Here's an alternate version of Muskie's retroactive 8-team playoff wherein from 1998-2013, the playoffs spots go to the 6 AQ conference champs plus the 2 remaining teams with the highest BCS rank. There is no specific rep from the non-AQ conferences. In 2014, the CFP replaces the BCS and the G5 get a rep.



Playoff Appearances
13: Ohio St.
12: Oklahoma
11: Alabama
9: Florida St.
8: USC
6: Florida, Oregon, Stanford, Virginia Tech, Wisconsin
5: Clemson, Georgia, LSU
4: Kansas St., Miami-FL, Michigan, Texas
3: Auburn, Penn St., TCU, UCF, Washington, West Virginia
2: Baylor, Cincinnati, Iowa, Louisville, Michigan St., Nebraska, Notre Dame, Tennessee
1: Boise St., California, Colorado, Connecticut, Georgia Tech, Houston, Illinois, Maryland, Mississippi St., Missouri, Oklahoma St., Oregon St., Pittsburgh, Purdue, Syracuse, Texas A&M, UCLA, Wake Forest, Washington St., WMU

(08-23-2019 10:45 PM)Nerdlinger Wrote:  Here's an alternate version of Muskie's retroactive 8-team playoff wherein from 1998-2013, the playoffs spots go to the 6 AQ conference champs plus the 2 remaining teams with the highest BCS rank. There is no specific rep from the non-AQ conferences. In 2014, the CFP replaces the BCS and the G5 get a rep.



Playoff Appearances
13: Ohio St.
12: Oklahoma
11: Alabama
9: Florida St.
8: USC
6: Florida, Oregon, Stanford, Virginia Tech, Wisconsin
5: Clemson, Georgia, LSU
4: Kansas St., Miami-FL, Michigan, Texas
3: Auburn, Penn St., TCU, UCF, Washington, West Virginia
2: Baylor, Cincinnati, Iowa, Louisville, Michigan St., Nebraska, Notre Dame, Tennessee
1: Boise St., California, Colorado, Connecticut, Georgia Tech, Houston, Illinois, Maryland, Mississippi St., Missouri, Oklahoma St., Oregon St., Pittsburgh, Purdue, Syracuse, Texas A&M, UCLA, Wake Forest, Washington St., WMU

The point was to simulate 5-1-2. You can’t do that if what you are really simulating is 6-2 and that’s not what we are aiming for.

(08-23-2019 10:56 PM)Fighting Muskie Wrote:  The point was to simulate 5-1-2. You can’t do that if what you are really simulating is 6-2 and that’s not what we are aiming for.

The thing of it is, there were 6 power conferences then, so to arbitrarily exclude the Big East after 2003 doesn't really reflect the facts on the ground.

And who is "we"?

(08-23-2019 10:45 PM)Nerdlinger Wrote:  Here's an alternate version of Muskie's retroactive 8-team playoff wherein from 1998-2013, the playoffs spots go to the 6 AQ conference champs plus the 2 remaining teams with the highest BCS rank. There is no specific rep from the non-AQ conferences. In 2014, the CFP replaces the BCS and the G5 get a rep.



Playoff Appearances
13: Ohio St.
12: Oklahoma
11: Alabama
9: Florida St.
8: USC
6: Florida, Oregon, Stanford, Virginia Tech, Wisconsin
5: Clemson, Georgia, LSU
4: Kansas St., Miami-FL, Michigan, Texas
3: Auburn, Penn St., TCU, UCF, Washington, West Virginia
2: Baylor, Cincinnati, Iowa, Louisville, Michigan St., Nebraska, Notre Dame, Tennessee
1: Boise St., California, Colorado, Connecticut, Georgia Tech, Houston, Illinois, Maryland, Mississippi St., Missouri, Oklahoma St., Oregon St., Pittsburgh, Purdue, Syracuse, Texas A&M, UCLA, Wake Forest, Washington St., WMU

That's 51 schools
7 of the 10 in the Big 12 (WVU in BE days)
10 of 14 in the Big 10 (NE in Big 12, MD in ACC)
9 of 14 in SEC (MO & A&M in Big 12)
9 of 14 in ACC (Miami, Syracuse, Pitt, UL in BE)
9 of 12 in Pac (CU in Big 12)
1 MAC, 1 MWC, 1 Independent, 4 AAC (UConn, Cin in BE)

Pretty similar % in all the P5 (64-75%) when you count participation in their prior conferences.

