they didn’t earn anything. 

2x account suspension survivor 

HS
idontsmellgood's picture

The committee has gotten the 4 teams correct every year. I have no issue with 4. Have to earn it, can’t expand the playoffs to let lesser deserving teams in. It’s not a participation trophy.

I disagree. One year they put in the 4 best (last year Alabama over OSU without winning their conference).  The next season they put in the most deserving (OSU over TCU and Baylor b/c no championship game, ND this year, Michigan State in 2015). You don't have to just earn it, you could win every game and still get left out. You could lay waste to the toughest schedule and have one slip up and get left out. Committee aside, the system sucks.

HS
BuckeyeRealist13's picture

Alabama was the better and more deserving team last year. 

2x account suspension survivor 

HS
thedewman10's picture

I’m all for redoing conferences and having a new division. Get rid of the Vanderbilts of the conferences that either never or rearely win a conference. Have a top division with 50-60 teams in it and you have to schedule all your games within that 50-60. Better games every week. Then a top 4 teams be accepted a little more. No more SOS argument or whatever. The other teams not in this would have their own division and playoffs. That way, a Vanderbilt or Indiana, would have a chance in that. It’s too watered down. Teams play now maybe, in a normal year, 3 games that really are good. And they probably make more money. Which is really what it’s all about anyway.

TheDewMan10

HS
Buck61's picture

add a relegation / promotion element to the mix, creating a static group would not be appropriate to up and coming schools. Take UCF as an example, if they were not in the original group they would be left out forever. Relegation is a method to make a school from living on the backs of successful programs and not attempting to compete at the highest level because they are protected.

Make the promotion element optional, a school may want to be at the top of the second tier and not compete against the upper group. Relegation should not be based on one year, a school could have a single bad season for reasons beyond their control.

HS
Dignan's picture

I thought of this, too. Each Power 5 conference could be tied to a "littler brother" Group of 5 conference. For example, the Big Ten is tied to the MAC. The MAC champion would have the opportunity to elevate to the B1G if they reach a certain points threshold (say, 15), as long as the last place B1G team also fails to reach a certain points threshold (say, 4).

On the run from Johnny Law...ain’t no trip to Cleveland.

HS
thedewman10's picture

I like both those ideas. UCF would not have been out however, because they would be included in the top 50-60...rememember, the conferences would be redone so they would be in one of the newer ones. Basically what the NFL did. It used to be AFC vs AFC only all year and then they had the top 2 play for a right to go to the super bowl. NFC did the same. Then they merged, play each other. I don’t like it as much with too many playoff teams. One year a team was 8-8 and got it. It also creates the scenario where the same 2 teams played during the season are now playing in the SB. There has to be something better for college.

TheDewMan10

HS
AesculusGlabra's picture

I like the idea! Would be interesting to go back and run all the numbers/teams from this year, to see how it adds up.

will ferrell yes GIF

HS
jpfbucks01's picture

i would reject this system out of hand

Sorry but beating Rutgers this past year should not count as double what a win over a UCF, or Fresno or Utah St or even UC would mean,, why because those 4 teams and possibly more were significantly better than Rutgers or even Oregon St were this year

All you have done is standardize a broadly based "opinion" into some kind of scoring system

Oregon State not only lost to us, they lost to 8-5 Nevada out of the Mt West, so should our win against them be worth double what a potential win of Boise St or Fresno St (who both beat Nevada and Fresno beat Arizona St) just because they were in a supposedly inferior conference?

And under this system UCF during the regular season would have earned only 11 points, 9 for wins over Group of 5 and 2 for beating Pitt and none for beating Navy and SCST. But 7-7 PItt who they beat by 31 points, would have earned 12 points for winning 6 games against Big 5 teams

So a 7-7 team would be ranked higher than an undefeated team,,,,,even though head up the undefeated team blew them out of a game that was 35-7 at half and 45 7 with 5 mins left to play

Nonsense

HS
Dignan's picture

There is no doubt that my system favors Power 5 teams, but so does every other system we've had. The Committee proved this year that even if a Group of 5 team goes 13-0 they're not getting in. And speaking of UCF, Navy would have counted, and they had a game against UNC cancelled, so that could have been another 2 points there. They were also conference champs, so that puts them at 14 total regulation points. But again, my system works because they played almost nothing but Group of 5 teams, and therefore haven't earned it on the field. The exception of Rutgers (bad P5 team) or UCF (good Go5 team) doesn't prove anything. There will always be exceptions no matter the system. 

