Comparing, since 2003, the 5-1-2 (6-1-1) vs. Straight 8.


Each year will be provided a 6-1-1 (during seasons with 6 power conferences) or 5-1-2 (during seasons with 5 power conferences) and Straight 8 bracket. The ranking numbers used are CFP and BCS. Why 2003? Because prior to 2003, the BCS only ranked the top 15. In case of co-champions, I used head-to-head then overall record, if necessary.

A couple observations:
- The power conference champions' rankings really tightened when the playoffs began (none lower than #9)
- 5-1-2/6-1-1
--- Non-power spot by conference (total: 17): MW (8), AAC (4), MAC (3), WAC (2), CUSA (0), SUN BELT (0)
--- At-large spot(s) by conference (total: 23): SEC (8), B1G (7), XII (5), IND (2), PAC (1), ACC (0), BIG EAST (0)

2003

6-1-1
#1 Oklahoma vs #11 Miami OH
#2 LSU vs #10 Kansas St
#3 USC vs #9 Miami FL
#4 Michigan vs #7 Florida St

Straight 8
#1 Oklahoma vs #8 Tennessee
#2 LSU vs #7 Florida St
#3 USC vs #6 Texas
#4 Michigan vs #5 Ohio St

2004

6-1-1
#1 USC vs #21 Pittsburgh
#2 Oklahoma vs #13 Michigan
#3 Auburn vs #8 Virginia Tech
#4 Texas vs #6 Utah

Straight 8
#1 USC vs #8 Virginia Tech
#2 Oklahoma vs #7 Georgia
#3 Auburn vs #6 Utah
#4 Texas vs #5 California

2005

6-1-1
#1 USC vs #22 Florida St
#2 Texas vs #14 TCU
#3 Penn St vs #11 West Virginia
#4 Ohio St vs #6 Georgia

Straight 8
#1 USC vs #8 Miami FL
#2 Texas vs #7 Georgia
#3 Penn St vs #6 Notre Dame
#4 Ohio St vs #5 Oregon

2006

6-1-1
#1 Ohio St vs #14 Wake Forest
#2 Florida vs #10 Oklahoma
#3 Michigan vs #8 Boise St
#5 USC vs #6 Louisville

Straight 8
#1 Ohio St vs #8 Boise St
#2 Florida vs #7 Wisconsin
#3 Michigan vs #6 Louisville
#4 LSU vs #5 USC

2007

6-1-1
#1 Ohio St vs #10 Hawaii
#2 LSU vs #9 West Virginia
#3 Virginia Tech vs #7 USC
#4 Oklahoma vs #5 Georgia

Straight 8
#1 Ohio St vs #8 Kansas
#2 LSU vs #7 USC
#3 Virginia Tech vs #6 Missouri
#4 Oklahoma vs #5 Georgia

2008

6-1-1
#1 Oklahoma vs #19 Virginia Tech
#2 Florida vs #12 Cincinnati
#3 Texas vs #8 Penn St
#5 USC vs #6 Utah

Straight 8
#1 Oklahoma vs #8 Penn St
#2 Florida vs #7 Texas Tech
#3 Texas vs #6 Utah
#4 Alabama vs #5 USC

2009

6-1-1
#1 Alabama vs #9 Georgia Tech
#2 Texas vs #8 Ohio St
#3 Cincinnati vs #7 Oregon
#4 TCU vs #5 Florida

Straight 8
#1 Alabama vs #8 Ohio St
#2 Texas vs #7 Oregon
#3 Cincinnati vs #6 Boise St
#4 TCU vs #5 Florida

2010

6-1-1
#1 Auburn vs NR Connecticut
#2 Oregon vs #13 Virginia Tech
#3 TCU vs #9 Michigan St
#4 Stanford vs #7 Oklahoma

Straight 8
#1 Auburn vs #8 Arkansas
#2 Oregon vs #7 Oklahoma
#3 TCU vs #6 Ohio St
#4 Stanford vs #5 Wisconsin

2011

6-1-1
#1 LSU vs #23 West Virginia
#2 Alabama vs #18 TCU
#3 Oklahoma St vs #15 Clemson
#5 Oregon vs # 10 Wisconsin

Straight 8
#1 LSU vs #8 Kansas St
#2 Alabama vs #7 Boise St
#3 Oklahoma St vs #6 Arkansas
#4 Stanford vs #5 Oregon

