CFP Discussion

1979jacket

Ramblin' Wreck
Messages
626
i see
1 vs 8 uga vs bama
2 vs 7 mich vs oregon
3 vs 6 fsu vs wash
4 vs 5 oreg st vs ohio st

oregon st as highest rated 4th conference gets a bye
 

1979jacket

Ramblin' Wreck
Messages
626
It is puff, puff, pass, not puff, puff, pufffffffffffff! Oregon State?
If I understand the current rules for next year - 6 conference champs make the playoffs with top 4 getting byes. Oregon St which could be in PAC2 or in Mountain West is the 4th highest ranked conference champ right now ahead of Utah from Big12 and Tulane from AAC. I'm using next year conference alignments which to me is what counts. Full lineup
1​
ugasec
2​
michbigi
3​
fsuacc
4​
oregstmwest
5​
ohio stbigi
6​
washbigi
7​
oregbigi
8​
texsec
9​
alasec
10​
pennstbigi
11​
utahbig12
12​
tulaac
 

UgaBlows

Helluva Engineer
Messages
6,803
FSU is the ACC realistic shot for the playoffs and they will need to stay undefeated. I think a 1 loss ACC Champ is behind a 1 loss Pac 12, Big Ten, SEC or Texas.
I have a feeling UL is gonna beat them in the acc championship game
 

Oldgoldandwhite

Helluva Engineer
Messages
5,760
If I understand the current rules for next year - 6 conference champs make the playoffs with top 4 getting byes. Oregon St which could be in PAC2 or in Mountain West is the 4th highest ranked conference champ right now ahead of Utah from Big12 and Tulane from AAC. I'm using next year conference alignments which to me is what counts. Full lineup
1​
ugasec
2​
michbigi
3​
fsuacc
4​
oregstmwest
5​
ohio stbigi
6​
washbigi
7​
oregbigi
8​
texsec
9​
alasec
10​
pennstbigi
11​
utahbig12
12​
tulaac
That’s what I’m afraid of. The BIG and SEC get half the spots. Based on them being ranked high early.
 

Oldgoldandwhite

Helluva Engineer
Messages
5,760
It's too bad that UL lost to Pitt. Two undefeated teams playing in the ACC Championship game would guarantee that the ACC had one of the four spots. But a one loss ACC champion has virtually no chance of getting in.
Depends. Usually chaos the last fourth of the season
 

TechPhi97

Ramblin' Wreck
Messages
749
Location
Davidson, NC
Yes. The difference is nobody that matters cares (or reacts) to our opinions. The AP poll gets entirely too much run & weight. I could not care less what 60+ journalists think about who's good at football. It's "interesting" info ... equally as interesting as the "opinions" folks who post here.


Correct. Computer algorithms/calculations can & do ... do a better job figuring out which teams are better than the next. Not perfectly... which is quite alright, but way better than a guy/girl trying to parse which team should be 14th or 23rd.

Even the CFB playoff committee claims to use statistical models to inform their rankings. The AP poll doesn't have any criteria applied to it as far as I know.

Imagine a world where no rankings are referenced or calculated or voted on before 6 weekends of football. How would we know which games are "marquee" ... How would GameDay figure out where to set up their show? People would be lost. 😅
When you say that computers / algorithms are better at picking the best teams, it begs the question: what do you mean by best teams? There are basically two ways to look at it; (1) best performance to date, or (2) predicted for the future. The models in (1) are still based on human inputs, because humans decide the factors to include and the definition of those factors. Take strength of schedule - based on record, record and point differential, record and points without garbage time, etc. Like Colley, these are pretty good at explaining the past.

Most models that do (1) also try to do (2), and honestly they do a terrible job. I looked at the FEI rankings for the year and compared their prediction of points differential for games against the final betting lines, and they were only right about 50% of the time. On my phone now, but will post a GSheet later today. These models are no better than Average Joe at predicting games, which isn’t surprising.
 

FredJacket

Helluva Engineer
Messages
6,238
Location
Fredericksburg, Virginia
When you say that computers / algorithms are better at picking the best teams, it begs the question: what do you mean by best teams? There are basically two ways to look at it; (1) best performance to date, or (2) predicted for the future. The models in (1) are still based on human inputs, because humans decide the factors to include and the definition of those factors. Take strength of schedule - based on record, record and point differential, record and points without garbage time, etc. Like Colley, these are pretty good at explaining the past.