On the run from Johnny Law...ain’t no trip to Cleveland.

HS
jpfbucks01's picture

You are right about Navy and the conference title, but even so, since the UNC game was in fact canceled they would have had exactly the same 12 points as Pitt a vastly inferior team that they trounced

I agree that both systems give them no chance. But your system gives greater rewards for beating terrible OOC teams from Power 5 rather than good teams from Big 5

A blow out win over a terrible Rutgers, Oregon st, Louisville, UNC, Kansas, or Akansas this past year should never count more than beating a 10-2 App St, or undefeated UCF, or 10-2 UC, or 10-3 UAB or Boise, or Fresno or Utah St

all those teams were definitively better than the Power 5 bottom dwellers I mentioned and i know this as those Power 5 teams all lost to Big 5 teams this year

Oregon St lost to Nevada, Rutgers to Buffalo, UNC got routed by ECU, Kansas lost to Nichols St, and Arky lost to both Colo St and North Texas

While UCF clobbered an ACC division winner in Pitt by over 30, Fresno beat UCLA and Arizona St, Troy beat Nebraska, UC beat UCLA, Temple beat MD, S Fla beat Illinois and Ga Tech, Houston beat Arizona, BYU beat Arizona and Wisconsin

This is not to say that this is true every year, but just as you cant arbitrarily assign the same value to clobbering Rutgers as you do to beating UM just because both are in the Big 10, and it makes no more sense to to so for beating Oregon St instead of Boise St, Boise was the better win for Okie St than our win over Oregon St, but your system would value to Oregon St win higher

HS
Dignan's picture

I definitely hear what you're saying. Every year there are a not insignificant number of Group of 5 teams that are demonstrably better than another not insignificant number of Power 5 teams. I'll keep thinking of a way to factor this in.

On the run from Johnny Law...ain’t no trip to Cleveland.

HS
jpfbucks01's picture

I have not thought this thru as much as your self but you could modify it similar to this

First the reality is ND is on par with being a Power 5 team and a win over them should count as though it was a power 5 win, BYU is close but not quite and all other independents should count as just Big 5 wins

2) Only tally points at the end of the season

3) Beat a Power 5 with 10 or more wins - 3 Pts (extra point for a win over 14 points)

4) Beat a power 5 over .500 but less than 10 wins - 2 pts (extra point for win over 21 points)

5) Beat a Big 5 with 10+ wins - 2 pts (extra point for win over 21 pts)

6) Beat a Big 5 over. 500 but less than 10 - 1 pt (extra point for win over 28 points)

7) Beat a Power 5 sub .500 - 1 pt (no extra points for margin)

8) Beat a Big 5 sub .500 or FCS or Ind - 0 pts (no extra points for margin)

9) loses to a sub .500 Power 5 lose 1 pt

10) Lose to a Big 5 sub .500 lose 3 pts

11) Win your conference - 5 pts

12) Lose to FCS - lose 5 points

Just a quick shot at this using OSU's schedule this past year

Oregon State - 1 point for win over sub .500 power 5

Rutgers - 1 point for win over sub .500 power 5

TCU - 2 points for win over middle Power 5 team

Tulane -1 point for win over middle Big 5 team 1 extra for margin

PSU - 2 pts for win over middle Power 5 team

Ind - 1 pt for win over sub .500 Power 5

Minn - 2 pts for win over middle power 5

Pur - 0 for loss to middle power 5 team (purdue was .500 at the end of the regular season)

Neb - 1 pt for win over sub .500 power 5 team

MSU - 2 pts for win over middle power 5 team

MD - 1 point for win over sub .500 Power 5

UM - 3 pts for win over power 5 with 10 or more wins, extra point for margin

NW - 2 pts for win over middle power 5 team plus 5 for conference title

Thats 26 points, if i loosen the point margin to say 21 or more, then they would pick up an extra point for the win over NW

May want to include a home/away/neutral factor as well

HS
armybuckinva's picture

I tried to calculate the points from the top 6 without digging into their opponents and their final records, so please bear with me as there could be some slop.

1. Oklahoma 36 points

2. Clemson 35 points

3. Alabama 33 points

4. Ohio State 33 points

5. Notre Dame 31 points

6. Georgia 25 points

Outside of Oklahoma being #1, this isn't a horrible final ranking.

HS
Oh Aych's picture

This isn't a horrible ranking, you're right. Perhaps Dignan will go back and score the CFB seasons the past few years using his system and then post what the results would have been.