(01-05-2019 12:11 PM)Fighting Muskie Wrote:  Here is a look at what a 5-1-2 8-team playoff would look like dating back to 2004, the first season that the Big East was without Miami and Virginia Tech. I used the current NY6 rotation. When a bowl is hosting a semi-final they'd also host a bowl on or around New Years Day featuring 2 of the 4 highest ranked teams not included in the playoffs. (* = at large selection)

2004
Rose: 1 USC vs 13 Mich
Orange: 2 Oklahoma vs 8 VT
Peach: 3 Auburn vs 6 Utah
Sugar: 4 Texas* vs 5 Cal*
Next 4 Out--
Peach: 7 Georgia vs 10 L'ville
Fiesta: 9 Boise St vs 11 LSU

2005
Fiesta: 1 USC vs 22 Florida St
Cotton: 2 Texas vs 11 WVU
Orange: 3 Penn St vs 7 Georgia
Peach: 4 Ohio St* vs 5 Oregon*
Next 4 Out--
Rose: 6 Notre Dame vs 8 Miami
Sugar: 9 Auburn vs 10 VT

2006
Rose: 1 Ohio St vs 14 WF
Peach: 2 Florida vs 10 Oklahoma
Fiesta: 3 Mich* vs 6 Louisville
Sugar: 4 LSU* vs 5 USC
Next 4 Out--
Orange: 7 Wisconsin vs 9 Auburn
Cotton: 8 Boise St vs 11 Notre Dame

2007
Rose: 1 Ohio St vs 9 WVU
Sugar: 2 LSU vs 7 USC
Orange: 3 VT vs 6 Mizz*
Cotton: 4 Oklahoma vs 5 Georgia*
Next 4 Out--
Peach: 8 Kansas vs 12 Florida
Fiesta: 10 Hawaii vs 11 Ariz St

2008
Cotton: 1 Oklahoma vs 19 VT
Orange: 2 Florida vs 8 Penn St
Fiesta: 3 Texas* vs 6 Utah
Peach: 4 Alabama* vs 5 USC
Next 4 Out--
Rose: 7 Texas Tech vs 10 Ohio St
Sugar: 9 Boise St vs 11 TCU

2009
Peach: 1 Alabama vs 9 GT
Sugar: 2 Texas vs 8 Ohio St
Fiesta: 3 Cincy vs 7 Oregon
Rose: 4 TCU* vs 5 Florida*
Next 4 Out--
Cotton: 6 Boise St vs 12 LSU
Orange: 10 Iowa vs 11 VT

2010
Orange: 1 Auburn vs 13 VT
Rose: 2 Oregon vs 7 Oklahoma
Cotton: 3 TCU vs 6 Ohio St*
Sugar: 4 Stanford* vs 5 Wisconsin
Next 4 Out--
Peach: 8 Arkansas vs 9 Mich St
Fiesta: 10 Boise St vs 11 LSU

2011
Cotton: 1 LSU vs 15 Clemson
Peach: 2 Alabama* vs 10 Wisconsin
Orange: 3 Okla St vs 7 Boise St
Fiesta: 4 Stanford vs 5 Oregon*
Next 4 Out--
Sugar: 6 Arkansas vs 8 Kan St
Rose: 9 S Car vs 11 VT

2012
Fiesta: 1 Notre Dame* vs NR Wisc (12-0 Ohio St and Penn St faced bowl bans)
Sugar: 2 Alabama vs 15 NIU
Peach: 3 Florida* vs 12 Florida St
Rose: 4 Oregon vs 5 Kan St
Next 4 Out--
Cotton: 6 Stanford vs 9 Texas A&M
Orange: 7 Georgia vs 8 LSU