2012

6-1-1
#1 Notre Dame vs NR Wisconsin
#2 Alabama vs #21 Louisville
#5 Kansas St vs #15 Northern Illinois
#6 Stanford vs #12 Florida St

Straight 8
#1 Notre Dame vs #8 LSU
#2 Alabama vs #7 Georgia
#3 Florida vs #6 Stanford
#4 Oregon vs #5 Kansas St

2013

6-1-1
#1 Florida St vs #20 Fresno St
#2 Auburn vs #15 Central Florida
#3 Alabama vs #6 Baylor
#4 Michigan St vs #5 Stanford

Straight 8
#1 Florida St vs #8 Missouri
#2 Auburn vs #7 Ohio St
#3 Alabama vs #6 Baylor
#4 Michigan St vs #5 Stanford

2014

5-1-2
#1 Alabama vs #20 Boise St
#2 Oregon vs #7 Mississippi St
#3 Florida St vs #6 TCU
#4 Ohio St vs #5 Baylor

Straight 8
#1 Alabama vs #8 Michigan St
#2 Oregon vs #7 Mississippi St
#3 Florida St vs #6 TCU
#4 Ohio St vs #5 Baylor

2015

5-1-2
#1 Clemson vs #18 Houston
#2 Alabama vs #7 Ohio St
#3 Michigan St vs #6 Stanford
#4 Oklahoma vs #5 Iowa

Straight 8
#1 Clemson vs #8 Notre Dame
#2 Alabama vs #7 Ohio St
#3 Michigan St vs #6 Stanford
#4 Oklahoma vs #5 Iowa

2016

5-1-2
#1 Alabama vs #18 Western Michigan
#2 Clemson vs #7 Oklahoma
#3 Ohio St vs #6 Michigan
#4 Washington vs #5 Penn St

Straight 8
#1 Alabama vs #8 Wisconsin
#2 Clemson vs #7 Oklahoma
#3 Ohio St vs #6 Michigan
#4 Washington vs #5 Penn St

2017

5-1-2
#1 Clemson vs #12 Central Florida
#2 Oklahoma vs #8 USC
#3 Georgia vs #6 Wisconsin
#4 Alabama vs #5 Ohio St

Straight 8
#1 Clemson vs #8 USC
#2 Oklahoma vs #7 Auburn
#3 Georgia vs #6 Wisconsin
#4 Alabama vs #5 Ohio St

2018

5-1-2
#1 Alabama vs #9 Washington
#2 Clemson vs #8 Central Florida
#3 Notre Dame vs #6 Ohio St
#4 Oklahoma vs #5 Georgia

Straight 8
#1 Alabama vs #8 Central Florida
#2 Clemson vs #7 Michigan
#3 Notre Dame vs #6 Ohio St
#4 Oklahoma vs #5 Georgia

2019

5-1-2
#1 LSU vs #17 Memphis
#2 Ohio St vs #7 Baylor
#3 Clemson vs #6 Oregon
#4 Oklahoma vs #5 Georgia

Straight 8
#1 LSU vs #8 Wisconsin
#2 Ohio St vs #7 Baylor
#3 Clemson vs #6 Oregon
#4 Oklahoma vs #5 Georgia

Sub-T15 6-1-1 teams:
#15 Clemson
#15 UCF
#15 NIU
#18 TCU
#19 Virginia Tech
#20 Fresno
#21 Pitt
#21 Louisville
#22 Florida St
#23 West Virginia
NR Wisconsin
NR UConn

Sub-T15 5-1-2 teams:
#17 Memphis
#18 WMU
#18 Houston
#20 Boise

One thing I noticed is the "tightening up" of teams since the advent of the CFP.

For example, in each of the six years of the CFP, there is just a one team difference between the two methods. In five of those six years, it is the substitution of a much lower ranked G5 team for a #8 ranked P5/Notre Dame team. The exception is 2018, when the G5 team was ranked in the top 8, so made it under both systems.

In contrast, I looked at the six years before the CFP, from 2008 - 2013, and in each of those years there was at least a two team difference between the systems, that is, the two systems produce six teams in common, but two different teams. In a couple years it is a 3 team difference.

