(01-27-2021 06:25 AM)slhNavy91 Wrote:  

(01-26-2021 11:18 PM)quo vadis Wrote:  

(01-15-2021 08:48 PM)slhNavy91 Wrote:  Wolfe has not reported all year.

Wolfe sensibly realized that there were too few OOC data points to make a meaningful analysis. Kudos to him for that.

Eyeballing it, which is all we can do, the SEC was the best overall conference and the SBC was the best G5 conference. Fortunately for the AAC, the CFP contract does call for the use of those computers, and with the skewed numbers, the AAC will get the most money as the top conference as determined by the CFP.

At USF, we won't complain about getting that money.

No.
The AAC had the best non-contract-bowl conference team.
The SBC had the best top three teams of non-contract bowl conferences, but the drop off was so severe that even top four teams, the edge went back to AAC. Then the SBC's middle and bottom were significantly worse, bad enough to negate the advantage of an extreme outlier good year by the top of the conference (even though that bottom four was for them outlier good tood too).

Top to bottom, AAC better - "more better" than mwc advantage in 2018 when mwc was outlier good and the AAC was outlier bad.

That's what post #8 says, for the numerically literate.

The problem is, post #8 relies on the 2020 MC, which is unreliable because of the paucity of data points. So all we can do is eyeball it.

The Colley-Matrix is garbage as a ranking system, but one thing it is good for is showing all the OOC games. Tossing out the FCS stuff, the Sun Belt was 15-15 vs FBS opponents, the AAC was 9-12. The SBC was 3-1 vs the P5, the AAC was 1-4. That's as top to bottom as you can get, all of the games.

(01-27-2021 09:44 AM)quo vadis Wrote:  

(01-27-2021 06:25 AM)slhNavy91 Wrote:  

(01-26-2021 11:18 PM)quo vadis Wrote:  

(01-15-2021 08:48 PM)slhNavy91 Wrote:  Wolfe has not reported all year.

Wolfe sensibly realized that there were too few OOC data points to make a meaningful analysis. Kudos to him for that.

Eyeballing it, which is all we can do, the SEC was the best overall conference and the SBC was the best G5 conference. Fortunately for the AAC, the CFP contract does call for the use of those computers, and with the skewed numbers, the AAC will get the most money as the top conference as determined by the CFP.

At USF, we won't complain about getting that money.

No.
The AAC had the best non-contract-bowl conference team.
The SBC had the best top three teams of non-contract bowl conferences, but the drop off was so severe that even top four teams, the edge went back to AAC. Then the SBC's middle and bottom were significantly worse, bad enough to negate the advantage of an extreme outlier good year by the top of the conference (even though that bottom four was for them outlier good tood too).

Top to bottom, AAC better - "more better" than mwc advantage in 2018 when mwc was outlier good and the AAC was outlier bad.

That's what post #8 says, for the numerically literate.

The problem is, post #8 relies on the 2020 MC, which is unreliable because of the paucity of data points. So all we can do is eyeball it.

The Colley-Matrix is garbage as a ranking system, but one thing it is good for is showing all the OOC games. Tossing out the FCS stuff, the Sun Belt was 15-15 vs FBS opponents, the AAC was 9-12. The SBC was 3-1 vs the P5, the AAC was 1-4. That's as top to bottom as you can get, all of the games.

So your eye test told you that Coastal Carolina was better than Cincinnati even after the loss to Liberty (Coastal was the first FBS team with a winning record that the Flames beat)?
Or does your eye test tell you that Georgia Southern was better than Memphis? (that's what the part about dropoff from top three average ranking to top four average ranking is)
Or in the middle, does your eye test say that Georgia State, Troy and Arkansas St are better than SMU, Tulane, and Houston?
Or does your eye test say that 0-10 ULM is something other than the lowest ranked FBS team, dragging down the overall conference assessment?

I just want to get this quo eyetest down in writing, so we all know exactly why we're throwing away data-driven things (imperfect though they may be).

(01-27-2021 09:44 AM)quo vadis Wrote:  

(01-27-2021 06:25 AM)slhNavy91 Wrote:  

(01-26-2021 11:18 PM)quo vadis Wrote:  

(01-15-2021 08:48 PM)slhNavy91 Wrote:  Wolfe has not reported all year.

Wolfe sensibly realized that there were too few OOC data points to make a meaningful analysis. Kudos to him for that.

Eyeballing it, which is all we can do, the SEC was the best overall conference and the SBC was the best G5 conference. Fortunately for the AAC, the CFP contract does call for the use of those computers, and with the skewed numbers, the AAC will get the most money as the top conference as determined by the CFP.

At USF, we won't complain about getting that money.

No.
The AAC had the best non-contract-bowl conference team.
The SBC had the best top three teams of non-contract bowl conferences, but the drop off was so severe that even top four teams, the edge went back to AAC. Then the SBC's middle and bottom were significantly worse, bad enough to negate the advantage of an extreme outlier good year by the top of the conference (even though that bottom four was for them outlier good tood too).

Top to bottom, AAC better - "more better" than mwc advantage in 2018 when mwc was outlier good and the AAC was outlier bad.

That's what post #8 says, for the numerically literate.

The problem is, post #8 relies on the 2020 MC, which is unreliable because of the paucity of data points. So all we can do is eyeball it.

The Colley-Matrix is garbage as a ranking system, but one thing it is good for is showing all the OOC games. Tossing out the FCS stuff, the Sun Belt was 15-15 vs FBS opponents, the AAC was 9-12. The SBC was 3-1 vs the P5, the AAC was 1-4. That's as top to bottom as you can get, all of the games.

As far as the Colley-Matrix, I have actually chuckled to myself from time to time awaiting this impassioned quo vadis adoption of Colley.

See, the "bias-free" in the headline of Colley's Bias Free Matrix Rankings really means that there is an over-reliance on W-L record and little weighting or statistical dive into whom the wins/losses were against. Kind of like quo has been doing with shallow, W-L only assessment. I knew that quo's profession of love for Colley was coming!

Look deeper - even just a little deeper at opponents' W-L and at opponents' opponents' W-L (aka strength of schedule) and it isn't as easy or as crystal clear as quo would like. Margin of victory, SunBelt can look good....based almost wholly on whuppin' up on the CUSA in their bowls (not even the wins over CUSA in regular season). That's just the first level beyond simplistic W-L record -- we're not talking about advanced stats or analytics or anything.

