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

I think often a ranking based on who would beat who (a power poll) is different from an AP kind of a poll.

In the AP, we almost never see an 8-3 team ranked ahead of a 10-1 team (both P5s).  They basically ranked teams on record, with some minor adjustments.  That 8-3 team could well have played a very tough schedule and looked good even in losing while the 10-1 team played almost nobody and edged a lot of weak teams.

Maybe the best "poll" is based on Vegas theoretical odds.

If the on-field results take a back seat to the beauty contest, then you'd be foolish not to run up the score. 

Except when Baylor beat TCU head to head, and finished with the same record, TCU was still higher based on the eyeball test, or maybe the "better loss" test since their loss was to Baylor, and Baylor's loss was to WVU.

Same would apply to 2015 MSU-OSU.  Each had 1 loss, OSU passed the eyeball test over MSU (even the most homer MSU fan can admit that), and OSU's one loss to MSU was certainly better than MSU's 1 loss to Nebraska.  And OSU had helmet factor.  But MSU was still ranked higher, because they were the conference champ, and OSU wasn't.

That's why I continue to believe helmet is not a factor, and yes a 1 loss Big Ten champ Indiana I believe would get in over a 1 loss non champ Alabama.

On the Baylor/TCU debate:  I agree and just want to point out that not only was TCU's loss to a better team (Baylor) but it was also only by a FG whereas Baylor's loss to WVU was by two TD's.  
On the second part, I have been having this debate with @utee94 for ages now and I never thought of that example, thank you.  MSU/tOSU 2015 is strong evidence that the committee is NOT biased in favor of helmet teams for all of the reasons you laid out.  

I strongly agree with @OrangeAfroMan on this issue.  I have always thought that the oddity of using H2H as a tiebreaker is that in most cases it rewards a team with a worse loss.  Ie, if Michigan loses to Rutgers but beats Ohio State such that both finish 8-1 it is odd to reward the team that lost to a horrible Rutgers team with a Championship.  I understand why we use it but I see no reason why the committee should be forced to use it.  

Another example:
If Notre Dame loses a game and finishes 11-1 while Michigan wins out to finish 12-1 I would absolutely put Michigan in the CFP over Notre Dame despite the H2H result.  Michigan would have a much better loss AND a much better collection of wins.  

I strongly agree with @OrangeAfroMan on this issue.  I have always thought that the oddity of using H2H as a tiebreaker is that in most cases it rewards a team with a worse loss.  Ie, if Michigan loses to Rutgers but beats Ohio State such that both finish 8-1 it is odd to reward the team that lost to a horrible Rutgers team with a Championship.  I understand why we use it but I see no reason why the committee should be forced to use it.  

Another example:
If Notre Dame loses a game and finishes 11-1 while Michigan wins out to finish 12-1 I would absolutely put Michigan in the CFP over Notre Dame despite the H2H result.  Michigan would have a much better loss AND a much better collection of wins.  

It also means Michigan has a win over OSU, while OSU has a win over Rutgers.
So the gap between the quality of the different win matches the gap in quality of loss.

One of the issues when trying to decide between these teams are so few of them play each other.  So if I get a result between two teams it means something to me.

I find it odd that other games mean more to you than a game between two teams you are comparing.

Lastly, I’m probably more interested in a team’s ceiling than their basement. Best win probably means a little more to me than worst loss. I know anyone is capable of getting crushed by Bama. I want the precious few capable of beating them.

One of the issues when trying to decide between these teams are so few of them play each other.  So if I get a result between two teams it means something to me.

I find it odd that other games mean more to you than a game between two teams you are comparing.

A sample size of 11 > a sample size of one (h2h)