CSNbbs
CFB's Philosopher's Stone (UPDATE 1/24, Full Draft w/Final Rankings) - Printable Version

+- CSNbbs (https://csnbbs.com)
+-- Forum: Active Boards (/forum-769.html)
+--- Forum: Lounge (/forum-564.html)
+---- Forum: College Sports and Conference Realignment (/forum-637.html)
+---- Thread: CFB's Philosopher's Stone (UPDATE 1/24, Full Draft w/Final Rankings) (/thread-1022817.html)

Pages: 1 2


CFB's Philosopher's Stone (UPDATE 1/24, Full Draft w/Final Rankings) - Crayton - 12-31-2025 11:33 PM

So, one of the datums missing from my rankings is a SOS metric. Should be pretty easy; probably just average out a team's opponent strengths.

In the journey of creating it, I found that you can actually make a rather human-computable SOR rating for teams. For the most part, each win in a SOR metric has the same marginal value. You can then use a SOS metric to set the baseline for each team and go up or down from there depending on a team's record.

For reference, here are my SOR values for a few select teams, given different records. Note how the slopes get steeper at the ends but are generally linear. Note also that these graphs are 'mostly' parallel, though some teams do get greater marginal value from each win.
[Image: attachment.php?aid=12279]

If we took these and assumed them to be both linear and parallel, we could approximate a SOR with a team's record + a SOS offset. I 'think' whatever is lost in these assumptions 'could' be made up for in the fact that the ranking is now human-computable (or, human-verifiable).

Here is the pre-CCG Top 25 using this method. I have added tie-breakers because I think this is the key part missing from most/all computer rankings. Tie-breakers make the "standings" more natural and allow them to be useful for conferences to use them too (Miami>Duke, UNM>UNLV). The right portion of the chart shows the Top 5 in each conference, using these standings.
[Image: attachment.php?aid=12280]

Pretty good? Aligns closely on the national and conference level. And given teams' SRS, you only need 2 formulas to calculate everyone's Adjusted Wins. This should score high in making national/conference "standings" that are popularly accessible. No more waiting for rankings on Sunday (or Tuesday). No more wondering how a group of humans will weigh certain wins or losses across the previous 3 months.

The tie-breakers are really what get this across the finishline for me. For most computer rankings, "math" ends up determining who earns the post-season reward; a few decimals up or down can make a big difference. But if there is head-to-head, that should be determinative; if one team has a better record (coupled with a better power-rating, SRS), then that too should be determinative.

Opportunities / Questions?

1) Because not all teams play 12 games (Army-Navy, CCG teams, mid-season standings), Win% would be more accurate than Win Count. On the other hand, "Wins" are more discrete and fan-accessible, I think. I'm showing the pre-CCG standings above in part because I didn't have a pretty way to "adjust" win count for teams playing other than 12 games. Would "adjusted Win %" still be as accessible as "Adjusted Wins"?

Another option is to +/- half wins for CCGs (roughly a .2 bonus to both teams) and teams who play less than 12 games (a .3 penalty if the 12th opponent would have been roughly equally matched). You can see I gave Navy 0.50 wins in the example above. The bonus and penalty may be more appropriate than straight Win%.

2) The formula I have posted for SRS is the internet-available definition, but the values I've listed next to each team in the chart have actually undergone 3 additional transformations that I have added over the years to make the final rankings "look" accurate. This off-season, I'll see if any of these can be consolidated and hopefully end up with 1 transformation without sacrificing ranking accuracy. To keep this chart human-computable, a column for that transformation should be added; hopefully the formula for that transformation is easily read. Better yet, I hope it can be rolled into the SRS calculation.

3) I threw the "-1" into the SOS calculation to balance the fact that I am only looking at the 8 most difficult opponents on a team's schedule. Across all 136 teams, SOS is pretty close to 0.0. But the "-1" doesn't have any impact on the standings. Is it nice to be able to see that Virginia, for example, has a pretty 'mid' schedule close to 0.0, or would it be better to save the extra subtraction step and only rarely come across teams with negative SOS adjustments?


RE: CFB's Philosopher's Stone ("real" SOR rankings that fans can calculate o... - ken d - 01-05-2026 08:41 AM

So, bottom line, you are saying the selection committee got the teams exactly right.

Edit: It's taken me a while to figure out exactly what you're saying, and I'm still not sure I understand it. I have no idea what SRS stands for, and an internet search came up empty.

