Just for comparison and discussion I am also listing who is left out. Rankings based on final selection committee rankings.

2014- P5 Champs
0 Loss FSU
1 Loss Alabama
1 Loss Oregon
1 Loss Ohio St
1 Loss Baylor

G5- 2 Loss Boise

At Large-
1 Loss TCU (Runner up due to tiebreaker)
2 Loss Mississippi State

Left out:
2 loss Michigan State (0-2 vs top 4)
3 Loss Ole Miss (beat Miss St, Boise, and Bama but lost three times including one to Memphis)


2015:
P5 Champs

Unbeaten Clemson
One Loss Alabama
One Loss Mich St
One Loss OU
Two Loss Stanford (Beat ND)

G5- One Loss Houston

At Large:
One Loss Iowa
One Loss Ohio St

Left Out
Two loss ND (lost to Clemson and Stanford)
Two loss FSU (Lost to Clemson and GT)

2016:
Unbeaten Alabama
One Loss Clemson
One Loss Washington
Two loss OU
Two loss Penn St

G5- Unbeaten WMU

At Large-
One Loss Ohio St
Two Loss Michigan

Left Out
3 Loss Wisconsin (Lost to 3 higher ranked B1G teams)
3 Loss USC


2017-

P5-
One Loss Clemson
One Loss UGA
One Loss OU
Two Loss Ohio St
Two Loss USC

G5- Unbeaten UCF

At Large-
One Loss Alabama
One Loss Wisconsin

Left Out
Three Loss Auburn (Lost to two of the top 4)
Two Loss Penn State


2018-

P5-
Unbeaten Alabama
Unbeaten Clemson
One Loss OU
One Loss Ohio State
Three Loss Washington

G5- Unbeaten UCF

At Large
Unbeaten Notre Dame
Two Loss UGA (Lost to Alabama but blown out by LSU)

Left Out:
Two Loss Michigan (blown out by Ohio St, lost to ND)
3 Loss Florida
3 Loss LSU



Conclusions:

Teams who had multiple shots at the best traded out for teams who didn’t really get one:

The lowest ranked teams that make it in with this model are every year pushing out first one out teams that have had multiple cracks at the top teams and failed. 2014 MichSt, 2015 ND, 2016 Wisconsin, 2017 Auburn, and 2018 Michigan all have multiple losses to the top teams and would be at best arguing for another chance. The teams replacing them most years (2014 Boise, 2015 UH, 2016 WMU, 2017 & 2018 UCF) are usually G5 teams that 60% of the time were unbeaten.

The teams left out really don’t have an argument left:

Among the first couple of teams left out each year, who has the best argument for inclusion? To me none of these arguments are strong.

Most years it works really well:
If we go back further autobids would give four and five loss teams a pass over better teams but fornthis sample it mostly provides fair access under a clear set of standards. Not perfect but mostly here. I’d prefer to have autobids as long as the champion P5 team has three or fewer losses. When you allow four and five lossmsquads in it loses its appeal IMO.

Thoughts? Who got screwed? Who doesn’t belong?

One point about the past is each P5 is larger and deeper than its BCS conference counterpart.

For this reason its become rare to see even 3 loss teams win a conference championship out of the P5.

If it were required that all conference champs be ranked in the Top 16, that should be very makeable for a P5 champ. Northwestern pulls the upset against Ohio St and they've got to be at least in the Top 16 for that.

It would take a non-top 25 P5 division winner to win a conference champ for them not to make the cut and I don't see how its even possible with the size of the conferences. It could happen with the XII though at only 10 teams.

1) Champs 16+ in playoff. Includes multiple G5's possibly.
2) At-larges if they are ranked in the Top 8 and if there is room.
3) Champs less than 16 are still guaranteed a NY8 bowl game.

NY8 adds LV Bowl and Citrus. Bowls rotate quarterfinals every other year. G5 team potentially could play in the Rose Bowl (Boise State).

Changes the 2014 National Champion to TCU. All other years crown the same champion.

If you go to page 2 of the threads I played out a 5-1-2 playoff all the way back to 2004, the year our 6 power conferences were effectively reduced to 5.

Occasionally some weak schools sneak in The 7th or 8th seed due to winning their conference but for the most part I think it produces a pretty fair field.

(01-17-2019 08:47 PM)IWokeUpLikeThis Wrote:  Changes the 2014 National Champion to TCU. All other years crown the same champion.

Sounds right.