For my part, I'm a fan of almost any system that makes the rankings objective. Take human opinions completely out of it. Use a relatively simple system, like Dignan's, or follow a more statistically complex system, like power rankings. In either case, teams can look at the system and see what they need to do, objectively, to meet their goals for the season. 

I(dah)O

HS
armybuckinva's picture

I get what you're saying.  I think everyone would have been more understanding with OSU's #6 ranking if there was a procedure created that valued all teams based off of similar factors.  Instead, we get the opinions of a dozen individuals and those people all have a different way evaluating who is who and where they belong.  I'm a process engineer so when I watch the CFP shows and listen to these people making the decisions, it literally drives me insane because they created the process but then never use it and if they never use then one could say definitively that the process is broken.  One year conference championships plays a big role, the next year nobody cares about them.

HS
BrutusB's picture

While I like the concept, you probably need to rethink those scores. Winning a game against a G5 shouldnt be worth nearly as much as winning a conference title. And there should be some penalty for losing - using your proposed numbers losing to Rutgers by 50 would be viewed the same as beating Citadel by 100.

HS
BUCKEYE3M's picture

Personally, I am a HUGE fan of removing the figure skating judges from the equation.  I think the thing I'd change is something to reward conference championships in some capacity.  In every other NCAA sanctioned sport, they matter a lot.  So, they should matter to a degree in football.

So, possible +3 or +4 for winning a conference title (which places it as a more valuable data point than beating the Rutgers/Oregon States/Kansas' of the world).
Or, you could place Conference Championships as the number one tie breaker (If two teams are tied in points, the team that won their respective conference wins. If each team won their respective conference, move to...).

HS
BUCKEYE3M's picture

Here is what the Top 10 in the final CFP poll would have looked like (EDIT: prior to bonus points):

Alabama: 24 points
Clemson: 25 points
Notre Dame: 22 Points
Oklahoma: 24 Points
Georgia: 18 Points
Ohio State: 25 Points
Michigan: 18 Points
UCF: 14 Points (UNC game canceled)
Washington: 19 Points
Florida: 15 Points

Source

HS
Triv's picture

I believe this is prior to his "bonus points" category being factored in, +1 for 11-16 point wins and +2 for wins over teams > 17 points

Sorry Urban, Woody is still my favorite

HS
BUCKEYE3M's picture

Yes, sorry, I didn't deep dive on the bonus points.

HS
Dignan's picture

Yes, these are the numbers that I came up with, but haven't had time to do the bonus points segment yet. I'll try to work on the bonus points for this year after work tonight, but I don't think the top four of OSU/Clemson/Alabama/Oklahoma will change.

On the run from Johnny Law...ain’t no trip to Cleveland.

HS
Dignan's picture

These numbers are making me rethink the bonus point threshold. Only 8 teams in the country would qualify as a double bonus win? That seems low. Also, the criticism of beating good Group of 5 teams vs. bad Power 5 teams is well-founded. Perhaps lowering the bonus thresholds would mitigate some of that, but I'll have to look at it with actual data. If only I didn't have a day job!

On the run from Johnny Law...ain’t no trip to Cleveland.

HS
BUCKEYE3M's picture

I think that beating Fresno State, for example, is fairly rewarded.  Keep in mind that the system cannot accurately predict from season to season which teams will be good or bad.  But, it absolutely should be designed to recognize the entirety of the respective leagues.  Part of the reason that UCF will end up with their record is the conference they play in, and the same is true of Rutgers.  

So, rewarding a win over UCF with +1 for being a G5 team is fair, because if they go on to win 9+ games, an additional +2 is applied.  Therefore, it is now on par with a standard win over a bowl eligible P5 team, and better than beating Oregon State.  I think this would be completely fair.

The entire purpose of a system is not to chase last year, meaning you cannot try to keep modifying a system to prevent something that happened once, to insure it never happens again, because you subject yourself to an entirely new set of loopholes and problems.  You have to have faith in what you created to be fair and true over time.  We tend to be prisoners of the moment in this regard. 

In other words, the system is not built to guarantee outcomes, but to provide equal opportunity for every team to make decisions knowing ahead of time what the pluses and minuses of their actions are.

HS
RunEddieRun1983's picture

Sounds better than what we have, but then it means ESPN won't be able to have a 13 hour show on their stupid channel and sell premium ads to companies to advertise and make them even more money.

Urban Meyer left an incredible legacy. 12/4/18 Ryan Day begins his.