Most models that do (1) also try to do (2), and honestly they do a terrible job. I looked at the FEI rankings for the year and compared their prediction of points differential for games against the final betting lines, and they were only right about 50% of the time. On my phone now, but will post a GSheet later today. These models are no better than Average Joe at predicting games, which isn’t surprising.

No argument from me. Yes... no computer model can perfectly account for all the variability. However... at least once the model is created (& refined) it provides a repeatable method to determine it's ratings for teams. I'd submit... pollsters methods aren't repeatable. I simply believe computers are better than polls... how much better? Difficult to say.

The reason nothing approaches perfect is humans play the games and are unpredictable.... nothing (computers or people) can account for all that. ...which is why we keep watching.

Your last sentence seems like a stretch if you meant it literally.
 

slugboy

Moderator
Staff member
Messages
11,475
No argument from me. Yes... no computer model can perfectly account for all the variability. However... at least once the model is created (& refined) it provides a repeatable method to determine it's ratings for teams. I'd submit... pollsters methods aren't repeatable. I simply believe computers are better than polls... how much better? Difficult to say.

The reason nothing approaches perfect is humans play the games and are unpredictable.... nothing (computers or people) can account for all that. ...which is why we keep watching.

Your last sentence seems like a stretch if you meant it literally.
Computer models can be biased, and don’t have a way to say “I watched these teams and Penn State tackles better and plays more solid defense”.

One the plus side, they don’t say to the press “we’re going to pick the teams that have earned it” one day and then turn around and pick a Bama team that didn’t make the conference championship the next. They should at least be consistent.
 

orientalnc

Helluva Engineer
Retired Staff
Messages
9,876
Location
Oriental, NC
Forget about models and polls for a while. Tomorrow there are four HUGE games that could impact the CFP and one of very special interest to us.

Michigan/Penn St - Penn State needs to win this game and have Michigan beat tOSU in two weeks.
Ole Miss/uga - A uga loss here could put their CFP hopes in jeopardy.
Washington/Utah and Oregon/USC are the key games in the PAC as Washington and Oregon appear headed for a rematch.
GT/Clemson - Could this be the year???
 

roadkill

Helluva Engineer
Messages
1,815
Why does it seem the pups have almost every big game at home?
Not only that, but they only have 4 true road games this year. And one of those is us, which could arguably be called a neutral site as of late.

Although I suspect they just got lucky with playing the best teams at home, makes you wonder if they don't have someone with an inside track in the SEC scheduling committee. Of course, conferences would never show favoritism in scheduling, right?
 

SOWEGA Jacket

Helluva Engineer
Messages
2,074
That’s what I’m afraid of. The BIG and SEC get half the spots. Based on them being ranked high early.
That‘s how this rigged sport has operated for over 100 years. The change will occur when it expands past 12 and then the teams who have been given preferential treatment will now have to play multiple legit teams in a row which will allow the teams relegated to “mid tier” status by the polls for decades to start showing what they can do. There will be a ton of upsets which will prove the farce that has been going on. Take UGA - yes, they are a top tier program, but they have a 2 game regular season with their scheduling and then Kirby has to game plan for a 4 team playoff. Let’s see how Kirby and his staff fare when they have to game plan for multiple legit teams week after week in a 24 or 32 team tourney without the bye weeks and powder puffs in between.

This is what the UGA’s, Bama’s, and Ohio States have been protected from for 100 years. And it’s always funny how when a big boy team is having a down year, say a 9-3 or 8-4 year they still get matched up in bowls with teams who are 7-5 or 6-6. It’s all about protecting the status quo. Just like we see in March Madness when 1s and 2s get upset we’ll see the same in football - because it happens in all sports with a true playoff.

Once the 12 team playoff brings in the viewership and dollars it will expand within a decade. That’s where teams like GT can finally get a fair shake on those years we are legit. Sure, UGA, Bama, Clemson, etc will get in every year but once the little guys who go 10-2 or 9-3 get in you’ll see the big boys pucker factor go way up.
 