But now it looks to me that you would have selected Notre Dame ahead of Miami, but neither would have made the CFP field because BYU had an adjusted win total better than both. If I am reading this right, you would have had three B1G teams with a first round bye and the ACC would have been shut out entirely. Your bracket would have looked like this:

8 Texas A&M vs 9 Oklahoma --> 1 Ohio State
5 Alabama vs 12 James Madison --> 4 Oregon

6 Ole Miss vs 11 Tulane --> 3 Georgia
7 BYU vs 10 Texas Tech --> 2 Indiana

Is that correct? Would the public have bought into that result?


RE: CFB's Philosopher's Stone ("real" SOR rankings that fans can calculate o... - Crayton - 01-05-2026 04:12 PM

(01-05-2026 08:41 AM)ken d Wrote:  So, bottom line, you are saying the selection committee got the teams exactly right.

Edit: It's taken me a while to figure out exactly what you're saying, and I'm still not sure I understand it. I have no idea what SRS stands for, and an internet search came up empty.

But now it looks to me that you would have selected Notre Dame ahead of Miami, but neither would have made the CFP field because BYU had an adjusted win total better than both. If I am reading this right, you would have had three B1G teams with a first round bye and the ACC would have been shut out entirely. Your bracket would have looked like this:

8 Texas A&M vs 9 Oklahoma --> 1 Ohio State
5 Alabama vs 12 James Madison --> 4 Oregon

6 Ole Miss vs 11 Tulane --> 3 Georgia
7 BYU vs 10 Texas Tech --> 2 Indiana

Is that correct? Would the public have bought into that result?

You can Google “SRS Football Rankings.” It stands for Simple Rating System and essentially just that formula shown. The actual values in the graphic I shared are modified a bit and I will work this Summer to get a cleaner formula that includes those modifications.

The takeaway here is not necessarily whether these are the correct teams. The teams should make sense and those probably are the correct teams… pre-CCGs.

The takeaway is if something ‘like’ this could replace all those opaque computer formulas and media polls, making conference and national races more accessible to fans.

Teams’ SOS adjustments will be more-or-less “known” after 4 weeks, allowing fans to track which teams are on track for a CCG or playoff berth. In the ACC, for example, Georgia Tech and Pittsburgh fans will get the benefit of playing a stronger schedule and ‘know’ they’ll win the tiebreaker over Virginia if they finish with the same record.

Not all 10-2 records are created equal, even in the same conference, and right now we all wait until Sunday to find out which teams get postseason berths. Like the NFL and other sports, discrete standings would allow this information to be known beforehand.


RE: CFB's Philosopher's Stone ("real" SOR rankings that fans can calculate o... - ken d - 01-05-2026 04:52 PM

(01-05-2026 04:12 PM)Crayton Wrote:  
(01-05-2026 08:41 AM)ken d Wrote:  So, bottom line, you are saying the selection committee got the teams exactly right.

Edit: It's taken me a while to figure out exactly what you're saying, and I'm still not sure I understand it. I have no idea what SRS stands for, and an internet search came up empty.

But now it looks to me that you would have selected Notre Dame ahead of Miami, but neither would have made the CFP field because BYU had an adjusted win total better than both. If I am reading this right, you would have had three B1G teams with a first round bye and the ACC would have been shut out entirely. Your bracket would have looked like this:

8 Texas A&M vs 9 Oklahoma --> 1 Ohio State
5 Alabama vs 12 James Madison --> 4 Oregon

6 Ole Miss vs 11 Tulane --> 3 Georgia
7 BYU vs 10 Texas Tech --> 2 Indiana

Is that correct? Would the public have bought into that result?

You can Google “SRS Football Rankings.” It stands for Simple Rating System and essentially just that formula shown. The actual values in the graphic I shared are modified a bit and I will work this Summer to get a cleaner formula that includes those modifications.

The takeaway here is not necessarily whether these are the correct teams. The teams should make sense and those probably are the correct teams… pre-CCGs.

The takeaway is if something ‘like’ this could replace all those opaque computer formulas and media polls, making conference and national races more accessible to fans.

Teams’ SOS adjustments will be more-or-less “known” after 4 weeks, allowing fans to track which teams are on track for a CCG or playoff berth. In the ACC, for example, Georgia Tech and Pittsburgh fans will get the benefit of playing a stronger schedule and ‘know’ they’ll win the tiebreaker over Virginia if they finish with the same record.