Part of that is probably due to realignment, the shrinkage of the AQ6 to the P5 occurred almost simultaneously with the switch from BCS to CFP. But I also think it is mind-set. Coaches and thus teams are responding to the different incentives of the 4-team CFP compared to the 2-team BCS. An 8-team playoff would have different incentives from either.

I personally think the 2014 - 2019 matchups produced by S8 are much better than that produced by 5-1-2. And importantly, even with the tightening up, there has always been a difference of at least one team. The two methods never produce the same eight teams even though that is not precluded by the systems.

(04-17-2020 01:38 PM)quo vadis Wrote:  One thing I noticed is the "tightening up" of teams since the advent of the CFP.

For example, in each of the six years of the CFP, there is just a one team difference between the two methods. In five of those six years, it is the substitution of a much lower ranked G5 team for a #8 ranked P5/Notre Dame team. The exception is 2018, when the G5 team was ranked in the top 8, so made it under both systems.

In contrast, I looked at the six years before the CFP, from 2008 - 2013, and in each of those years there was at least a two team difference between the systems, that is, the two systems produce six teams in common, but two different teams. In a couple years it is a 3 team difference.

Part of that is probably due to realignment, the shrinkage of the AQ6 to the P5 occurred almost simultaneously with the switch from BCS to CFP. But I also think it is mind-set. Coaches and thus teams are responding to the different incentives of the 4-team CFP compared to the 2-team BCS. An 8-team playoff would have different incentives from either.

I personally think the 2014 - 2019 matchups produced by S8 are much better than that produced by 5-1-2. And importantly, even with the tightening up, there has always been a difference of at least one team. The two methods never produce the same eight teams even though that is not precluded by the systems.

Or maybe the committee is tightening up the ratings differences? It definitely seems like the G5 get less consideration than they did in the poll systems. And traditional powers get more of a break. You also see the polls drift closer to the CFP rankings later in the year.

(04-17-2020 01:51 PM)bullet Wrote:  

(04-17-2020 01:38 PM)quo vadis Wrote:  One thing I noticed is the "tightening up" of teams since the advent of the CFP.

For example, in each of the six years of the CFP, there is just a one team difference between the two methods. In five of those six years, it is the substitution of a much lower ranked G5 team for a #8 ranked P5/Notre Dame team. The exception is 2018, when the G5 team was ranked in the top 8, so made it under both systems.

In contrast, I looked at the six years before the CFP, from 2008 - 2013, and in each of those years there was at least a two team difference between the systems, that is, the two systems produce six teams in common, but two different teams. In a couple years it is a 3 team difference.

Part of that is probably due to realignment, the shrinkage of the AQ6 to the P5 occurred almost simultaneously with the switch from BCS to CFP. But I also think it is mind-set. Coaches and thus teams are responding to the different incentives of the 4-team CFP compared to the 2-team BCS. An 8-team playoff would have different incentives from either.

I personally think the 2014 - 2019 matchups produced by S8 are much better than that produced by 5-1-2. And importantly, even with the tightening up, there has always been a difference of at least one team. The two methods never produce the same eight teams even though that is not precluded by the systems.

Or maybe the committee is tightening up the ratings differences? It definitely seems like the G5 get less consideration than they did in the poll systems. And traditional powers get more of a break. You also see the polls drift closer to the CFP rankings later in the year.

The use of computer ratings in the BCS is a big difference. In general, computer ratings are/were favorable to the best "G5" or "non-BCS" teams because most computer formulas heavily weight the basic W-L record, giving any 12-0 or 11-1 team a big boost when compared to a team with more than one loss, even if there are substantial differences in strength of schedule and quality wins.

The committee, in contrast, seems to lean toward the "Who did they play and who did they beat" frame of reference, which is always going to favor teams that have good records against tough schedules.

You should also do 5-3

#20 Fresno State knocking out Missouri in 2013
#20 Boise State knocking out Michigan State in 2014
#18 Houston knocking out Notre Dame in 2015
#18 Western Michigan knocking out Wisconsin in 2016
#7 Michigan knocking out #9 Washington (P12 champs) in 2018

These are ones that would not go over well (2012 is a disaster the B1G champ not rated, the Big Eats Champ rated 21st; at least NIU was #15)

#15 Northern Illinois over not top 20 schools in 2012
#12 UCF over #8 Auburn in 2017
#8 UCF over #7 Michigan in 2018 (really Washington #9 as P12 champs bump Michigan #7)
#15 Memphis over #8 Wisconsin in 2019

These would work fine, as bumped schools are neither conference champs nor top seeds.