FredJacket

Helluva Engineer
Messages
6,238
Location
Fredericksburg, Virginia
Not only that, but they only have 4 true road games this year. And one of those is us, which could arguably be called a neutral site as of late.

Although I suspect they just got lucky with playing the best teams at home, makes you wonder if they don't have someone with an inside track in the SEC scheduling committee. Of course, conferences would never show favoritism in scheduling, right?

I realize this is a GT forum AND I hate UGA... but the lack of objectivity is pretty extreme in some cases. UGA was scheduled to travel to Oklahoma this season. That got canceled (by the SEC) when realignment happened.

They have future opponents like Clemson, UCLA, FSU & Loiusville.

Not to mention... Ga Tech accounts for one UGA's "cupcakes" out of conference.

TO HELL WITH GEORGIA!!
 

roadkill

Helluva Engineer
Messages
1,815
I realize this is a GT forum AND I hate UGA... but the lack of objectivity is pretty extreme in some cases. UGA was scheduled to travel to Oklahoma this season. That got canceled (by the SEC) when realignment happened.

They have future opponents like Clemson, UCLA, FSU & Loiusville.

Not to mention... Ga Tech accounts for one UGA's "cupcakes" out of conference.

TO HELL WITH GEORGIA!!
I was speaking of their 2023 season game locations. Not about their SoS or future opponents. Did I say anything that was not factual?

Also, do you have a good source for the SEC canceling the Oklahoma game due to realignment? I ask because Texas-Alabama happened as planned.
 

bobongo

Helluva Engineer
Messages
7,550
I was speaking of their 2023 season game locations. Not about their SoS or future opponents. Did I say anything that was not factual?

Also, do you have a good source for the SEC canceling the Oklahoma game due to realignment? I ask because Texas-Alabama happened as planned.

The Southeastern Conference has directed the University of Georgia and the University of Tennessee to postpone scheduled football games versus Oklahoma because the transition of Oklahoma into the SEC will not allow for the involved institutions to fulfill their respective contractual nonconference home-and-home appearance obligations.

Because the second nonconference game in each of the Georgia-Oklahoma and Tennessee-Oklahoma series is scheduled to take place after Oklahoma joins the SEC in 2025, the conference is directing the postponement of the Georgia at Oklahoma game in 2023 and the Oklahoma at Tennessee game in 2024 until such time that those matchups become part of the conference rotation of games in future years.
 

FredJacket

Helluva Engineer
Messages
6,238
Location
Fredericksburg, Virginia
I was speaking of their 2023 season game locations. Not about their SoS or future opponents. Did I say anything that was not factual?

Also, do you have a good source for the SEC canceling the Oklahoma game due to realignment? I ask because Texas-Alabama happened as planned.
The out of whack game locations were directly related to cancelation of road game at Oklahoma.

Sorry... my comment about lack of objectivity was not aimed at you ... more a general feeling I have about a lot of posts on this subject (UGA & SEC biases). For record... I think you can find evidence of bias but not nearly approaching the amount discussed in many posts.
 

CEB

Helluva Engineer
Messages
2,551
Forget about models and polls for a while. Tomorrow there are four HUGE games that could impact the CFP and one of very special interest to us.

Michigan/Penn St - Penn State needs to win this game and have Michigan beat tOSU in two weeks.
Ole Miss/uga - A uga loss here could put their CFP hopes in jeopardy.
Washington/Utah and Oregon/USC are the key games in the PAC as Washington and Oregon appear headed for a rematch.
GT/Clemson - Could this be the year???
Michigan v PSU is huge test for Mich in my opinion. Yes, Michigan looks complete. Yes, they are beating people big. Look harder at their schedule though…. They SHOULD be beating those people big. They have played no one yet. I said in another post that Michigan is being rewarded for playing even less of a schedule than UGA.
Michigan has a crazy November slate, starting tomorrow. Not to mention the other garbage swirling around the program right now…We’re about to learn a lot about Michigan tomorrow.

Other games you mention could have huge impact depending on outcome but Michigan v Penn St will DEFINITELY have an impact regardless.
 
Top