HS
CreekBuck's picture

1 base point for ND, Army and BYU?   Same as Group of 5? I don't know about that for ND? Why schedule anyone them is they are worth the same as Charlotte or UTEP, ?  Obviously bonus points might get you more but ony if you win and they continue to win.

I would be in favor of some kind of point based system with less human bias.  But it would be up to humans to setup the point system and there is were the bias would be interjected.

This is similar to OHSAA FB Playoff tiered point system (Harbin Playoff) Point Calculation).  Which separate divisions by .5 point steps and then is also based on gaining points from the wins of the teams you beat. So beating a D1 team is worth 6.5 pts and D2 6, D3 5.5 ... so forth to D7.  Those are called 1st level points and divided by number of games you played. The wins of your "victims" are similarly weighted and divided and added as 2nd level points.  Now the controversy is in the makeup of the divisions mainly based on school population and "evenly" placing similar number of teams in each division. It's controversial because city school players are primarily bound by geographical area, where as parocial schools can recruit with no geographical bounds and subsequently can keep enroll at whatever size they want, can support.  OHSAA has recently put in a winning element into the formula to move schools up and down divisions (think Premiere league soccer) based on previous year’s W/L record. However, that part of the formula is small but has changed some teams up and down 1 division.  I doubt we see more than 1 division change unless the population in a school drastically changes. Anyway more than you ever wanted to know about Ohio High school football.

https://www.ohsaa.org/sports/football/computerrankingsinfo

You win with people.

HS
jmmysms's picture

Thank You. I thought that Ohio has used or was using this system. A long time ago I played on a very good undefeated team. An undefeated team beat another undefeated team and jumped over us into the playoffs.

Another method would be to get one point for every win your opponents had, lose one for playing down in division and possibly a two point kicker for winning your conference. Yes if you play in the tough SEC you won't get as many points as each team beats each other up. There are lots of mathematical options.

I chuckle at the figure skating comparison. 

JMMYSMS

HS
D Mac 24's picture

Doing just Ohio St and Alabama this year with bonus points - Ohio State = 37, Alabama = 30.  Alabama had 3 non-power 5 wins, and only one win vs. a 9 win team. 

HS
Dignan's picture

I did not factor in the bowl games, so the number I came up with for OSU is 30. 11 wins against Power 5 schools (22), 1 win against Go5 school (1), and conference champion (2). Then they got +1 for PSU (16 points), +1 for MSU (12 points), +1 for Northwestern (16 points), and +2 for Michigan (18 points). Haven't done any other schools yet.

On the run from Johnny Law...ain’t no trip to Cleveland.

HS
D Mac 24's picture

Ahhhhh..... I did not separate their win versus Northwestern and a big 10 championship so I had extra points there.    I actually really like this idea because it is simple and a little bit more objective. I also like Joel Klatt’s idea where there are factors that go into having teams just be able to make the consideration field. 

HS
Dignan's picture

I calculated the totals for the top 6 schools, and here they are, in order:

1. Clemson, 31 points. 25 regulation points, plus 6 bonus points: +1 Tx A&M (12), +1 GT (11), +1 Syracuse (14), +1 NC State (13), +1 Duke (11), +1 Pitt (12)

2. Alabama, 30 points. 24 regulation points, plus 6 bonus points: +1 Tx A&M (12), +1 Mizzou (12), +1 LSU (14), +1 Miss. St. (12), +2 Georgia (18)

3. Ohio State, 30 points. 25 regulation points, plus 5 bonus points: +1 PSU (16), +1 MSU (12), +2 Michigan (18), +1 Northwestern (16) [OSU loses tiebreaker because of the loss.]

4. Notre Dame, 28 points. 22 regulation points, plus 6 bonus points: +2 Michigan (18), +1 Stanford (13), +1 Pitt (12), +1 Northwestern (16), +1 Syracuse (14)

5. Oklahoma, 28 points. 24 regulation points, plus 4 bonus points: +1 Iowa St (13), +1 WVU (14), +2 Texas (17) [OU loses tiebreaker to Notre Dame because of their loss.]

6. Georgia, 22 points. 18 regulation points, plus 4 bonus points: +1 Mizzou (12), +1 Florida (13), +1 Kentucky (14), +1 GT (11)

So in my system the playoff would have been Clemson v. Notre Dame, and Alabama v. Ohio State.

On the run from Johnny Law...ain’t no trip to Cleveland.

HS