Final CFP top 10 each year:

2018
1 Alabama 13-0
2 Clemson 13-0
3 Notre Dame 12-0
4 Oklahoma 12-1
5 Georgia 11-2
6 Ohio St. 12-1
7 Michigan 10-2
8 Central Florida 12-0
9 Washington 10-3
10 Florida 9-3

2017
1 Clemson 12-1
2 Oklahoma 12-1
3 Georgia 12-1
4 Alabama 11-1
5 Ohio St. 11-2
6 Wisconsin 12-1
7 Auburn 10-3
8 USC 11-2
9 Penn St. 10-2
10 Miami (FL) 10-2

2016
1 Alabama 13-0
2 Clemson 12-1
3 Ohio St. 11-1
4 Washington 12-1
5 Penn St. 11-2
6 Michigan 10-2
7 Oklahoma 10-2
8 Wisconsin 10-3
9 USC 9-3
10 Colorado 10-3

2015
1 Clemson 13-0
2 Alabama 12-1
3 Michigan St. 12-1
4 Oklahoma 11-1
5 Iowa 12-1
6 Stanford 11-2
7 Ohio St. 11-1
8 Notre Dame 10-2
9 Florida St. 10-2
10 North Carolina 11-2

2014
1 Alabama 12-1
2 Oregon 12-1
3 Florida St. 13-0
4 Ohio St. 12-1
5 Baylor 11-1
6 TCU 11-1
7 Mississippi St. 10-2
8 Michigan St. 10-2
9 Mississippi 9-3
10 Arizona 10-3

If you did a 5-0-3 instead of 5-1-2, you get
Michigan St. instead of Boise St. In bowls Boise beat Arizona and MSU beat Baylor
Notre Dame instead of UH In bowls UH whipped FSU. Notre Dame got stomped by Ohio St.
Wisconsin instead of W. Michigan Wisconsin beat WMU in Cotton Bowl
Auburn instead of UCF UCF beat Auburn in bowl.
Michigan instead of UCF MI got stomped in bowl. UCF lost.

Doesn't seem like any great loss to allow the G5 champ.

(01-17-2019 10:09 PM)bullet Wrote:  If you did a 5-0-3 instead of 5-1-2, you get
Michigan St. instead of Boise St. In bowls Boise beat Arizona and MSU beat Baylor
Notre Dame instead of UH In bowls UH whipped FSU. Notre Dame got stomped by Ohio St.
Wisconsin instead of W. Michigan Wisconsin beat WMU in Cotton Bowl
Auburn instead of UCF UCF beat Auburn in bowl.
Michigan instead of UCF MI got stomped in bowl. UCF lost.

Doesn't seem like any great loss to allow the G5 champ.

Each year it’s taking someone that already lost to multiple 8 team playoff squads and giving someone new a shot.

(01-17-2019 09:50 PM)Fighting Muskie Wrote:  If you go to page 2 of the threads I played out a 5-1-2 playoff all the way back to 2004, the year our 6 power conferences were effectively reduced to 5.

Occasionally some weak schools sneak in The 7th or 8th seed due to winning their conference but for the most part I think it produces a pretty fair field.

Which is why I am a bigger fan of conditional autobids.

P5 champs have to have fewer than four losses.

BCS top 10 going back to 2004

2013
1 Florida St. 13-0
2 Auburn 12-1
3 Alabama 11-1
4 Michigan St. 12-1
5 Stanford 11-2
6 Baylor 11-1
7 Ohio St. 12-1
8 Missouri 11-2
9 South Carolina 10-2
10 Oregon 10-2

2012
1 Notre Dame 12-0
2 Alabama 12-1
3 Florida 11-1
4 Oregon 11-1
5 Kansas St. 11-1
6 Stanford 11-2
7 Georgia 11-2
8 LSU 10-2
9 Texas A&M 10-2
10 South Carolina 10-2

2011
1 LSU 13-0
2 Alabama 11-1
3 Oklahoma St. 11-1
4 Stanford 11-1
5 Oregon 11-2
6 Arkansas 10-2
7 Boise St. 11-1
8 Kansas St. 10-2
9 South Carolina 10-2
10 Wisconsin 11-2

2010
1 Auburn 13-0
2 Oregon 12-0
3 TCU 12-0
4 Stanford 11-1
5 Wisconsin 11-1
6 Ohio St. 11-1
7 Oklahoma 11-2
8 Arkansas 10-2
9 Boise St. 11-1
10 Michigan St. 11-1