Not all 10-2 records are created equal, even in the same conference, and right now we all wait until Sunday to find out which teams get postseason berths. Like the NFL and other sports, discrete standings would allow this information to be known beforehand.

How does anyone know what their team's SOS will be after only four weeks? And, even if they could know it, what good would it do them? Few teams have yet played a meaningful number of conference games, and teams that look like sure playoff bets after four weeks have been known to fade in the stretch. Just look at LSU this year.

I could see trying to apply something like this after all CCGs have been played, but not before. So what's the point?


RE: CFB's Philosopher's Stone ("real" SOR rankings that fans can calculate o... - Crayton - 01-05-2026 06:09 PM

(01-05-2026 04:52 PM)ken d Wrote:  
(01-05-2026 04:12 PM)Crayton Wrote:  
(01-05-2026 08:41 AM)ken d Wrote:  So, bottom line, you are saying the selection committee got the teams exactly right.

Edit: It's taken me a while to figure out exactly what you're saying, and I'm still not sure I understand it. I have no idea what SRS stands for, and an internet search came up empty.

But now it looks to me that you would have selected Notre Dame ahead of Miami, but neither would have made the CFP field because BYU had an adjusted win total better than both. If I am reading this right, you would have had three B1G teams with a first round bye and the ACC would have been shut out entirely. Your bracket would have looked like this:

8 Texas A&M vs 9 Oklahoma --> 1 Ohio State
5 Alabama vs 12 James Madison --> 4 Oregon

6 Ole Miss vs 11 Tulane --> 3 Georgia
7 BYU vs 10 Texas Tech --> 2 Indiana

Is that correct? Would the public have bought into that result?

You can Google “SRS Football Rankings.” It stands for Simple Rating System and essentially just that formula shown. The actual values in the graphic I shared are modified a bit and I will work this Summer to get a cleaner formula that includes those modifications.

The takeaway here is not necessarily whether these are the correct teams. The teams should make sense and those probably are the correct teams… pre-CCGs.

The takeaway is if something ‘like’ this could replace all those opaque computer formulas and media polls, making conference and national races more accessible to fans.

Teams’ SOS adjustments will be more-or-less “known” after 4 weeks, allowing fans to track which teams are on track for a CCG or playoff berth. In the ACC, for example, Georgia Tech and Pittsburgh fans will get the benefit of playing a stronger schedule and ‘know’ they’ll win the tiebreaker over Virginia if they finish with the same record.

Not all 10-2 records are created equal, even in the same conference, and right now we all wait until Sunday to find out which teams get postseason berths. Like the NFL and other sports, discrete standings would allow this information to be known beforehand.

How does anyone know what their team's SOS will be after only four weeks? And, even if they could know it, what good would it do them? Few teams have yet played a meaningful number of conference games, and teams that look like sure playoff bets after four weeks have been known to fade in the stretch. Just look at LSU this year.

I could see trying to apply something like this after all CCGs have been played, but not before. So what's the point?

Any ranking can be done after CCGs and a good portion can give rankings that are just as acceptable. The challenge I’d like to tackle is getting a ranking system that CAN be tracked week-to-week.

Instead of Texas fans saying they need to beat A&M, hope for “some” upsets ahead of them, and enlist their coach to stump the media for enhanced standing, Texas fans should be able to say “we need to beat A&M and have 3 of the following 6 teams get upset.”

Can we know SOS after only 4 weeks?

That is a great question. I should run the OP standings at that point in the season and see how valid my presumption is. As you note, teams like LSU (and FSU) can see dramatic dips in fortunes after impressive season starts. There are a few mitigating factors though:

1) most teams ‘are’ as strong at the beginning of the year as they are at the end.

2) the teams which experience upsets after 4 weeks are generally upset by conference-mates, having a balancing affect on other conference teams’ SOS.

3) any unexpected result is ultimately only 1 of 12 games on a team’s schedule and only 1 of 144 when looking at another team’s SOS. Any unexpected results will generally balance.

But the proof is in the pudding. Either tonight or this weekend, I’ll look to see how SOS changes week-to-week. If it changes radically, my purpose may be defeated.


RE: CFB's Philosopher's Stone ("real" SOR rankings that fans can calculate o... - Hokie Mark - 01-05-2026 09:51 PM

Methods like this often lead to false conclusions because SoS in itself is not nearly as important as people have been brainwashed to think it is. For instance, the OP's chart would imply that no way Miami beats Ohio State. That's demonstrably FALSE.