2013
Orange: 1 Florida St vs 15 UCF
Cotton: 2 Auburn vs 7 Ohio St*
Sugar: 3 Alabama* vs 6 Baylor
Rose: 4 Mich St vs 5 Stanford
Next 4 Out--
Fiesta: 8 Mizz vs 10 Oregon
Peach: 9 S Car vs 11 Oklahoma

2014
Peach: 1 Alabama vs 20 Boise St
Fiesta: 2 Oregon vs 7 Miss St*
Orange: 3 Florida St vs 6 TCU
Cotton: 4 Ohio St 5 Baylor*
Next 4 Out--
Rose: 8 Mich St vs 10 Ariz
Sugar: 9 Ole Miss vs 11 Kan St

2015
Peach: 1 Clemson vs 18 Houston
Sugar: 2 Alabama vs 7 Ohio St*
Rose: 3 Mich St vs 6 Stanford
Fiesta: 4 Oklahoma vs 5 Iowa*
Next 4 Out--
Orange: 8 Notre Dame vs 9 Florida St
Cotton: 10 UNC vs 11 TCU

2016
Sugar: 1 Alabama vs 15 WMU
Orange: 2 Clemson vs 7 Oklahoma
Cotton: 3 Ohio St* vs Mich*
Rose: 4 Wash vs 5 Iowa
Next 4 Out--
Fiesta: 8 Wisconsin vs 9 USC
Peach: 10 Colorado vs 11 Florida St

2017
Orange: 1 Clemson vs 12 UCF
Cotton: 2 Oklahoma vs 11 Wash
Peach: 3 Georgia* vs 6 Wisconsin*
Fiesta: 4 Alabama vs 5 Ohio St
Next 4 Out--
Sugar: 7 Auburn vs 10 Miami
Rose: 8 USC vs 9 Penn St

2018
Sugar: 1 Alabama vs 9 Wash
Peach: 2 Clemson vs 8 UCF
Rose: 3 Notre Dame* vs 6 Ohio St
Fiesta: 4 Oklahoma vs 5 Georgia*
Next 4 Out--
Orange: 7 Mich vs 10 Florida
Cotton: 11 LSU vs 12 Penn St

For those who don’t like that I simply downgraded the BE to a G5 league I’ve gone back through and tried to approximate how things would have went if during the BCS years the conferences were aligned as they are now:

2004: With 6 Utah in the PAC 12 (they lose to PAC 12 champ USC in the reg season) the G5 spot instead goes to 9 Boise St (11-0). Utah still goes to the playoffs when Cal loses to USC in the PAC 12 CCG and they slide in the standings

2005: 11 WVU is in the Big 12 and they don’t get the G5 spot. NR Boise St (9-3) get the G5 spot in a rough year for the G5

2006: 6 Louisville wins the ACC (not the BE) and 14 WF does not; G5 spot goes to 8 Boise St (12-0)

2007: 9 WVU comes in 3rd in the Big 12 standings; Oklahoma upsets Kansas (not Missouri) in the B12 CCG; 10 Hawaii (12-0) gets the G5 spot

2008: As in 2004, 6 Utah loses to USC in the regular season; the G5 slot goes to 9 Boise St (12-0)

2009: 3 Cincinnati still gets the G5 spot; TCU loses the Big 12 CCG to Texas and fall out of playoff contention opening the door for 10 Boise St (11-1) to claim the first ever at-large berth for a G5

2010: 3 TCU defeats 7 Oklahoma in the Big 12 CCG and Oklahoma misses the playoffs; 10 Boise St (11-1) grabs the G5 spot

2011: No Changes; 7 Boise St (11-1) still represents the G5

2012: Rutgers wins the Big 10 East but falls to Nebraska in the CCG and the Big 10 sends a ranked team to the playoffs; 15 NIU still represents the G5

2013: No Changes; 15 UCF still represents the G5

The premise of this threat is crap.