I think this screams for both the G5 and major conference champions (see 2012) be ranked top 15 to secure their spots. This would in turn force the FBS conferences to end divisions and have the best two teams in their CCG qualifier. If the same ratings are used only one P5 champ would be denied in the last 8 years, and 4 of the last 8 G5 would qualify.

What happens in a straight 8 like 2012 when the SEC represents half the field while the Big Ten and ACC aren’t represented at all?

(04-17-2020 04:11 PM)SkullyMaroo Wrote:  What happens in a straight 8 like 2012 when the SEC represents half the field while the Big Ten and ACC aren’t represented at all?

Ohio State was 12-0 that season but on a postseason ban. They were ranked #3 in the last AP poll before the bowl games.

(04-17-2020 04:22 PM)Wedge Wrote:  

(04-17-2020 04:11 PM)SkullyMaroo Wrote:  What happens in a straight 8 like 2012 when the SEC represents half the field while the Big Ten and ACC aren’t represented at all?

Ohio State was 12-0 that season but on a postseason ban. They were ranked #3 in the last AP poll before the bowl games.

We are splitting hairs. What happens in a straight 8 like 2012 when the SEC represents half the field while the Big Ten and ACC aren’t represented at all, the former because Ohio State isn’t eligible for postseason?

More succinctly, what happens in a straight 8 like 2012 when the SEC represents half the field?

(04-17-2020 04:59 PM)SkullyMaroo Wrote:  

(04-17-2020 04:22 PM)Wedge Wrote:  

(04-17-2020 04:11 PM)SkullyMaroo Wrote:  What happens in a straight 8 like 2012 when the SEC represents half the field while the Big Ten and ACC aren’t represented at all?

Ohio State was 12-0 that season but on a postseason ban. They were ranked #3 in the last AP poll before the bowl games.

We are splitting hairs. What happens in a straight 8 like 2012 when the SEC represents half the field while the Big Ten and ACC aren’t represented at all, the former because Ohio State isn’t eligible for postseason?

More succinctly, what happens in a straight 8 like 2012 when the SEC represents half the field?

The 2012 example requires the Big Ten's best team to be on probation, which makes that season a fluke by definition -- unless in the future there is a huge increase in top teams getting postseason bans. In the absence of that surge in postseason bans, two P5 conferences getting shut out of a top 8 is a scenario we shouldn't lose much sleep over. Certainly shouldn't build an entire playoff system based on fearing that scenario.

As for just one conference getting shut out -- several folks have already mentioned that any significant possibility of a P5 conference being left out of a top 8 might be enough for them to say, "If we don't have a guarantee in an 8-team playoff, then there's not enough reason to expand the playoff at all."

How much interest will wane if all the teams come from the same conference?

(04-17-2020 02:05 PM)Wedge Wrote:  

(04-17-2020 01:51 PM)bullet Wrote:  

(04-17-2020 01:38 PM)quo vadis Wrote:  One thing I noticed is the "tightening up" of teams since the advent of the CFP.

For example, in each of the six years of the CFP, there is just a one team difference between the two methods. In five of those six years, it is the substitution of a much lower ranked G5 team for a #8 ranked P5/Notre Dame team. The exception is 2018, when the G5 team was ranked in the top 8, so made it under both systems.

In contrast, I looked at the six years before the CFP, from 2008 - 2013, and in each of those years there was at least a two team difference between the systems, that is, the two systems produce six teams in common, but two different teams. In a couple years it is a 3 team difference.

Part of that is probably due to realignment, the shrinkage of the AQ6 to the P5 occurred almost simultaneously with the switch from BCS to CFP. But I also think it is mind-set. Coaches and thus teams are responding to the different incentives of the 4-team CFP compared to the 2-team BCS. An 8-team playoff would have different incentives from either.

I personally think the 2014 - 2019 matchups produced by S8 are much better than that produced by 5-1-2. And importantly, even with the tightening up, there has always been a difference of at least one team. The two methods never produce the same eight teams even though that is not precluded by the systems.