We started down this path before, and reflecting on that earlier exchange had me waiting for quo's Damascene conversion to the Church of Colley Matrix.

(01-27-2021 06:25 AM)slhNavy91 Wrote:  No.
The AAC had the best non-contract-bowl conference team.
The SBC had the best top three teams of non-contract bowl conferences, but the drop off was so severe that even top four teams, the edge went back to AAC. Then the SBC's middle and bottom were significantly worse, bad enough to negate the advantage of an extreme outlier good year by the top of the conference (even though that bottom four was for them outlier good tood too).

Top to bottom, AAC better - "more better" than mwc advantage in 2018 when mwc was outlier good and the AAC was outlier bad.

That's what post #8 says, for the numerically literate.

Here's the Sun Belt-AAC matchups (records are conference records).
Memphis (5-3) beats Ark State (2-6) by 13
Tulane (3-5) beats South Alabama (3-5) by 3
SMU (4-3) beats Texas State (1-7) by 7
Georgia State (4-4) beats ECU (3-5) by 20

I'm not seeing any advantage the AAC having over the SBC in the middle nd bottom. In 2 of those games the AAC team had a significantly better conference record and won by margins of 7 and 13, and in the two more even games you have a 20-point Sun Belt win and a 7-point AAC win.

I do expect the AAC to be #1, though, because with the lack of non-conference games more deference is going to be given to years before 2020. I just hope the Sun Belt is at least ahead of the Mountain West.

My 89-year-old father watched a lot of college football this year (as he does each season). He closely follows the AAC (due to Cincy and Memphis) and the Sun Belt (because he has a soft spot in his heart for the G5). He said the American and the Sun Belt, top to bottom, were very close in terms of quality, competitiveness, etc., this past campaign.

I realize that's worth no more than my thinking that, hypothetically, actress Jessica Chastain would be better suited to handle Bill Dazzle's eccentricities and questionable language during a dinner date than actress Alyson Hannigan would be.

(01-27-2021 12:26 PM)EigenEagle Wrote:  

(01-27-2021 06:25 AM)slhNavy91 Wrote:  No.
The AAC had the best non-contract-bowl conference team.
The SBC had the best top three teams of non-contract bowl conferences, but the drop off was so severe that even top four teams, the edge went back to AAC. Then the SBC's middle and bottom were significantly worse, bad enough to negate the advantage of an extreme outlier good year by the top of the conference (even though that bottom four was for them outlier good tood too).

Top to bottom, AAC better - "more better" than mwc advantage in 2018 when mwc was outlier good and the AAC was outlier bad.

That's what post #8 says, for the numerically literate.

Here's the Sun Belt-AAC matchups (records are conference records).
Memphis (5-3) beats Ark State (2-6) by 13
Tulane (3-5) beats South Alabama (3-5) by 3
SMU (4-3) beats Texas State (1-7) by 7
Georgia State (4-4) beats ECU (3-5) by 20

I'm not seeing any advantage the AAC having over the SBC in the middle nd bottom. In 2 of those games the AAC team had a significantly better conference record and won by margins of 7 and 13, and in the two more even games you have a 20-point Sun Belt win and a 7-point AAC win.

I do expect the AAC to be #1, though, because with the lack of non-conference games more deference is going to be given to years before 2020. I just hope the Sun Belt is at least ahead of the Mountain West.

Compare 1 vs 1, 2 vs 2, etc down the line. Do it down to #10 vs #10, or do the stats trick of removing the AAC's #6 so that you get both 1 vs 1 and last vs last.
Massey Composite the AAC #x is a higher rank than the SBC #x 8 out of 10 or 7 out of 10 times.
Do it with Sagarin -- which is essentially a predictor not an ordinal ranker -- and the advantage goes to AAC 9 out of 10 or 8 out of 10 times.
I used Massey's matchup widget (in his own ranking system, not the composite) and got 10 out of 10 predicted wins for the AAC.

That's the top to bottom matchups.

The line you bolded, though, was about conferences' average ranking. I was saying that the Massey Composite difference between #6 AAC and #7 SunBelt is greater than the 2018 difference between #6 mwc and #7 AAC.

The 2020 CFP payout rankings, by the five former BCS systems that published this year, should be final -- see post #1 here. SBC #7 ahead of mwc.

Also, in years past poster MadMartigan -- I believe a Boise fan -- has been here in years past with years' worth of comparisons of top 4 average MC ranking, middle 4 average MC ranking, and bottom 4 average MC ranking. That's kind of what I was going for in post #8.
Top 1 advantage AAC
Top 3 advantage SBC
Top 4 advantage AAC
Middle 4 advantage AAC
Middle 3 vs Middle 2 (since neither of us has 12 members) advantage AAC
Bottom 4 advantage AAC

(01-27-2021 10:00 AM)slhNavy91 Wrote:  

(01-27-2021 09:44 AM)quo vadis Wrote:  

(01-27-2021 06:25 AM)slhNavy91 Wrote:  

(01-26-2021 11:18 PM)quo vadis Wrote:  

(01-15-2021 08:48 PM)slhNavy91 Wrote:  Wolfe has not reported all year.

Wolfe sensibly realized that there were too few OOC data points to make a meaningful analysis. Kudos to him for that.

Eyeballing it, which is all we can do, the SEC was the best overall conference and the SBC was the best G5 conference. Fortunately for the AAC, the CFP contract does call for the use of those computers, and with the skewed numbers, the AAC will get the most money as the top conference as determined by the CFP.

At USF, we won't complain about getting that money.

No.
The AAC had the best non-contract-bowl conference team.
The SBC had the best top three teams of non-contract bowl conferences, but the drop off was so severe that even top four teams, the edge went back to AAC. Then the SBC's middle and bottom were significantly worse, bad enough to negate the advantage of an extreme outlier good year by the top of the conference (even though that bottom four was for them outlier good tood too).

Top to bottom, AAC better - "more better" than mwc advantage in 2018 when mwc was outlier good and the AAC was outlier bad.

That's what post #8 says, for the numerically literate.