That's because people keep forgetting this vital fact which should be obvious to all ball-knowers: SoS is a measure of how strong your opponents were, NOT how strong your team is! The latter is really what we want to know, and SoS is just not that.

Is SoS part of a Power Rating calculation? Sure! But it's the power rating that matters, not the strength of schedule!

[/rant]


RE: CFB's Philosopher's Stone ("real" SOR rankings that fans can calculate o... - Crayton - 01-06-2026 05:52 AM

(01-05-2026 09:51 PM)Hokie Mark Wrote:  Methods like this often lead to false conclusions because SoS in itself is not nearly as important as people have been brainwashed to think it is. For instance, the OP's chart would imply that no way Miami beats Ohio State. That's demonstrably FALSE.

That's because people keep forgetting this vital fact which should be obvious to all ball-knowers: SoS is a measure of how strong your opponents were, NOT how strong your team is! The latter is really what we want to know, and SoS is just not that.

Is SoS part of a Power Rating calculation? Sure! But it's the power rating that matters, not the strength of schedule!

[/rant]

You may miss the point here. The point is not to rate teams by who would win on a neutral field in Arlington. The point is to rank teams by “Strength of Record”. And I am trying to calibrate the two components (SOS and Record) such that they can be easily added together to create that rating.


RE: CFB's Philosopher's Stone ("real" SOR rankings that fans can calculate o... - Dr Dawgma III - 01-06-2026 07:40 AM

Unfortunately, college football has moved away from objective decision making: division > conference > CFP

If we could rewind to 2011 era alignment and use the five power conferences, I believe even the top four conference champ bye method would have worked. Due to the SEC and B1G's predatory ways, the top programs are now concentrated in three conferences instead of five.


RE: CFB's Philosopher's Stone ("real" SOR rankings that fans can calculate on pen+paper) - Fighting Muskie - 01-06-2026 09:03 AM

Any particular reason why you’re only looking at the top 8 opponents for SOS? To me there’s a big difference between having your weakest opponents as teams you’ve got a 90%+ chance of beating vs having those teams be teams that you have ~70% chance of beating. Ignoring them rewards the cupcake eaters and I think the formula should punish teams for scheduling de facto wins.

I could see maybe only evaluating the top 11 regular season games to avoid having to determine the strengths of all the FCS rent-a-victims, but even then, you’re punishing the teams who scheduling 12 FBS programs.


RE: CFB's Philosopher's Stone ("real" SOR rankings that fans can calculate o... - ken d - 01-06-2026 10:22 AM

(01-05-2026 06:09 PM)Crayton Wrote:  
(01-05-2026 04:52 PM)ken d Wrote:  
(01-05-2026 04:12 PM)Crayton Wrote:  
(01-05-2026 08:41 AM)ken d Wrote:  So, bottom line, you are saying the selection committee got the teams exactly right.

Edit: It's taken me a while to figure out exactly what you're saying, and I'm still not sure I understand it. I have no idea what SRS stands for, and an internet search came up empty.

But now it looks to me that you would have selected Notre Dame ahead of Miami, but neither would have made the CFP field because BYU had an adjusted win total better than both. If I am reading this right, you would have had three B1G teams with a first round bye and the ACC would have been shut out entirely. Your bracket would have looked like this:

8 Texas A&M vs 9 Oklahoma --> 1 Ohio State
5 Alabama vs 12 James Madison --> 4 Oregon

6 Ole Miss vs 11 Tulane --> 3 Georgia
7 BYU vs 10 Texas Tech --> 2 Indiana

Is that correct? Would the public have bought into that result?

You can Google “SRS Football Rankings.” It stands for Simple Rating System and essentially just that formula shown. The actual values in the graphic I shared are modified a bit and I will work this Summer to get a cleaner formula that includes those modifications.

The takeaway here is not necessarily whether these are the correct teams. The teams should make sense and those probably are the correct teams… pre-CCGs.

The takeaway is if something ‘like’ this could replace all those opaque computer formulas and media polls, making conference and national races more accessible to fans.

Teams’ SOS adjustments will be more-or-less “known” after 4 weeks, allowing fans to track which teams are on track for a CCG or playoff berth. In the ACC, for example, Georgia Tech and Pittsburgh fans will get the benefit of playing a stronger schedule and ‘know’ they’ll win the tiebreaker over Virginia if they finish with the same record.

Not all 10-2 records are created equal, even in the same conference, and right now we all wait until Sunday to find out which teams get postseason berths. Like the NFL and other sports, discrete standings would allow this information to be known beforehand.