First off, there were 6 power conferences from 2004-2012, the Big East being the 6th. So no school from the Big East would qualify as "access".

The MWC would have nearly every access from 2004-12, thanks to having TCU and Boise State, as well as BYU and Utah.

You are counting schools like Pitt, WV, Syracuse, Louisville, Rutgers as "G5" which is absurd. The only schools that changed status after 2012 were Utah and TCU joining power conferences with USF, UConn and Cincy getting left out.

Stupid concept.

(08-24-2019 04:36 PM)Stugray2 Wrote:  The premise of this threat is crap.

First off, there were 6 power conferences from 2004-2012, the Big East being the 6th. So no school from the Big East would qualify as "access".

The MWC would have nearly every access from 2004-12, thanks to having TCU and Boise State, as well as BYU and Utah.

You are counting schools like Pitt, WV, Syracuse, Louisville, Rutgers as "G5" which is absurd. The only schools that changed status after 2012 were Utah and TCU joining power conferences with USF, UConn and Cincy getting left out.

Stupid concept.

And the gray cloud shows up to try and ruin another thread.

Let me teach you things about science and math. I’m concerned that these topics might not have covered at your alma mater and they may have given you a phony degree.

If you are trying to test a theory you need to have a data set to examine the bigger the better. You and others have been trying to argue straight 8 on a 5 year data set which in the world of statistics is frankly a meaningless joke because it’s no where near the size you need to draw any useful conclusions.

You’ve essentially flipped 5 coins and had them all land heads and then declared that coins will always land heads.

Since a larger data set is required we need to look at more seasons. But when you are running an experiment conditions need to be the same—i.e. we need all the schools to be in the exact same conferences as they are now. Since those conditions are not available it becomes necessary to run a statistical simulation of how schools would have performed had they played each season in their current conferences.

Don’t jump to call something crap because you don’t understand it.

(08-24-2019 05:06 PM)Fighting Muskie Wrote:  Let me teach you things about science and math. I’m concerned that these topics might not have covered at your alma mater and they may have given you a phony degree.

If you are trying to test a theory you need to have a data set to examine the bigger the better. You and others have been trying to argue straight 8 on a 5 year data set which in the world of statistics is frankly a meaningless joke because it’s no where near the size you need to draw any useful conclusions.

As I tried to explain earlier, the bolded part is true *only if the additional data used to make the set larger is relevant*. If it's not, then it is certainly NOT better to include that extra data because that pollutes it. Better to have a small pool of relevant data than a large pool that is contaminated with bad data. It's like if you are in a desert and very thirsty, better to have a half gallon of fresh water than a full gallon of salt water.

And as I've explained your BCS-years data points are salt water.

You can use statistical analysis and computer generation to create a predictive data set. It’s used in all sorts of fields, including sports. Billy Bean? Moneyball?

My methods were somewhat crude but an advanced statistician could, with relative ease, predict let’s say Louisville, Pitt, Syracuse, and BC would have performed in 2004 had they been in the ACC and playing an ACC schedule. The same goes with any of the other teams who have jumped around since then.

We all know that there were 6 big conferences in 2004. Analytics can tell us how teams would have performed if there were 5.

(08-24-2019 05:49 PM)quo vadis Wrote:  

(08-24-2019 05:06 PM)Fighting Muskie Wrote:  Let me teach you things about science and math. I’m concerned that these topics might not have covered at your alma mater and they may have given you a phony degree.

If you are trying to test a theory you need to have a data set to examine the bigger the better. You and others have been trying to argue straight 8 on a 5 year data set which in the world of statistics is frankly a meaningless joke because it’s no where near the size you need to draw any useful conclusions.

As I tried to explain earlier, the bolded part is true *only if the additional data used to make the set larger is relevant*. If it's not, then it is certainly NOT better to include that extra data because that pollutes it. Better to have a small pool of relevant data than a large pool that is contaminated with bad data. It's like if you are in a desert and very thirsty, better to have a half gallon of fresh water than a full gallon of salt water.

And as I've explained your BCS-years data points are salt water.