2009
1 Alabama 13-0
2 Texas 13-0
3 Cincinnati 12-0
4 TCU 12-0
5 Florida 12-1
6 Boise St. 13-0
7 Oregon 10-2
8 Ohio St. 10-2
9 Iowa 10-2
10 Georgia Tech 11-2

2008
1 Florida 12-1
2 Oklahoma 12-1
3 Texas 11-1
4 Alabama 12-1
5 USC 11-1
6 Utah 12-0
7 Texas Tech 11-1
8 Penn St. 11-1
9 Boise St. 12-0
10 Ohio St. 10-2

2007
1 Ohio St. 11-1
2 LSU 11-2
3 Virginia Tech 11-2
4 Oklahoma 11-2
5 Georgia 10-2
6 Missouri 11-2
7 USC 10-2
8 Kansas 11-1
9 West Virginia 10-2
10 Hawaii 12-0

2006
1 Ohio St. 12-0
2 Florida 12-1
3 Michigan 11-1
4 LSU 10-2
5 USC 10-2
6 Louisville 11-1
7 Wisconsin 11-1
8 Boise St. 12-0
9 Auburn 10-2
10 Oklahoma 11-2

2005
1 USC 12-0
2 Texas 12-0
3 Penn St. 10-1
4 Ohio St. 9-2
5 Oregon 10-1
6 Notre Dame 9-2
7 Georgia 10-2
8 Miami FL 9-2
9 Auburn 9-2
10 Virginia Tech 10-2

2004
1 USC 11-0
2 Oklahoma 12-0
3 Auburn 12-0
4 Texas 10-1
5 California 10-1
6 Utah 11-0
7 Georgia 9-2
8 Virginia Tech 9-2
9 Louisville 10-1
10 Boise St. 11-0

And 1998-2003 BCS final top 10

2003
1 Oklahoma 12-1
2 LSU 12-1
3 USC 11-1
4 Michigan 10-2
5 Ohio St. 10-2
6 Texas 10-2
7 Florida St. 10-2
8 Tennessee 10-2
9 Miami FL 10-2
10 Kansas St. 10-3

2002
1 Miami FL 12-0
2 Ohio St. 13-0
3 Georgia 11-1
4 USC 10-2
5 Iowa 11-1
6 Washington St. 9-2
7 Oklahoma 11-2
8 Kansas St. 10-2
9 Notre Dame 10-2
10 Texas 10-2

2001
1 Miami FL 11-0
2 Nebraska 10-1
3 Colorado 10-2
4 Oregon 10-1
5 Florida 9-2
6 Tennessee 10-2
7 Texas 10-2
8 Illinois 10-1
9 Stanford 9-2
10 Maryland 10-1

2000
1 Oklahoma 12-0
2 Florida St. 10-1
3 Miami FL 10-1
4 Washington 10-1
5 Virginia Tech 10-1
6 Oregon St. 10-1
7 Florida 10-2
8 Nebraska 9-2
9 Kansas St. 10-3
10 Oregon 9-2

1999
1 Florida St. 11-0
2 Virginia Tech 11-0
3 Nebraska 11-1
4 Alabama 10-2
5 Tennessee 9-2
6 Kansas St. 10-1
7 Wisconsin 9-2
8 Michigan 9-2
9 Michigan St. 9-2
10 Florida 9-3

1998
1 Tennessee 12-0
2 Florida St. 11-1
3 Kansas St. 11-1
4 Ohio St. 10-1
5 UCLA 10-1
6 Texas A&M 11-2
7 Arizona 10-1
8 Florida 9-2
9 Wisconsin 10-1
10 Tulane 11-0

(01-18-2019 10:56 AM)1845 Bear Wrote:  

(01-17-2019 09:50 PM)Fighting Muskie Wrote:  If you go to page 2 of the threads I played out a 5-1-2 playoff all the way back to 2004, the year our 6 power conferences were effectively reduced to 5.

Occasionally some weak schools sneak in The 7th or 8th seed due to winning their conference but for the most part I think it produces a pretty fair field.

Which is why I am a bigger fan of conditional autobids.

P5 champs have to have fewer than four losses.

Wow, so in your system a 3-loss P5 champ could get in over a 1-loss P5 non-champ? Even if the 1-loss team beat the 3-loss team in the season?

That's a very low bar and pretty crazy IMO under any circumstances, but especially given the highly uncertain way conferences choose their champs.

Much better just to go straight-8. That way, you just get our best approximation of the best 8 teams without muddling the waters with automatic bids for conference champs.