The problem is, post #8 relies on the 2020 MC, which is unreliable because of the paucity of data points. So all we can do is eyeball it.

The Colley-Matrix is garbage as a ranking system, but one thing it is good for is showing all the OOC games. Tossing out the FCS stuff, the Sun Belt was 15-15 vs FBS opponents, the AAC was 9-12. The SBC was 3-1 vs the P5, the AAC was 1-4. That's as top to bottom as you can get, all of the games.

So your eye test told you that Coastal Carolina was better than Cincinnati even after the loss to Liberty (Coastal was the first FBS team with a winning record that the Flames beat)?
Or does your eye test tell you that Georgia Southern was better than Memphis? (that's what the part about dropoff from top three average ranking to top four average ranking is)
Or in the middle, does your eye test say that Georgia State, Troy and Arkansas St are better than SMU, Tulane, and Houston?
Or does your eye test say that 0-10 ULM is something other than the lowest ranked FBS team, dragging down the overall conference assessment?

I just want to get this quo eyetest down in writing, so we all know exactly why we're throwing away data-driven things (imperfect though they may be).

LOL ... by their nature "eye tests" are purely subjective. In this case, my eye was on the conferences as a whole, not on specific matchups. FWIW, I thought Cincy was better than Coastal, but would have liked to have seen it settled on the field.

As for Colley, I don't endorse anything about his method or his rankings. They are IMO ridiculous. What I said was the on his web page, he provides a nice table that summarizes all OOC games played by all conferences. He gives you the record of each conference vs every other conference, vs independents, vs FCS, and the overall records. Pus, the scores of every OOC game played by every conference. Which makes it easy to see how each conference did vs all the other conferences.

Such as being able to see that the SBC was 15-15 vs FBS while the AAC was 9-12 vs FBS, etc.

Let's face it: The AAC was fortunate that the computers were borked by low data points, because the SBC was likely the best G5 conference. That's also a big slap-down to the notion that the AAC has achieved escape-velocity in terms of creating separation from the rest of the G5.

Being the second-place G5 conference two of the past three years is a bad look in terms of claiming clear tweener status.

That's just how it is.

(01-27-2021 01:56 PM)quo vadis Wrote:  

(01-27-2021 10:00 AM)slhNavy91 Wrote:  

(01-27-2021 09:44 AM)quo vadis Wrote:  

(01-27-2021 06:25 AM)slhNavy91 Wrote:  

(01-26-2021 11:18 PM)quo vadis Wrote:  Wolfe sensibly realized that there were too few OOC data points to make a meaningful analysis. Kudos to him for that.

Eyeballing it, which is all we can do, the SEC was the best overall conference and the SBC was the best G5 conference. Fortunately for the AAC, the CFP contract does call for the use of those computers, and with the skewed numbers, the AAC will get the most money as the top conference as determined by the CFP.

At USF, we won't complain about getting that money.

No.
The AAC had the best non-contract-bowl conference team.
The SBC had the best top three teams of non-contract bowl conferences, but the drop off was so severe that even top four teams, the edge went back to AAC. Then the SBC's middle and bottom were significantly worse, bad enough to negate the advantage of an extreme outlier good year by the top of the conference (even though that bottom four was for them outlier good tood too).

Top to bottom, AAC better - "more better" than mwc advantage in 2018 when mwc was outlier good and the AAC was outlier bad.

That's what post #8 says, for the numerically literate.

The problem is, post #8 relies on the 2020 MC, which is unreliable because of the paucity of data points. So all we can do is eyeball it.

The Colley-Matrix is garbage as a ranking system, but one thing it is good for is showing all the OOC games. Tossing out the FCS stuff, the Sun Belt was 15-15 vs FBS opponents, the AAC was 9-12. The SBC was 3-1 vs the P5, the AAC was 1-4. That's as top to bottom as you can get, all of the games.

So your eye test told you that Coastal Carolina was better than Cincinnati even after the loss to Liberty (Coastal was the first FBS team with a winning record that the Flames beat)?
Or does your eye test tell you that Georgia Southern was better than Memphis? (that's what the part about dropoff from top three average ranking to top four average ranking is)
Or in the middle, does your eye test say that Georgia State, Troy and Arkansas St are better than SMU, Tulane, and Houston?
Or does your eye test say that 0-10 ULM is something other than the lowest ranked FBS team, dragging down the overall conference assessment?

I just want to get this quo eyetest down in writing, so we all know exactly why we're throwing away data-driven things (imperfect though they may be).

LOL ... by their nature "eye tests" are purely subjective. In this case, my eye was on the conferences as a whole, not on specific matchups. FWIW, I thought Cincy was better than Coastal, but would have liked to have seen it settled on the field.

As for Colley, I don't endorse anything about his method or his rankings. They are IMO ridiculous. What I said was the on his web page, he provides a nice table that summarizes all OOC games played by all conferences. He gives you the record of each conference vs every other conference, vs independents, vs FCS, and the overall records. Pus, the scores of every OOC game played by every conference. Which makes it easy to see how each conference did vs all the other conferences.

Such as being able to see that the SBC was 15-15 vs FBS while the AAC was 9-12 vs FBS, etc.

Let's face it: The AAC was fortunate that the computers were borked by low data points, because the SBC was likely the best G5 conference. That's also a big slap-down to the notion that the AAC has achieved escape-velocity in terms of creating separation from the rest of the G5.

Being the second-place G5 conference two of the past three years is a bad look in terms of claiming clear tweener status.

That's just how it is.

Second place non-contract-bowl conference once out of the last three years, 2018. That's also second place once out of the last five years.

According to the CFP payout, 2020 will have AAC first. According to the Massey Composite, 2020 will have AAC first.

To get to TWO second places of the non-contract-bowl conferences you need the entire CFP era.
Five out of seven years in first
Five out of seven years earned the NY6 bid
Four out of seven years closer to #5 than to #8-10
Three out of seven years closer to #5 than to #7-10
Escape velocity, Achieved. P6/Tweener, Claimable.

To say "second place two of the past three years," you really need to make sure you specify "In quo's imaginary world." Or something along those lines.
And just like the actual second place once out of the last five years needed an outlier good year from the mwc and an outlier bad year from the AAC, the quo's-imaginary-world second place required an extreme outlier good year from the SunBelt.

.