It's ok to do it this way as a "what if" for a message board, a way to mark time in the CFB offseason.

An actual statistical analysis would have to find a way to reliably take into account the fact that many of the outcomes of games in the last 20 years, particularly late season games, would have been much different if all teams were playing in the framework of an 8-team playoff.

Some teams with a chance at an 8-team playoff would be more motivated and play better. Some teams would fall victim to nerves if a playoff spot was on the line. Some teams with a chance to play spoiler and knock an opponent out of at-large consideration would be far more motivated in November than if they were merely playing out the string in a disappointing season. These factors would have changed the outcome of many games over a 20-year period, and none of that is taken into account by simply looking at data from years when there was only a 2-team BCS title game or even a 4-team playoff.

(08-24-2019 04:36 PM)Stugray2 Wrote:  The premise of this threat is crap.

First off, there were 6 power conferences from 2004-2012, the Big East being the 6th. So no school from the Big East would qualify as "access".

The MWC would have nearly every access from 2004-12, thanks to having TCU and Boise State, as well as BYU and Utah.

You are counting schools like Pitt, WV, Syracuse, Louisville, Rutgers as "G5" which is absurd. The only schools that changed status after 2012 were Utah and TCU joining power conferences with USF, UConn and Cincy getting left out.

Stupid concept.

You do understand that the playoff would be going forward, and not truly retroactively, right?

Maybe your understanding of how it works is crap. It is not rocket science

Time for a bump— 5-1-2 is a hot topic once more

(12-16-2019 10:44 AM)Fighting Muskie Wrote:  Time for a bump— 5-1-2 is a hot topic once more

3 times in your analysis you had a #4 OU vs. #5 UGA. Given the Rose Bowl a couple years ago when they were 2 and 3, it seems like a pretty good pairing.

(12-16-2019 12:45 PM)bullet Wrote:  

(12-16-2019 10:44 AM)Fighting Muskie Wrote:  Time for a bump— 5-1-2 is a hot topic once more

3 times in your analysis you had a #4 OU vs. #5 UGA. Given the Rose Bowl a couple years ago when they were 2 and 3, it seems like a pretty good pairing.

Oklahoma would have had a total of 10 appearances in the field of 8 and Georgia would have had 5. It’s a little uncanny that they paired as 4 and 5 in three separate years, both times with Oklahoma at 4 and the Big 12 champ facing at large Georgia.

(12-16-2019 02:32 PM)Fighting Muskie Wrote:  

(12-16-2019 12:45 PM)bullet Wrote:  

(12-16-2019 10:44 AM)Fighting Muskie Wrote:  Time for a bump— 5-1-2 is a hot topic once more

3 times in your analysis you had a #4 OU vs. #5 UGA. Given the Rose Bowl a couple years ago when they were 2 and 3, it seems like a pretty good pairing.

Oklahoma would have had a total of 10 appearances in the field of 8 and Georgia would have had 5. It’s a little uncanny that they paired as 4 and 5 in three separate years, both times with Oklahoma at 4 and the Big 12 champ facing at large Georgia.

There is the flaw in this approach. They were ranked 4 and 5 because they were being positioned as acceptable recipients for that final slot by the CFP committee (whose rankings vary widely from those of the AP). If an 8 team model were being considered I would posit that the placement of 8 & 9 would be handled very similarly and that those finishing 4 and 5 would be much more reflective of their actual standing for seeding purposes verses entry purposes.

The CFP would probably be a little more delicate in the way they evaluate spots 6-9 as they’d have more bearing on who makes the championship bracket field and who gets a high dollar consolation game, particularly when it comes to the at large teams.

(12-16-2019 03:48 PM)JRsec Wrote:  

(12-16-2019 02:32 PM)Fighting Muskie Wrote:  

(12-16-2019 12:45 PM)bullet Wrote:  

(12-16-2019 10:44 AM)Fighting Muskie Wrote:  Time for a bump— 5-1-2 is a hot topic once more

3 times in your analysis you had a #4 OU vs. #5 UGA. Given the Rose Bowl a couple years agonizing when they were 2 and 3, it seems like a pretty good pairing.