How does anyone know what their team's SOS will be after only four weeks? And, even if they could know it, what good would it do them? Few teams have yet played a meaningful number of conference games, and teams that look like sure playoff bets after four weeks have been known to fade in the stretch. Just look at LSU this year.

I could see trying to apply something like this after all CCGs have been played, but not before. So what's the point?

Any ranking can be done after CCGs and a good portion can give rankings that are just as acceptable. The challenge I’d like to tackle is getting a ranking system that CAN be tracked week-to-week.

Instead of Texas fans saying they need to beat A&M, hope for “some” upsets ahead of them, and enlist their coach to stump the media for enhanced standing, Texas fans should be able to say “we need to beat A&M and have 3 of the following 6 teams get upset.”

Can we know SOS after only 4 weeks?

That is a great question. I should run the OP standings at that point in the season and see how valid my presumption is. As you note, teams like LSU (and FSU) can see dramatic dips in fortunes after impressive season starts. There are a few mitigating factors though:

1) most teams ‘are’ as strong at the beginning of the year as they are at the end.

2) the teams which experience upsets after 4 weeks are generally upset by conference-mates, having a balancing affect on other conference teams’ SOS.

3) any unexpected result is ultimately only 1 of 12 games on a team’s schedule and only 1 of 144 when looking at another team’s SOS. Any unexpected results will generally balance.

But the proof is in the pudding. Either tonight or this weekend, I’ll look to see how SOS changes week-to-week. If it changes radically, my purpose may be defeated.

As a starting point to develop an algorithm that gives a clear picture throughout the season, you must first have an algorithm that produces a logical result at the point in the season when you have all the data you are going to get. That's after the CCGs.

This particular algorithm seems to not produce that. It has BYU ranked 7th ahead of Texas Tech at #11, despite the fact that Tech beat BYU twice convincingly, both late in the season. Tech would make the field because they were the third highest ranked conference champion. But if logic says BYU should be ranked below them, that is to say at #12, then Notre Dame would move up to #9 and replace BYU in the field of 12.

So, it seems you would need to tweak the algorithm in a way that resolves the BYU dilemma without creating a different dilemma. I think we have to consider the possibility that there is no algorithm that would do that with any consistency. If that is the case, then the flaw may be within us, the fans, for wanting a ranking system in the first place rather than some combination of conference standings and wild card games.


RE: CFB's Philosopher's Stone ("real" SOR rankings that fans can calculate o... - Crayton - 01-06-2026 02:57 PM

(01-06-2026 09:03 AM)Fighting Muskie Wrote:  Any particular reason why you’re only looking at the top 8 opponents for SOS? To me there’s a big difference between having your weakest opponents as teams you’ve got a 90%+ chance of beating vs having those teams be teams that you have ~70% chance of beating. Ignoring them rewards the cupcake eaters and I think the formula should punish teams for scheduling de facto wins.

I could see maybe only evaluating the top 11 regular season games to avoid having to determine the strengths of all the FCS rent-a-victims, but even then, you’re punishing the teams who scheduling 12 FBS programs.

A SOR metric naturally depends more on the difficulty of the opponents nearer your record. The difficulty of Boise last year or JMU this year going 12-1 is indicated most by who their #1 and #2 opponents were. Yes, opponents #9-#12 have an impact as well, but it is far less.

Of course assigning these values properly is a job for a computer. Assigning these values with a simple average can be done by a human and taking the Top 8 encompasses the value of the most pertinent opponents.

In the theoretical world where such a ranking as this is used, would the cliff at opponent #9 influence scheduling practices? Maybe, but strong scheduling practices would still have a positive impact. Texas would still benefit from scheduling Ohio State over Wyoming. And while Vanderbilt scheduling VT may seem like a waste, the Hokies have more upside than say the Zips. I don’t think scheduling strategies would be impacted.


RE: CFB's Philosopher's Stone ("real" SOR rankings that fans can calculate o... - Fighting Muskie - 01-06-2026 04:10 PM

(01-06-2026 02:57 PM)Crayton Wrote:  
(01-06-2026 09:03 AM)Fighting Muskie Wrote:  Any particular reason why you’re only looking at the top 8 opponents for SOS? To me there’s a big difference between having your weakest opponents as teams you’ve got a 90%+ chance of beating vs having those teams be teams that you have ~70% chance of beating. Ignoring them rewards the cupcake eaters and I think the formula should punish teams for scheduling de facto wins.