What is this "smilie" icon supposed to connote or signify, anyway: (?)


a) "A hot toddy and the National Enquirer - - my favorite combination!"

b) "I can't believe I'm reading the comics - has someone spiked my cocoa?

c) "Reading comprehension is shot - - I hope this espresso wakes me up."

d) a and b

e) none of the above

f) all of the above

(01-27-2021 02:27 PM)slhNavy91 Wrote:  

(01-27-2021 01:56 PM)quo vadis Wrote:  

(01-27-2021 10:00 AM)slhNavy91 Wrote:  

(01-27-2021 09:44 AM)quo vadis Wrote:  

(01-27-2021 06:25 AM)slhNavy91 Wrote:  No.
The AAC had the best non-contract-bowl conference team.
The SBC had the best top three teams of non-contract bowl conferences, but the drop off was so severe that even top four teams, the edge went back to AAC. Then the SBC's middle and bottom were significantly worse, bad enough to negate the advantage of an extreme outlier good year by the top of the conference (even though that bottom four was for them outlier good tood too).

Top to bottom, AAC better - "more better" than mwc advantage in 2018 when mwc was outlier good and the AAC was outlier bad.

That's what post #8 says, for the numerically literate.

The problem is, post #8 relies on the 2020 MC, which is unreliable because of the paucity of data points. So all we can do is eyeball it.

The Colley-Matrix is garbage as a ranking system, but one thing it is good for is showing all the OOC games. Tossing out the FCS stuff, the Sun Belt was 15-15 vs FBS opponents, the AAC was 9-12. The SBC was 3-1 vs the P5, the AAC was 1-4. That's as top to bottom as you can get, all of the games.

So your eye test told you that Coastal Carolina was better than Cincinnati even after the loss to Liberty (Coastal was the first FBS team with a winning record that the Flames beat)?
Or does your eye test tell you that Georgia Southern was better than Memphis? (that's what the part about dropoff from top three average ranking to top four average ranking is)
Or in the middle, does your eye test say that Georgia State, Troy and Arkansas St are better than SMU, Tulane, and Houston?
Or does your eye test say that 0-10 ULM is something other than the lowest ranked FBS team, dragging down the overall conference assessment?

I just want to get this quo eyetest down in writing, so we all know exactly why we're throwing away data-driven things (imperfect though they may be).

LOL ... by their nature "eye tests" are purely subjective. In this case, my eye was on the conferences as a whole, not on specific matchups. FWIW, I thought Cincy was better than Coastal, but would have liked to have seen it settled on the field.

As for Colley, I don't endorse anything about his method or his rankings. They are IMO ridiculous. What I said was the on his web page, he provides a nice table that summarizes all OOC games played by all conferences. He gives you the record of each conference vs every other conference, vs independents, vs FCS, and the overall records. Pus, the scores of every OOC game played by every conference. Which makes it easy to see how each conference did vs all the other conferences.

Such as being able to see that the SBC was 15-15 vs FBS while the AAC was 9-12 vs FBS, etc.

Let's face it: The AAC was fortunate that the computers were borked by low data points, because the SBC was likely the best G5 conference. That's also a big slap-down to the notion that the AAC has achieved escape-velocity in terms of creating separation from the rest of the G5.

Being the second-place G5 conference two of the past three years is a bad look in terms of claiming clear tweener status.

That's just how it is.

Second place non-contract-bowl conference once out of the last three years, 2018. That's also second place once out of the last five years.

According to the CFP payout, 2020 will have AAC first. According to the Massey Composite, 2020 will have AAC first.

(snip)

To say "second place two of the past three years," you really need to make sure you specify "In quo's imaginary world." Or something along those lines.
And just like the actual second place once out of the last five years needed an outlier good year from the mwc and an outlier bad year from the AAC, the quo's-imaginary-world second place required an extreme outlier good year from the SunBelt.

I've already agreed that, thanks, to computers being borked this year, the AAC will be the official CFP First-Place G5 conference.

But they weren't actually the best G5 conference, the SBC was.

That's just how it actually is, though I'm glad that, thanks to the CFP having no mechanism to get around borked computers in a special year like this, USF will collect that first-place money. What should have happened was that the CFP declares the computers unreliable for this year and has the committee itself evaluation the conferences. That would have likely resulted in the SBC being named #1, because, well, they had the #1 record among the G5 conferences. But that process doesn't exist.

To perhaps the largest point, though, the point about the AAC being a clear-cut separate from the other G5: It doesn't matter if the AAC finishes second once every three years to the same specific G5, or whether they finish second once every three years to a random different G5 just having an exceptional supernova year. It adds up to the same thing - no separation. Because the issue isn't separation between the AAC and each individual other G5 in a one-on-one sense, but the AAC vs the G5 collective.

(01-27-2021 09:59 PM)quo vadis Wrote:  

(01-27-2021 02:27 PM)slhNavy91 Wrote:  

(01-27-2021 01:56 PM)quo vadis Wrote:  

(01-27-2021 10:00 AM)slhNavy91 Wrote:  

(01-27-2021 09:44 AM)quo vadis Wrote:  The problem is, post #8 relies on the 2020 MC, which is unreliable because of the paucity of data points. So all we can do is eyeball it.

The Colley-Matrix is garbage as a ranking system, but one thing it is good for is showing all the OOC games. Tossing out the FCS stuff, the Sun Belt was 15-15 vs FBS opponents, the AAC was 9-12. The SBC was 3-1 vs the P5, the AAC was 1-4. That's as top to bottom as you can get, all of the games.

So your eye test told you that Coastal Carolina was better than Cincinnati even after the loss to Liberty (Coastal was the first FBS team with a winning record that the Flames beat)?
Or does your eye test tell you that Georgia Southern was better than Memphis? (that's what the part about dropoff from top three average ranking to top four average ranking is)
Or in the middle, does your eye test say that Georgia State, Troy and Arkansas St are better than SMU, Tulane, and Houston?
Or does your eye test say that 0-10 ULM is something other than the lowest ranked FBS team, dragging down the overall conference assessment?

I just want to get this quo eyetest down in writing, so we all know exactly why we're throwing away data-driven things (imperfect though they may be).