Or maybe the committee is tightening up the ratings differences? It definitely seems like the G5 get less consideration than they did in the poll systems. And traditional powers get more of a break. You also see the polls drift closer to the CFP rankings later in the year.

The use of computer ratings in the BCS is a big difference. In general, computer ratings are/were favorable to the best "G5" or "non-BCS" teams because most computer formulas heavily weight the basic W-L record, giving any 12-0 or 11-1 team a big boost when compared to a team with more than one loss, even if there are substantial differences in strength of schedule and quality wins.

The committee, in contrast, seems to lean toward the "Who did they play and who did they beat" frame of reference, which is always going to favor teams that have good records against tough schedules.

The committee is generally very big on "who did they beat?" That hurts the G5. And in contrast, they ignore "who did they lose to?" They don't really punish, at least P5 teams, for a loss to a weak team.

(04-17-2020 02:05 PM)Stugray2 Wrote:  You should also do 5-3

#20 Fresno State knocking out Missouri in 2013
#20 Boise State knocking out Michigan State in 2014
#18 Houston knocking out Notre Dame in 2015
#18 Western Michigan knocking out Wisconsin in 2016
#7 Michigan knocking out #9 Washington (P12 champs) in 2018

These are ones that would not go over well (2012 is a disaster the B1G champ not rated, the Big Eats Champ rated 21st; at least NIU was #15)

#15 Northern Illinois over not top 20 schools in 2012
#12 UCF over #8 Auburn in 2017
#8 UCF over #7 Michigan in 2018 (really Washington #9 as P12 champs bump Michigan #7)
#15 Memphis over #8 Wisconsin in 2019

These would work fine, as bumped schools are neither conference champs nor top seeds.

I think this screams for both the G5 and major conference champions (see 2012) be ranked top 15 to secure their spots. This would in turn force the FBS conferences to end divisions and have the best two teams in their CCG qualifier. If the same ratings are used only one P5 champ would be denied in the last 8 years, and 4 of the last 8 G5 would qualify.

2014 Boise beat Arizona in the Fiesta.
2015 Houston beat FSU in the Peach.

(04-17-2020 01:51 PM)bullet Wrote:  

(04-17-2020 01:38 PM)quo vadis Wrote:  One thing I noticed is the "tightening up" of teams since the advent of the CFP.

For example, in each of the six years of the CFP, there is just a one team difference between the two methods. In five of those six years, it is the substitution of a much lower ranked G5 team for a #8 ranked P5/Notre Dame team. The exception is 2018, when the G5 team was ranked in the top 8, so made it under both systems.

In contrast, I looked at the six years before the CFP, from 2008 - 2013, and in each of those years there was at least a two team difference between the systems, that is, the two systems produce six teams in common, but two different teams. In a couple years it is a 3 team difference.

Part of that is probably due to realignment, the shrinkage of the AQ6 to the P5 occurred almost simultaneously with the switch from BCS to CFP. But I also think it is mind-set. Coaches and thus teams are responding to the different incentives of the 4-team CFP compared to the 2-team BCS. An 8-team playoff would have different incentives from either.

I personally think the 2014 - 2019 matchups produced by S8 are much better than that produced by 5-1-2. And importantly, even with the tightening up, there has always been a difference of at least one team. The two methods never produce the same eight teams even though that is not precluded by the systems.

Or maybe the committee is tightening up the ratings differences? It definitely seems like the G5 get less consideration than they did in the poll systems. And traditional powers get more of a break. You also see the polls drift closer to the CFP rankings later in the year.

The claim of an AP or Coaches poll "drift" to the CFP has been claimed by some but never demonstrated. In fact, when you walk through the rankings history, there seems to be as much of a pull in the other direction if not more (CFP following, not leading, the AP). Makes sense too, since the AP and Coaches polls come out before the CFP poll.

About the latter, it's been shown that the computers of 10 years ago also ranked the top non-AQ teams higher than the top G5 teams get ranked today. E.g., in say 2010, when a Boise or TCU would rise to #3 or#4 in the BCS rankings and the human polls, they would rank similarly high in the computers. When a Memphis in 2019 is ranked #17 by the CFP, their computer ranking is around #17 as well.

This suggests that the top non-AQ teams of 10 years ago were just better, relative to the competition, than the top G5 teams are today.