I could see maybe only evaluating the top 11 regular season games to avoid having to determine the strengths of all the FCS rent-a-victims, but even then, you’re punishing the teams who scheduling 12 FBS programs.

A SOR metric naturally depends more on the difficulty of the opponents nearer your record. The difficulty of Boise last year or JMU this year going 12-1 is indicated most by who their #1 and #2 opponents were. Yes, opponents #9-#12 have an impact as well, but it is far less.

Of course assigning these values properly is a job for a computer. Assigning these values with a simple average can be done by a human and taking the Top 8 encompasses the value of the most pertinent opponents.

In the theoretical world where such a ranking as this is used, would the cliff at opponent #9 influence scheduling practices? Maybe, but strong scheduling practices would still have a positive impact. Texas would still benefit from scheduling Ohio State over Wyoming. And while Vanderbilt scheduling VT may seem like a waste, the Hokies have more upside than say the Zips. I don’t think scheduling strategies would be impacted.

I still think that methodology is skewed in favor of teams who have a few really tough opponents but then pad their schedule with easy wins.

Imagine 2 teams have the same score when evaluating the top 8 opponents but team A’s bottom 4 are ranked 89, 92, 96, and 101 while team B’s bottom 4 are 120, 128, 134, and and an FCS. Do these teams really have the same schedule strength?


RE: CFB's Philosopher's Stone ("real" SOR rankings that fans can calculate o... - Hokie Mark - 01-06-2026 04:19 PM

(01-06-2026 05:52 AM)Crayton Wrote:  
(01-05-2026 09:51 PM)Hokie Mark Wrote:  Methods like this often lead to false conclusions because SoS in itself is not nearly as important as people have been brainwashed to think it is. For instance, the OP's chart would imply that no way Miami beats Ohio State. That's demonstrably FALSE.

That's because people keep forgetting this vital fact which should be obvious to all ball-knowers: SoS is a measure of how strong your opponents were, NOT how strong your team is! The latter is really what we want to know, and SoS is just not that.

Is SoS part of a Power Rating calculation? Sure! But it's the power rating that matters, not the strength of schedule!

[/rant]

You may miss the point here. The point is not to rate teams by who would win on a neutral field in Arlington. The point is to rank teams by “Strength of Record”. And I am trying to calibrate the two components (SOS and Record) such that they can be easily added together to create that rating.

To what end? What do you see that ranking used for?


RE: CFB's Philosopher's Stone ("real" SOR rankings that fans can calculate o... - Hokie Mark - 01-06-2026 04:26 PM

(01-06-2026 10:22 AM)ken d Wrote:  [quote='Crayton' pid='20506823' dateline='1767654580']
As a starting point to develop an algorithm that gives a clear picture throughout the season, you must first have an algorithm that produces a logical result at the point in the season when you have all the data you are going to get. That's after the CCGs.

This particular algorithm seems to not produce that. It has BYU ranked 7th ahead of Texas Tech at #11, despite the fact that Tech beat BYU twice convincingly, both late in the season. Tech would make the field because they were the third highest ranked conference champion. But if logic says BYU should be ranked below them, that is to say at #12, then Notre Dame would move up to #9 and replace BYU in the field of 12.

So, it seems you would need to tweak the algorithm in a way that resolves the BYU dilemma without creating a different dilemma. I think we have to consider the possibility that there is no algorithm that would do that with any consistency. If that is the case, then the flaw may be within us, the fans, for wanting a ranking system in the first place rather than some combination of conference standings and wild card games.

This speaks to my point. There's nothing illogical in saying "BYU played a stronger schedule than Texas Tech" and also "Texas Tech is better than BYU" -- those two things are different, like "higher" and "in front of" are different.

In my estimation, people throughout CFB (not just this thread) are spending WAY too much time and energy on calculating strength of schedule and/or record, which frankly, is not the really important question!

Power ratings aren't that hard to calculate - but they sometimes produce results that some people just don't like very much!


RE: CFB's Philosopher's Stone ("real" SOR rankings that fans can calculate o... - Crayton - 01-06-2026 08:01 PM

(01-06-2026 04:10 PM)Fighting Muskie Wrote:  
(01-06-2026 02:57 PM)Crayton Wrote:  
(01-06-2026 09:03 AM)Fighting Muskie Wrote:  Any particular reason why you’re only looking at the top 8 opponents for SOS? To me there’s a big difference between having your weakest opponents as teams you’ve got a 90%+ chance of beating vs having those teams be teams that you have ~70% chance of beating. Ignoring them rewards the cupcake eaters and I think the formula should punish teams for scheduling de facto wins.