LOL ... by their nature "eye tests" are purely subjective. In this case, my eye was on the conferences as a whole, not on specific matchups. FWIW, I thought Cincy was better than Coastal, but would have liked to have seen it settled on the field.

As for Colley, I don't endorse anything about his method or his rankings. They are IMO ridiculous. What I said was the on his web page, he provides a nice table that summarizes all OOC games played by all conferences. He gives you the record of each conference vs every other conference, vs independents, vs FCS, and the overall records. Pus, the scores of every OOC game played by every conference. Which makes it easy to see how each conference did vs all the other conferences.

Such as being able to see that the SBC was 15-15 vs FBS while the AAC was 9-12 vs FBS, etc.

Let's face it: The AAC was fortunate that the computers were borked by low data points, because the SBC was likely the best G5 conference. That's also a big slap-down to the notion that the AAC has achieved escape-velocity in terms of creating separation from the rest of the G5.

Being the second-place G5 conference two of the past three years is a bad look in terms of claiming clear tweener status.

That's just how it is.

Second place non-contract-bowl conference once out of the last three years, 2018. That's also second place once out of the last five years.

According to the CFP payout, 2020 will have AAC first. According to the Massey Composite, 2020 will have AAC first.

(snip)

To say "second place two of the past three years," you really need to make sure you specify "In quo's imaginary world." Or something along those lines.
And just like the actual second place once out of the last five years needed an outlier good year from the mwc and an outlier bad year from the AAC, the quo's-imaginary-world second place required an extreme outlier good year from the SunBelt.

I've already agreed that, thanks, to computers being borked this year, the AAC will be the official CFP First-Place G5 conference.

But they weren't actually the best G5 conference, the SBC was.

That's just how it actually is, though I'm glad that, thanks to the CFP having no mechanism to get around borked computers in a special year like this, USF will collect that first-place money. What should have happened was that the CFP declares the computers unreliable for this year and has the committee itself evaluation the conferences. That would have likely resulted in the SBC being named #1, because, well, they had the #1 record among the G5 conferences. But that process doesn't exist.

To perhaps the largest point, though, the point about the AAC being a clear-cut separate from the other G5: It doesn't matter if the AAC finishes second once every three years to the same specific G5, or whether they finish second once every three years to a random different G5 just having an exceptional supernova year. It adds up to the same thing - no separation. Because the issue isn't separation between the AAC and each individual other G5 in a one-on-one sense, but the AAC vs the G5 collective.

Ahh yes the always air tight Quo this is how I feel argument.

His final weak troll attempt when beat from pillar to post with facts. It's sad.

I assume Quo, you hope for the sort of similar emotional reaction from posters. I think that you'll disappointed in the SLH response. You should try Jed, he seems to struggle with that sort of thing. I think we would all find it vastly more entertaining than the vicious intellectual beating you are currently receiving from an opponent who not only out matches and out classes you, but comes with receipts.

(01-27-2021 02:27 PM)slhNavy91 Wrote:  

(01-27-2021 01:56 PM)quo vadis Wrote:  

(01-27-2021 10:00 AM)slhNavy91 Wrote:  

(01-27-2021 09:44 AM)quo vadis Wrote:  

(01-27-2021 06:25 AM)slhNavy91 Wrote:  No.
The AAC had the best non-contract-bowl conference team.
The SBC had the best top three teams of non-contract bowl conferences, but the drop off was so severe that even top four teams, the edge went back to AAC. Then the SBC's middle and bottom were significantly worse, bad enough to negate the advantage of an extreme outlier good year by the top of the conference (even though that bottom four was for them outlier good tood too).

Top to bottom, AAC better - "more better" than mwc advantage in 2018 when mwc was outlier good and the AAC was outlier bad.

That's what post #8 says, for the numerically literate.

The problem is, post #8 relies on the 2020 MC, which is unreliable because of the paucity of data points. So all we can do is eyeball it.

The Colley-Matrix is garbage as a ranking system, but one thing it is good for is showing all the OOC games. Tossing out the FCS stuff, the Sun Belt was 15-15 vs FBS opponents, the AAC was 9-12. The SBC was 3-1 vs the P5, the AAC was 1-4. That's as top to bottom as you can get, all of the games.

So your eye test told you that Coastal Carolina was better than Cincinnati even after the loss to Liberty (Coastal was the first FBS team with a winning record that the Flames beat)?
Or does your eye test tell you that Georgia Southern was better than Memphis? (that's what the part about dropoff from top three average ranking to top four average ranking is)
Or in the middle, does your eye test say that Georgia State, Troy and Arkansas St are better than SMU, Tulane, and Houston?
Or does your eye test say that 0-10 ULM is something other than the lowest ranked FBS team, dragging down the overall conference assessment?

I just want to get this quo eyetest down in writing, so we all know exactly why we're throwing away data-driven things (imperfect though they may be).

LOL ... by their nature "eye tests" are purely subjective. In this case, my eye was on the conferences as a whole, not on specific matchups. FWIW, I thought Cincy was better than Coastal, but would have liked to have seen it settled on the field.

As for Colley, I don't endorse anything about his method or his rankings. They are IMO ridiculous. What I said was the on his web page, he provides a nice table that summarizes all OOC games played by all conferences. He gives you the record of each conference vs every other conference, vs independents, vs FCS, and the overall records. Pus, the scores of every OOC game played by every conference. Which makes it easy to see how each conference did vs all the other conferences.

Such as being able to see that the SBC was 15-15 vs FBS while the AAC was 9-12 vs FBS, etc.

Let's face it: The AAC was fortunate that the computers were borked by low data points, because the SBC was likely the best G5 conference. That's also a big slap-down to the notion that the AAC has achieved escape-velocity in terms of creating separation from the rest of the G5.

Being the second-place G5 conference two of the past three years is a bad look in terms of claiming clear tweener status.

That's just how it is.

Second place non-contract-bowl conference once out of the last three years, 2018. That's also second place once out of the last five years.

According to the CFP payout, 2020 will have AAC first. According to the Massey Composite, 2020 will have AAC first.

To get to TWO second places of the non-contract-bowl conferences you need the entire CFP era.
Five out of seven years in first
Five out of seven years earned the NY6 bid
Four out of seven years closer to #5 than to #8-10
Three out of seven years closer to #5 than to #7-10
Escape velocity, Achieved. P6/Tweener, Claimable.