Also, straight-8 is much safer legally, as it doesn't involve a different standard for P5 and G5 champs.

(01-18-2019 11:39 AM)quo vadis Wrote:  

(01-18-2019 10:56 AM)1845 Bear Wrote:  

(01-17-2019 09:50 PM)Fighting Muskie Wrote:  If you go to page 2 of the threads I played out a 5-1-2 playoff all the way back to 2004, the year our 6 power conferences were effectively reduced to 5.

Occasionally some weak schools sneak in The 7th or 8th seed due to winning their conference but for the most part I think it produces a pretty fair field.

Which is why I am a bigger fan of conditional autobids.

P5 champs have to have fewer than four losses.

Wow, so in your system a 3-loss P5 champ could get in over a 1-loss P5 non-champ? Even if the 1-loss team beat the 3-loss team in the season?

1- I would consider moving it to three losses.
2- A one Loss non champ will usually already make it. The only times where I see the following:
- a one loss P5 team that isn’t a league champ
- That also isn’t in the two wildcards
- Or doesn’t slip in when a champ has 4 losses

The only cases over the last 20 seasons are:
2007 Kansas- Didn’t win its division and lost to 2 loss Mizzou. 3rd best in its conference
2006 Wisconsin- One of three big ten teams and ranked below 2 Loss LSU as an at-Large.

One was a low SOS team that didn’t win its division. The other was a third best team in its own conference.

Neither really has much of an argument.

Quote:That's a very low bar and pretty crazy IMO under any circumstances, but especially given the highly uncertain way conferences choose their champs.

Much better just to go straight-8. That way, you just get our best approximation of the best 8 teams without muddling the waters with automatic bids for conference champs.

Also, straight-8 is much safer legally, as it doesn't involve a different standard for P5 and G5 champs.

Straight eight would have serious issues. Unbeaten G5’s de-facto excluded by constantly moving goalposts (sorry but you magically are ranked 9th again!), certain schools benefiting from the eye test and others not, and constant subjectivity replacing objective standard. In each of the five playoff years we’d have seen a team with multiple shots at the top teams replaced by a team who didn’t get a shot. That’s a good thing.

Straight 8 is an improvement over 4 but it’s inferior to the. Other models out there for 8.

You could add a wild card play-in round and expand the field to 9-12 teams. Probably not necessary, but it would give 1-2 of the Left Out teams a spot and usually incorporate all teams ranked in the top-8.

So, this year may be Georgia and Michigan play for the right to reach the quarterfinals?

#1 Alabama v. #8 Washington
#2 Clemson v. #7 UCF
#3 Notre Dame v. wild card winner
#4 Oklahoma v. #5 Ohio St.

Something like that?

(01-18-2019 11:58 AM)1845 Bear Wrote:  

(01-18-2019 11:39 AM)quo vadis Wrote:  

(01-18-2019 10:56 AM)1845 Bear Wrote:  

(01-17-2019 09:50 PM)Fighting Muskie Wrote:  If you go to page 2 of the threads I played out a 5-1-2 playoff all the way back to 2004, the year our 6 power conferences were effectively reduced to 5.

Occasionally some weak schools sneak in The 7th or 8th seed due to winning their conference but for the most part I think it produces a pretty fair field.

Which is why I am a bigger fan of conditional autobids.

P5 champs have to have fewer than four losses.

Wow, so in your system a 3-loss P5 champ could get in over a 1-loss P5 non-champ? Even if the 1-loss team beat the 3-loss team in the season?

1- I would consider moving it to three losses.
2- A one Loss non champ will usually already make it. The only times where I see the following:
- a one loss P5 team that isn’t a league champ
- That also isn’t in the two wildcards
- Or doesn’t slip in when a champ has 4 losses

The only cases over the last 20 seasons are:
2007 Kansas- Didn’t win its division and lost to 2 loss Mizzou. 3rd best in its conference
2006 Wisconsin- One of three big ten teams and ranked below 2 Loss LSU as an at-Large.

One was a low SOS team that didn’t win its division. The other was a third best team in its own conference.

Neither really has much of an argument.

Quote:That's a very low bar and pretty crazy IMO under any circumstances, but especially given the highly uncertain way conferences choose their champs.

Much better just to go straight-8. That way, you just get our best approximation of the best 8 teams without muddling the waters with automatic bids for conference champs.

Also, straight-8 is much safer legally, as it doesn't involve a different standard for P5 and G5 champs.