I could see maybe only evaluating the top 11 regular season games to avoid having to determine the strengths of all the FCS rent-a-victims, but even then, you’re punishing the teams who scheduling 12 FBS programs.

A SOR metric naturally depends more on the difficulty of the opponents nearer your record. The difficulty of Boise last year or JMU this year going 12-1 is indicated most by who their #1 and #2 opponents were. Yes, opponents #9-#12 have an impact as well, but it is far less.

Of course assigning these values properly is a job for a computer. Assigning these values with a simple average can be done by a human and taking the Top 8 encompasses the value of the most pertinent opponents.

In the theoretical world where such a ranking as this is used, would the cliff at opponent #9 influence scheduling practices? Maybe, but strong scheduling practices would still have a positive impact. Texas would still benefit from scheduling Ohio State over Wyoming. And while Vanderbilt scheduling VT may seem like a waste, the Hokies have more upside than say the Zips. I don’t think scheduling strategies would be impacted.

I still think that methodology is skewed in favor of teams who have a few really tough opponents but then pad their schedule with easy wins.

Imagine 2 teams have the same score when evaluating the top 8 opponents but team A’s bottom 4 are ranked 89, 92, 96, and 101 while team B’s bottom 4 are 120, 128, 134, and and an FCS. Do these teams really have the same schedule strength?

The trick is that I am not really trying to measure Strength of Schedule but “portion of Strength of Record that is not attributable to Record” with greatest accuracy for Top 25 teams. ESPN’s “SOS” is similarly skewed toward the Top 25. “SOS” is simply an imminently available and publicly relatable term to describe this measure.

Reminder that I am trying to approximate those wavy intersecting lines in the OP with straight parallel lines. I used regression analysis to identify Top 8 as the best average for approximating SOR for Top 25 teams.

As the formulas are ‘okay’ approximations of “True” SOR, there must be tie breakers (within 1.0 Adjusted Win) to provide clarity at the edges. In your A-B example, while one team may have a slight edge over the other, SOR should not be the final say. Head-to-head is absent from most ranking algorithms though it ‘should’ be used in a case like this.

And, ultimately, any standing (even the NFL) sacrifices a perfectly accurate rank for a perfectly legible one. Miami didn’t cry foul when they got left out of the ACCCG. The Bucs didn’t cry foul when the Falcons convoluted their head-to-head edge over Carolina. Why? Because the scenarios were fully known beforehand.


RE: CFB's Philosopher's Stone ("real" SOR rankings that fans can calculate o... - Crayton - 01-06-2026 08:08 PM

(01-06-2026 04:26 PM)Hokie Mark Wrote:  
(01-06-2026 10:22 AM)ken d Wrote:  [quote='Crayton' pid='20506823' dateline='1767654580']
As a starting point to develop an algorithm that gives a clear picture throughout the season, you must first have an algorithm that produces a logical result at the point in the season when you have all the data you are going to get. That's after the CCGs.

This particular algorithm seems to not produce that. It has BYU ranked 7th ahead of Texas Tech at #11, despite the fact that Tech beat BYU twice convincingly, both late in the season. Tech would make the field because they were the third highest ranked conference champion. But if logic says BYU should be ranked below them, that is to say at #12, then Notre Dame would move up to #9 and replace BYU in the field of 12.

So, it seems you would need to tweak the algorithm in a way that resolves the BYU dilemma without creating a different dilemma. I think we have to consider the possibility that there is no algorithm that would do that with any consistency. If that is the case, then the flaw may be within us, the fans, for wanting a ranking system in the first place rather than some combination of conference standings and wild card games.

This speaks to my point. There's nothing illogical in saying "BYU played a stronger schedule than Texas Tech" and also "Texas Tech is better than BYU" -- those two things are different, like "higher" and "in front of" are different.

In my estimation, people throughout CFB (not just this thread) are spending WAY too much time and energy on calculating strength of schedule and/or record, which frankly, is not the really important question!

Power ratings aren't that hard to calculate - but they sometimes produce results that some people just don't like very much!

Power ratings and SOR ratings measure different things, yes. Texas Tech can be better than BYU while at the same time BYU has the more impressive SOR.