To say "second place two of the past three years," you really need to make sure you specify "In quo's imaginary world." Or something along those lines.
And just like the actual second place once out of the last five years needed an outlier good year from the mwc and an outlier bad year from the AAC, the quo's-imaginary-world second place required an extreme outlier good year from the SunBelt.

If the question is, "Which non-P5 conference had the highest FB rankings over the past three years?", here's the answer based on the 2018-2020 Massey Composite rankings:

...........................2018..........2019..........2020*......Mean Rank:
PAC-12...................#5............#4..............#5.............#4.67
ACC.......................#4.............#6.............#6..............#5.33
FB Independents..#6.............#9.............#4*.............#6.33*
AAC.......................#8.............#5.............#7..............#6.67
MWC......................#7.............#7............#10.............#8.00
SBC.......................#9.............#8.............#8..............#8.33
MAC.......................#10...........#11.............#9..............#10.00
CUSA.....................#11..........#10.............#11.............#10.67

*NOTE: 2020 rankings for FBS independents did not include data for Connecticut and New Mexico, because they did not play. The combined record of BYU (#18), Army (#72), Liberty (#75), & UMass (#126) was 30-8.

NOTE: In 2016, the rankings were: #6 AAC, #7 FBS Independents, #8 MWC, #9 SBC, #10 MAC, #11 CUSA. In 2017, the rankings were: #6 FBS Independents, #7 AAC, #8 MWC, #9 MAC, #10 CUSA, #11 SBC.

Here are the conference rankings based on the RPI:

......................2018............2019............2020............Mean Rank
ACC..................#5................#6...............#4................#5.00
PAC-12.............#4................#5...............#8................#5.67
AAC...................#6...............#4...............#9................#6.33
Independents....#7................#9...............#6.................#7.33
SBC...................#11..............#8..............#5...................#8.00
MWC.................#8................#7..............#10..................#8.33
MAC.................#10...............#11.............#7...................#9.33
CUSA................#9................#10............#11..................#10.00

Combining the RPI and Massey Composite rankings:

...........................Mean of RPI & Massey Composite Ranks (2018-2020):
PAC-12....................................#4.835
ACC........................................#5.165
AAC........................................#6.500
FB Independents......................#6.830
SBC........................................#8.165
MWC......................................#8.165
MAC.......................................#9.665
CUSA.....................................#10.33


If the FB Independents are excluded, it looks like this:

...........................Mean of RPI & Massey Composite Ranks (2018-2020):

PAC-12....................................#4.835

ACC........................................#5.165

AAC........................................#6.500

SBC........................................#8.165

MWC......................................#8.165

MAC.......................................#9.665

CUSA.....................................#10.33

.................................................................................................................

The mean rank of the AAC (6.50) is more similar to the mean rank of the lowest P5 conference - - the ACC (5.165) - - than it is to the mean rank of the next highest G5 conferences (SBC and MWC (8.165).

Difference between mean ranks of the AAC and ACC: 1.335 ranks.

Difference between mean ranks of the AAC and SBC/MWC: 1.665 ranks.

.

Q: Is the mean rank of the AAC approximately half-way between the mean ranks of the lowest-ranked P5 conference (the ACC) and the highest-ranked non-AAC, non-P5 conferences (the SBC and MWC)?

A: Yes.

.

Q: Is the mean rank of the FBS Independents, as a group (not including Notre Dame), also approximately half-way between the mean ranks of the lowest-ranked P5 conference (the ACC) and the highest-ranked non-AAC, non-P5 conferences (the SBC and MWC)?

A: Yes.

(01-27-2021 10:48 PM)Foreverandever Wrote:  

(01-27-2021 09:59 PM)quo vadis Wrote:  

(01-27-2021 02:27 PM)slhNavy91 Wrote:  

(01-27-2021 01:56 PM)quo vadis Wrote:  

(01-27-2021 10:00 AM)slhNavy91 Wrote:  So your eye test told you that Coastal Carolina was better than Cincinnati even after the loss to Liberty (Coastal was the first FBS team with a winning record that the Flames beat)?
Or does your eye test tell you that Georgia Southern was better than Memphis? (that's what the part about dropoff from top three average ranking to top four average ranking is)
Or in the middle, does your eye test say that Georgia State, Troy and Arkansas St are better than SMU, Tulane, and Houston?
Or does your eye test say that 0-10 ULM is something other than the lowest ranked FBS team, dragging down the overall conference assessment?

I just want to get this quo eyetest down in writing, so we all know exactly why we're throwing away data-driven things (imperfect though they may be).

LOL ... by their nature "eye tests" are purely subjective. In this case, my eye was on the conferences as a whole, not on specific matchups. FWIW, I thought Cincy was better than Coastal, but would have liked to have seen it settled on the field.

As for Colley, I don't endorse anything about his method or his rankings. They are IMO ridiculous. What I said was the on his web page, he provides a nice table that summarizes all OOC games played by all conferences. He gives you the record of each conference vs every other conference, vs independents, vs FCS, and the overall records. Pus, the scores of every OOC game played by every conference. Which makes it easy to see how each conference did vs all the other conferences.

Such as being able to see that the SBC was 15-15 vs FBS while the AAC was 9-12 vs FBS, etc.

Let's face it: The AAC was fortunate that the computers were borked by low data points, because the SBC was likely the best G5 conference. That's also a big slap-down to the notion that the AAC has achieved escape-velocity in terms of creating separation from the rest of the G5.

Being the second-place G5 conference two of the past three years is a bad look in terms of claiming clear tweener status.

That's just how it is.

Second place non-contract-bowl conference once out of the last three years, 2018. That's also second place once out of the last five years.

According to the CFP payout, 2020 will have AAC first. According to the Massey Composite, 2020 will have AAC first.

(snip)

To say "second place two of the past three years," you really need to make sure you specify "In quo's imaginary world." Or something along those lines.
And just like the actual second place once out of the last five years needed an outlier good year from the mwc and an outlier bad year from the AAC, the quo's-imaginary-world second place required an extreme outlier good year from the SunBelt.

I've already agreed that, thanks, to computers being borked this year, the AAC will be the official CFP First-Place G5 conference.

But they weren't actually the best G5 conference, the SBC was.