Straight eight would have serious issues. Unbeaten G5’s de-facto excluded by constantly moving goalposts (sorry but you magically are ranked 9th again!), certain schools benefiting from the eye test and others not, and constant subjectivity replacing objective standard. In each of the five playoff years we’d have seen a team with multiple shots at the top teams replaced by a team who didn’t get a shot. That’s a good thing.

Straight 8 is an improvement over 4 but it’s inferior to the. Other models out there for 8.

Agree.

(01-18-2019 12:20 PM)YNot Wrote:  You could add a wild card play-in round and expand the field to 9-12 teams. Probably not necessary, but it would give 1-2 of the Left Out teams a spot and usually incorporate all teams ranked in the top-8.

So, this year may be Georgia and Michigan play for the right to reach the quarterfinals?

#1 Alabama v. #8 Washington
#2 Clemson v. #7 UCF
#3 Notre Dame v. wild card winner
#4 Oklahoma v. #5 Ohio St.

Something like that?

Why would anyone want to do that?

The only reason I see for a 9-12 team expansion with a play-in round would be to give extra slots to G5 champs.

For example an 11 team tourney:
5 P5 champs in to quarter-finals
3 (or 4) wildcards and 3 (or 2) G5 champs play-in to get a slot in the quarter-finals.

If the wildcards excluded division champs, then you effectively have a 16 team tournament and the P5 ccgs are a game in the round of 16.

(01-18-2019 11:58 AM)1845 Bear Wrote:  

(01-18-2019 11:39 AM)quo vadis Wrote:  

(01-18-2019 10:56 AM)1845 Bear Wrote:  

(01-17-2019 09:50 PM)Fighting Muskie Wrote:  If you go to page 2 of the threads I played out a 5-1-2 playoff all the way back to 2004, the year our 6 power conferences were effectively reduced to 5.

Occasionally some weak schools sneak in The 7th or 8th seed due to winning their conference but for the most part I think it produces a pretty fair field.

Which is why I am a bigger fan of conditional autobids.

P5 champs have to have fewer than four losses.

Wow, so in your system a 3-loss P5 champ could get in over a 1-loss P5 non-champ? Even if the 1-loss team beat the 3-loss team in the season?

1- I would consider moving it to three losses.
2- A one Loss non champ will usually already make it. The only times where I see the following:
- a one loss P5 team that isn’t a league champ
- That also isn’t in the two wildcards
- Or doesn’t slip in when a champ has 4 losses

The only cases over the last 20 seasons are:
2007 Kansas- Didn’t win its division and lost to 2 loss Mizzou. 3rd best in its conference
2006 Wisconsin- One of three big ten teams and ranked below 2 Loss LSU as an at-Large.

One was a low SOS team that didn’t win its division. The other was a third best team in its own conference.

Neither really has much of an argument.

Quote:That's a very low bar and pretty crazy IMO under any circumstances, but especially given the highly uncertain way conferences choose their champs.

Much better just to go straight-8. That way, you just get our best approximation of the best 8 teams without muddling the waters with automatic bids for conference champs.

Also, straight-8 is much safer legally, as it doesn't involve a different standard for P5 and G5 champs.

Straight eight would have serious issues. Unbeaten G5’s de-facto excluded by constantly moving goalposts (sorry but you magically are ranked 9th again!), certain schools benefiting from the eye test and others not, and constant subjectivity replacing objective standard. In each of the five playoff years we’d have seen a team with multiple shots at the top teams replaced by a team who didn’t get a shot. That’s a good thing.

If we're going to go on past history, as I've already pointed out, during the CFP era, 24 out of 25 P5 champs would have made the playoffs under straight-8, So again, a rule in search of a problem.

And IMO, we certainly don't want to craft a playoff specifically around G5 possibilities, because the odds that any G5 could ever win the national title is extremely low, at least no better than any similarly-ranked P5 team that didn't win its conference. The odds that a zero-loss #10 Boise could win isn't any better than the chances that a 2-loss #10 Florida could win. So no need to cater to either.

Beyond that, as I've explained, the issue here is flexibility. The major problem with autobids is that it is "automatic", or nearly so depending on what your ever-shifting rank or won-loss cutoff happens to be today. You've even acknowledged that in part by advocating a loss-cutoff, because you realize it is possible for a P5 conference to produce an obviously unworthy champion. So why should we trust that process for any P5 champ?

Given the uncertainty associated with much of this, better to have flexible standards that allow for consideration of all factors, including being a champ.