The reason I regularly choose SOR over a Power rating is because I think in circumstances like meting out post-season benefit (CCG, Playoff bid) we should reward teams who best accomplish the objective of the game: winning.

Once we start giving post season benefit to teams who score the most points or rack up the most yards or tally more first downs, we are diverting the objective of the game AWAY from winning.

That said, you’ll note I do use my power ratings (SRS) as a tie-breaker but only when coupled with a superior record and not if it would reverse head-to-head.


RE: CFB's Philosopher's Stone ("real" SOR rankings that fans can calculate o... - ken d - 01-06-2026 08:33 PM

(01-06-2026 04:26 PM)Hokie Mark Wrote:  
(01-06-2026 10:22 AM)ken d Wrote:  [quote='Crayton' pid='20506823' dateline='1767654580']
As a starting point to develop an algorithm that gives a clear picture throughout the season, you must first have an algorithm that produces a logical result at the point in the season when you have all the data you are going to get. That's after the CCGs.

This particular algorithm seems to not produce that. It has BYU ranked 7th ahead of Texas Tech at #11, despite the fact that Tech beat BYU twice convincingly, both late in the season. Tech would make the field because they were the third highest ranked conference champion. But if logic says BYU should be ranked below them, that is to say at #12, then Notre Dame would move up to #9 and replace BYU in the field of 12.

So, it seems you would need to tweak the algorithm in a way that resolves the BYU dilemma without creating a different dilemma. I think we have to consider the possibility that there is no algorithm that would do that with any consistency. If that is the case, then the flaw may be within us, the fans, for wanting a ranking system in the first place rather than some combination of conference standings and wild card games.

This speaks to my point. There's nothing illogical in saying "BYU played a stronger schedule than Texas Tech" and also "Texas Tech is better than BYU" -- those two things are different, like "higher" and "in front of" are different.

In my estimation, people throughout CFB (not just this thread) are spending WAY too much time and energy on calculating strength of schedule and/or record, which frankly, is not the really important question!

Power ratings aren't that hard to calculate - but they sometimes produce results that some people just don't like very much!

Much of the problem with all of this is that there is really no consensus about whether a national championship tournament should include the "best" teams, or the "most deserving" teams.

For most of of college football's history, it didn't matter much who claimed to be the best, or who someone else proclaimed who that was, or if more than one school or pundit made competing claims. Having controversy just added to the appeal of college football and didn't cost anybody anything.

Today, when millions of dollars are at stake, it matters. In, say, 1980, I don't believe many fans believed that AP voters or coaches were sufficiently biased as to overcome the balancing effect of large numbers of voters with different biases. But today? I suspect that many fans, maybe even a majority of fans, question the integrity of the people chosen to serve on the selection committee. This, despite the inescapable fact that their output closely matches the results of both human rankings and computer rankings nearly every year.

So we look to computer algorithms to take the subjectivity out of the selection process, even though many of us believe even those algorithms are rigged. Even though the NFL's process for qualifying for their playoffs doesn't even pretend to include either the "best" or "most deserving" teams, most fans don't question the fairness of that process. No SOS, SOR, or any other acronym for them. With 32 teams in 8 divisions is all they need, because all 32 teams benefit from their process.

I don't believe we will ever see college football evolve in a way that will match the NFL's formula. Until we do, this debate will continue on forums like this. But then, we wouldn't have it any other way.


RE: CFB's Philosopher's Stone ("real" SOR rankings that fans can calculate o... - Usajags - 01-06-2026 08:37 PM

This is what happens when you let nerds decide who is in the CFP.


RE: CFB's Philosopher's Stone ("real" SOR rankings that fans can calculate on pen+paper) - Crayton - 01-07-2026 06:05 AM

(01-06-2026 08:37 PM)Usajags Wrote:  This is what happens when you let nerds decide who is in the CFP.

The nerds in the newsroom have been picking champions or playoff fields for nearly 100 years. Business suits have now been doing it for a dozen years. This data geek wants to give control back to the jocks on the field.

Back in the day, the NFL awarded its championship to the team with the most Wins (they played differing amounts of games until the 30s). The data geeks convinced them to use something called “percentages” to determine their champion. After the Merger, data geeks introduced the “half win” for games that ended in a tie, as a better way of calculating win percentage than simply ignoring the tie.


RE: CFB's Philosopher's Stone ("real" SOR rankings that fans can calculate on pen+paper) - Dr Dawgma III - 01-07-2026 04:01 PM

Is this one of those formulas where a team is rewarded for losing to "good" teams?