Perhaps the BEST B1G Forum anywhere, here at College Football Fan Site, CFB51!!!

If Georgia's gonna keep on being so disappointed and then losing the Sugar Bowl, it's long past time to reset expectations.

But it's such a convenient excuse for the SEC, every time they lose to an "undeserving" B12 team.  Bowl record for B12 vs. SEC was 3-1 last year.  I'm sure all of those SEC teams were just terribly disappointed and "didn't want to be there."  Same shit we hear every year.  Lather, rinse, repeat.

I don't mean it as any "excuse", in some ways I view it as worse than showing up and getting beat.  My point is they likely are disappointed in their season and that may have some impact on how they play.  Losing 3/5ths of their OL is probably more of a factor.  Baylor is probably better than any team they beat this year, at least as good as.  

And sometimes a huge blowout.
31-0
59-20
38-0
24-7
24-6
30-3
44-16

Just because the result of the game was a huge blowout doesn't necessarily mean that the teams were all that unevenly matched.  We all know that sometimes a superior team loses to an inferior team.  It happens.  Illinois wasn't better than Wisconsin this year but when they played, Illinois won.  It happens because sometimes teams have a great game and sometimes teams have a terrible game.  

The same applies to games among relative equals.  If one of them has a great game and one of them has a terrible game we are going to see a blowout.  It happens.  Ohio State wasn't 59-0 better than UW in 2014 but one night in Indianapolis Ohio State played 59-0 better than Wisconsin.  

I think one has to spell out what you want in a playoff before thinking about how it should be designed.  Do you want to identify the "best team" consistently?

Good luck with that.

That's impossible which is why people should stop attempting to do so.

they've been attempting to do it since the 1800's, it's not going to stop

But have they really?  I think there's a difference between a news wire service poll attempting to rank the "best" teams at the end of the season, versus an Alliance/BCS/playoff attempting to matchup the "best" teams and have them play and produce a "champion."  

They've been doing the former for almost a 100 years, but the latter has only occurred in the last 20.  Before that, we were participating in the Old Bowl system where occasionally we might get a 1-2 matchup but often we had 2 or even 3 bowls with the 1,2,3,4,5 teams in them, any of which might end up being designated by a wire service poll as the "champion" after the games were played.  And of course, before that, they didn't even rank them after the bowl games, and it was just the regular season that mattered.

All of that to say, this insistence on finding and determining the SOLE national champion is a very new thing in the history of D1-A college football.  It wasn't that way for the first 130 years or so...

The question is whether this is just a sample size issue, or indicative of a wider problem in the polls/committee?

To be honest, with proper seeding, a 1 seed should win the whole thing more than 25% of the time, but less than 100%. The question is how close to those two extremes is "correct", and that question boils down to just how much difference there is between the strength of #1 vs #2, #3, and #4. Still, a #1 seed in my opinion should be expected to win the whole thing maybe 30% of the time at minimum, or 40% of the time at maximum.

This year, for example, Vegas has LSU as a 14 point favorite over OU, and ESPN's FPI has LSU's win probability at 67.5%. If we make the assumption that LSU, as the higher seeded [and therefore supposedly stronger] team than OSU/Clemson would have a >50% win probability in that game, it would still require a 60% win probability in the final (over a team perceived significantly stronger than OU this year) to reach 40% win probability for the #1 seed.

So we've had 5 instances of the CFP, and the expected number of times for a #1 seed to win is somewhere between 1.5 and 2 times. That it's been 0 in only 5 instances might be simple due to small sample size.

I think it is mostly just sample size.  In the mathematical model that you built here the #1 seeds should be about 4-2 in six semi-finals.  They are 3-2 so far in five.  That is only a one game difference.  Then they should be a little better than 50/50 in CG's or roughly 1.5 out of the three so far.  That is only a game and a half difference.  

Another thing to note is that the #4 seeds have been feast-or-famine.  Two (tOSU in 2014 and Bama in 2017) won the whole thing but the other three each went 0-1.  I think part of that is that two things:

• As it turned out tOSU was actually BETTER in 2014 after their starting QB went down.  Nobody would have expected that.  
• Bama in 2017 was "better" than #4 but they didn't "deserve" a better seed because they didn't win their conference.  I think if the committee had simply seeded based on how good they thought the teams were rather than on what they deserved, Bama would have been higher and possibly #1.  

Or it might not... You all know my thoughts on how to crown a champion in college football. Some might argue that if a committee made up of CFB experts, whose sole solitary job is to find the best four teams in the land (and seed them according to strength), and they can't pick the #1 team, how could pollsters be any better? Even worse, if the #4 team has a winning record and has won 2 out of 5 so far, might it be that they're actively seeding BADLY such that the consensus weakest team of the 4 is the second best performing seed, it suggests that maybe the "experts" don't know as much as we thought.

It's for that reason that I eschew the idea that only the "best teams" deserve to be in the BCS or CFP. Because we're notoriously bad at determining who the "best teams" are. So I fall back on the hybrid system of "most deserving" teams plus a few "at large" teams that are deemed worthy but didn't meet the objective criteria.

Go to an eight-team playoff. Either the 5 power conference champs plus 3 at-large, or the 5 power conference champs plus 1 highest ranked G5 conference champ, plus 2 at-large.

Because if even the committee can't get seeding right, how do we even know they're getting the best 4 teams right? ESPN FPI has five teams higher than OU's FPI this year. Are we really sure they're the fourth best team in the land?

I have long assumed that we are heading for an eight-team playoff with:

• The five P5 Champions
• The highest ranked G5 Champion
• Two at-large teams

The thing I don't like about it is that it will inevitably lead to some bad P5 Champions getting in.  Some team will have a weak OOC, lose a couple conference games but sneak into a CG based on a tiebreaker and pull off an upset then head to the CFP at 9-4 and ranked #15.  To me, that seriously diminishes the "every game matters" (or at least could matter) concept that has been so integral to this sport for so long.  

All of that to say, this insistence on finding and determining the SOLE national champion is a very new thing in the history of D1-A college football.  It wasn't that way for the first 130 years or so...

this is true

and I'm not a fan

under the old system we could easily have had Clemson/Ohio St. and LSU/OU in bowls

or LSU vs OSU

the plus 1 game will be fun to watch as a fan, but I don't see it as any more of a big thing than a MNC from the 80's or 90's - including a split in 97 for Nebraska and Michigan
That would have been a fun game to watch as a fan