That's just how it actually is, though I'm glad that, thanks to the CFP having no mechanism to get around borked computers in a special year like this, USF will collect that first-place money. What should have happened was that the CFP declares the computers unreliable for this year and has the committee itself evaluation the conferences. That would have likely resulted in the SBC being named #1, because, well, they had the #1 record among the G5 conferences. But that process doesn't exist.

To perhaps the largest point, though, the point about the AAC being a clear-cut separate from the other G5: It doesn't matter if the AAC finishes second once every three years to the same specific G5, or whether they finish second once every three years to a random different G5 just having an exceptional supernova year. It adds up to the same thing - no separation. Because the issue isn't separation between the AAC and each individual other G5 in a one-on-one sense, but the AAC vs the G5 collective.

Ahh yes the always air tight Quo this is how I feel argument.

His final weak troll attempt when beat from pillar to post with facts. It's sad.

(01-28-2021 02:13 AM)jedclampett Wrote:  

(01-27-2021 02:27 PM)slhNavy91 Wrote:  

(01-27-2021 01:56 PM)quo vadis Wrote:  

(01-27-2021 10:00 AM)slhNavy91 Wrote:  

(01-27-2021 09:44 AM)quo vadis Wrote:  The problem is, post #8 relies on the 2020 MC, which is unreliable because of the paucity of data points. So all we can do is eyeball it.

The Colley-Matrix is garbage as a ranking system, but one thing it is good for is showing all the OOC games. Tossing out the FCS stuff, the Sun Belt was 15-15 vs FBS opponents, the AAC was 9-12. The SBC was 3-1 vs the P5, the AAC was 1-4. That's as top to bottom as you can get, all of the games.

So your eye test told you that Coastal Carolina was better than Cincinnati even after the loss to Liberty (Coastal was the first FBS team with a winning record that the Flames beat)?
Or does your eye test tell you that Georgia Southern was better than Memphis? (that's what the part about dropoff from top three average ranking to top four average ranking is)
Or in the middle, does your eye test say that Georgia State, Troy and Arkansas St are better than SMU, Tulane, and Houston?
Or does your eye test say that 0-10 ULM is something other than the lowest ranked FBS team, dragging down the overall conference assessment?

I just want to get this quo eyetest down in writing, so we all know exactly why we're throwing away data-driven things (imperfect though they may be).

LOL ... by their nature "eye tests" are purely subjective. In this case, my eye was on the conferences as a whole, not on specific matchups. FWIW, I thought Cincy was better than Coastal, but would have liked to have seen it settled on the field.

As for Colley, I don't endorse anything about his method or his rankings. They are IMO ridiculous. What I said was the on his web page, he provides a nice table that summarizes all OOC games played by all conferences. He gives you the record of each conference vs every other conference, vs independents, vs FCS, and the overall records. Pus, the scores of every OOC game played by every conference. Which makes it easy to see how each conference did vs all the other conferences.

Such as being able to see that the SBC was 15-15 vs FBS while the AAC was 9-12 vs FBS, etc.

Let's face it: The AAC was fortunate that the computers were borked by low data points, because the SBC was likely the best G5 conference. That's also a big slap-down to the notion that the AAC has achieved escape-velocity in terms of creating separation from the rest of the G5.

Being the second-place G5 conference two of the past three years is a bad look in terms of claiming clear tweener status.

That's just how it is.

Second place non-contract-bowl conference once out of the last three years, 2018. That's also second place once out of the last five years.

According to the CFP payout, 2020 will have AAC first. According to the Massey Composite, 2020 will have AAC first.

To get to TWO second places of the non-contract-bowl conferences you need the entire CFP era.
Five out of seven years in first
Five out of seven years earned the NY6 bid
Four out of seven years closer to #5 than to #8-10
Three out of seven years closer to #5 than to #7-10
Escape velocity, Achieved. P6/Tweener, Claimable.

To say "second place two of the past three years," you really need to make sure you specify "In quo's imaginary world." Or something along those lines.
And just like the actual second place once out of the last five years needed an outlier good year from the mwc and an outlier bad year from the AAC, the quo's-imaginary-world second place required an extreme outlier good year from the SunBelt.

If the question is, "Which non-P5 conference had the highest FB rankings over the past three years?", here's the answer based on the 2018-2020 Massey Composite rankings:

...........................2018..........2019..........2020*......Mean Rank:
PAC-12...................#5............#4..............#5.............#4.67
ACC.......................#4.............#6.............#6..............#5.33
FB Independents..#6.............#9.............#4*.............#6.33*
AAC.......................#8.............#5.............#7..............#6.67
MWC......................#7.............#7............#10.............#8.00
SBC.......................#9.............#8.............#8..............#8.33
MAC.......................#10...........#11.............#9..............#10.00
CUSA.....................#11..........#10.............#11.............#10.67

*NOTE: 2020 rankings for FBS independents did not include data for Connecticut and New Mexico, because they did not play. The combined record of BYU (#18), Army (#72), Liberty (#75), & UMass (#126) was 30-8.

NOTE: In 2016, the rankings were: #6 AAC, #7 FBS Independents, #8 MWC, #9 SBC, #10 MAC, #11 CUSA. In 2017, the rankings were: #6 FBS Independents, #7 AAC, #8 MWC, #9 MAC, #10 CUSA, #11 SBC.

Pursuant to this and your other captures...

JUST the ordinal ranking doesn't capture the separation -- it's much more evident in the amplitude.
Looking at Massey Composite for the two datasets below.