Oklahoma would have had a total of 10 appearances in the field of 8 and Georgia would have had 5. It’s a little uncanny that they paired as 4 and 5 in three separate years, both times with Oklahoma at 4 and the Big 12 champ facing at large Georgia.

There is the flaw in this approach. They were ranked 4 and 5 because they were being positioned as acceptable recipients for that final slot by the CFP committee (whose rankings vary widely from those of the AP). If an 8 team model were being considered I would posit that the placement of 8 & 9 would be handled very similarly and that those finishing 4 and 5 would be much more reflective of their actual standing for seeding purposes verses entry purposes.

Using a computer would be best. A person, regardless of what they intend, will naturally bend their rankings at the break-off point of a playoff field.

(12-16-2019 09:01 PM)BePcr07 Wrote:  

(12-16-2019 03:48 PM)JRsec Wrote:  

(12-16-2019 02:32 PM)Fighting Muskie Wrote:  

(12-16-2019 12:45 PM)bullet Wrote:  

(12-16-2019 10:44 AM)Fighting Muskie Wrote:  Time for a bump— 5-1-2 is a hot topic once more

3 times in your analysis you had a #4 OU vs. #5 UGA. Given the Rose Bowl a couple years agonizing when they were 2 and 3, it seems like a pretty good pairing.

Oklahoma would have had a total of 10 appearances in the field of 8 and Georgia would have had 5. It’s a little uncanny that they paired as 4 and 5 in three separate years, both times with Oklahoma at 4 and the Big 12 champ facing at large Georgia.

There is the flaw in this approach. They were ranked 4 and 5 because they were being positioned as acceptable recipients for that final slot by the CFP committee (whose rankings vary widely from those of the AP). If an 8 team model were being considered I would posit that the placement of 8 & 9 would be handled very similarly and that those finishing 4 and 5 would be much more reflective of their actual standing for seeding purposes verses entry purposes.

Using a computer would be best. A person, regardless of what they intend, will naturally bend their rankings at the break-off point of a playoff field.

And every computer program is only as good as the bias of their programmer.

(12-16-2019 09:09 PM)JRsec Wrote:  

(12-16-2019 09:01 PM)BePcr07 Wrote:  

(12-16-2019 03:48 PM)JRsec Wrote:  

(12-16-2019 02:32 PM)Fighting Muskie Wrote:  

(12-16-2019 12:45 PM)bullet Wrote:  3 times in your analysis you had a #4 OU vs. #5 UGA. Given the Rose Bowl a couple years agonizing when they were 2 and 3, it seems like a pretty good pairing.

Oklahoma would have had a total of 10 appearances in the field of 8 and Georgia would have had 5. It’s a little uncanny that they paired as 4 and 5 in three separate years, both times with Oklahoma at 4 and the Big 12 champ facing at large Georgia.

There is the flaw in this approach. They were ranked 4 and 5 because they were being positioned as acceptable recipients for that final slot by the CFP committee (whose rankings vary widely from those of the AP). If an 8 team model were being considered I would posit that the placement of 8 & 9 would be handled very similarly and that those finishing 4 and 5 would be much more reflective of their actual standing for seeding purposes verses entry purposes.

Using a computer would be best. A person, regardless of what they intend, will naturally bend their rankings at the break-off point of a playoff field.

And every computer program is only as good as the bias of their programmer.

True but the parameters are preset and generally predictable while a person can have an unexpected change in mind.

(04-18-2019 08:35 AM)goofus Wrote:  I always thought 2008 was a good argument for an 8-team playoff because there were so many teams with a legitimate claim to deserving a shot at the National championship game.

The Big 12 was especially tricky because it had 3 teams that all finished in 11-1 in 3-way tie for first. For that reason, I believe a selection committee would put 3 teams from the Big 12 in the 8 team field and take TT over Bama.

They’d stick with Bama IMO. Tech lost by over 40 points in their loss while Bama’s Loss to Tebow was much more forgiveable