AAC relative to the best G4
2020.....2.94 better - CORRECTION 2.99 better
2019.....14.29 better
2018.....2.48 lower
2017.....17.12 better
2016.....9.78 better
2015.....14.3 better
2014.....7.2 lower

As the #6, five years out of seven, AAC averages 11.69 better
When not #6, only twice in the CFP, AAC averages 4.84 behind

AAC relative to the lowest contract-bowl conference
2020.....9.58 lower - CORRECTION 9.45 lower
2019.....0.12 better
2018.....24.61 lower
2017.....14.98 lower
2016.....16.66 lower
2015.....14.2 lower
2014......48.73 lower - CORRECTION 35.2 lower
On this one, I'll highlight that indeed 2014 and 2018 are the outliers. 2020 (caveat/disclaimer assumed for strange year) in fact was not a down year looking through this lens.
Corrections hat tip to quo vadis in another post derived from this one

(01-28-2021 09:27 AM)slhNavy91 Wrote:  

(01-28-2021 02:13 AM)jedclampett Wrote:  

(01-27-2021 02:27 PM)slhNavy91 Wrote:  

(01-27-2021 01:56 PM)quo vadis Wrote:  

(01-27-2021 10:00 AM)slhNavy91 Wrote:  So your eye test told you that Coastal Carolina was better than Cincinnati even after the loss to Liberty (Coastal was the first FBS team with a winning record that the Flames beat)?
Or does your eye test tell you that Georgia Southern was better than Memphis? (that's what the part about dropoff from top three average ranking to top four average ranking is)
Or in the middle, does your eye test say that Georgia State, Troy and Arkansas St are better than SMU, Tulane, and Houston?
Or does your eye test say that 0-10 ULM is something other than the lowest ranked FBS team, dragging down the overall conference assessment?

I just want to get this quo eyetest down in writing, so we all know exactly why we're throwing away data-driven things (imperfect though they may be).

LOL ... by their nature "eye tests" are purely subjective. In this case, my eye was on the conferences as a whole, not on specific matchups. FWIW, I thought Cincy was better than Coastal, but would have liked to have seen it settled on the field.

As for Colley, I don't endorse anything about his method or his rankings. They are IMO ridiculous. What I said was the on his web page, he provides a nice table that summarizes all OOC games played by all conferences. He gives you the record of each conference vs every other conference, vs independents, vs FCS, and the overall records. Pus, the scores of every OOC game played by every conference. Which makes it easy to see how each conference did vs all the other conferences.

Such as being able to see that the SBC was 15-15 vs FBS while the AAC was 9-12 vs FBS, etc.

Let's face it: The AAC was fortunate that the computers were borked by low data points, because the SBC was likely the best G5 conference. That's also a big slap-down to the notion that the AAC has achieved escape-velocity in terms of creating separation from the rest of the G5.

Being the second-place G5 conference two of the past three years is a bad look in terms of claiming clear tweener status.

That's just how it is.

Second place non-contract-bowl conference once out of the last three years, 2018. That's also second place once out of the last five years.

According to the CFP payout, 2020 will have AAC first. According to the Massey Composite, 2020 will have AAC first.

To get to TWO second places of the non-contract-bowl conferences you need the entire CFP era.
Five out of seven years in first
Five out of seven years earned the NY6 bid
Four out of seven years closer to #5 than to #8-10
Three out of seven years closer to #5 than to #7-10
Escape velocity, Achieved. P6/Tweener, Claimable.

To say "second place two of the past three years," you really need to make sure you specify "In quo's imaginary world." Or something along those lines.
And just like the actual second place once out of the last five years needed an outlier good year from the mwc and an outlier bad year from the AAC, the quo's-imaginary-world second place required an extreme outlier good year from the SunBelt.

If the question is, "Which non-P5 conference had the highest FB rankings over the past three years?", here's the answer based on the 2018-2020 Massey Composite rankings:

...........................2018..........2019..........2020*......Mean Rank:
PAC-12...................#5............#4..............#5.............#4.67
ACC.......................#4.............#6.............#6..............#5.33
FB Independents..#6.............#9.............#4*.............#6.33*
AAC.......................#8.............#5.............#7..............#6.67
MWC......................#7.............#7............#10.............#8.00
SBC.......................#9.............#8.............#8..............#8.33
MAC.......................#10...........#11.............#9..............#10.00
CUSA.....................#11..........#10.............#11.............#10.67

*NOTE: 2020 rankings for FBS independents did not include data for Connecticut and New Mexico, because they did not play. The combined record of BYU (#18), Army (#72), Liberty (#75), & UMass (#126) was 30-8.

NOTE: In 2016, the rankings were: #6 AAC, #7 FBS Independents, #8 MWC, #9 SBC, #10 MAC, #11 CUSA. In 2017, the rankings were: #6 FBS Independents, #7 AAC, #8 MWC, #9 MAC, #10 CUSA, #11 SBC.

Pursuant to this and your other captures...

JUST the ordinal ranking doesn't capture the separation -- it's much more evident in the amplitude.
Looking at Massey Composite for the two datasets below.

AAC relative to the best G4
2020.....2.94 better
2019.....14.29 better
2018.....2.48 lower
2017.....17.12 better
2016.....9.78 better
2015.....14.3 better
2014.....7.2 lower

As the #6, five years out of seven, AAC averages 11.69 better
When not #6, only twice in the CFP, AAC averages 4.84 behind

AAC relative to the lowest contract-bowl conference
2020.....9.58 lower
2019.....0.12 better
2018.....24.61 lower
2017.....14.98 lower
2016.....16.66 lower
2015.....14.2 lower
2014......48.73 lower
On this one, I'll highlight that indeed 2014 and 2018 are the outliers. 2020 (caveat/disclaimer assumed for strange year) in fact was not a down year looking through this lens.

True, though the ordinal and cardinal data tell essentially the same story when summarized over multiple years.

(01-28-2021 02:44 AM)jedclampett Wrote:  Here are the conference rankings based on the RPI:

......................2018............2019............2020............Mean Rank
ACC..................#5................#6...............#4................#5.00
PAC-12.............#4................#5...............#8................#5.67
AAC...................#6...............#4...............#9................#6.33
Independents....#7................#9...............#6.................#7.33
SBC...................#11..............#8..............#5...................#8.00
MWC.................#8................#7..............#10..................#8.33
MAC.................#10...............#11.............#7...................#9.33
CUSA................#9................#10............#11..................#10.00

See, this shows how off these computers are this year. While yes, the RPI does correctly capture the SBC being the #1 G5 conference in 2020, the notion that the AAC is behind the MAC as well is ridiculous. The MAC played only two OOC games all year, winning them both vs a CUSA and a MW team. Playing two games is just way too few data points to make comparisons, but somehow the RPI (and most other computers) tried to anyway, with silly results.

That's why it makes no sense for you and "Navy" to keep posting computer numbers for 2020. They are meaningless for true analytical purposes, though they will determine who gets the most CFP money.