1. Clemson
2. Bama

3. Georgia
4. TCU (Resume's much better W/Ls than UGA, but got insanely lucky to win last 2 games while UGA has been taking teams to the woodshed)

5. Penn State (I don't really think Penn State is actually that good... think they get nuked by the first team with a good d-line that they play)
6. Wazzu (swing your sword)
7. Washington (think Washington >> Penn State > Wazzu in power, but Washington's resume is terrible)

8. Oklahoma (loss to ISU at home is bad, but recoverable. Probably still make playoffs if they win out. Dominant win @ OSU is still pretty great)
9. San Diego State (win @ ASU, beat Stanford at home, beat NIU who is basically a mid-tier G5 team (won @ Nebraska by 4, lost to BC by 3))
10. Ohio State (Resume is totally devoid, but power rating is still high)
11. Miami
12. UCF
13. USC
14. NC State (Loss to USCe where they outgained them huge is looking like a total fluke... but man USCe sucks)
15. Virginia Tech
16. Michigan (Offense is bad and should feel bad)
17. Auburn (nothing even resembling a decent win, but the loss @ Clemson is good I guess?)
18. Notre Dame
19. Michigan State
20. Wisconsin (Call me when they play someone in the top 50)

21. Oklahoma State
22. Texas Tech (Beat ASU and a very decent Houston team, only loss is by 7 to Ok State)
23. Stanford
24. South Florida
25. West Virginia

Close:
Virginia, Georgia Tech

Pretty clear tiers forming in the rankings at the point.

1. Miami

Biggest gap conceivable be the human mind

2. Everyone else

Looks good to me.

So I've been doing some serious thinking, and we gotta figure out a way to have Miami and Georgia play each other in the playoffs.

Let's get to work!

LOL at anyone ranking Wazzu ahead of UW especialky a Pac-12 fan. Nice troll.

1 Alabama (1) Still a conference road win
2 Clemson (2) I have a theory that Wake Forest sucks
3 Georgia (4) Vaguely remember losing Eason was supposed to matter?
4 TCU (6) A rough game but I got no other undefeated with a comparable overall resume
5 Washington (7)
6 Washington St (9)
7 Penn St (8)
8 Ohio St (10)
9 Michigan (5) I don't see it as that bad of a loss, plus they destroy Penn State
10 Virginia Tech (12)
11 Oklahoma (3) I know it's stupid to put them below tOSU but they also barely handled Baylor last week. Iowa State is a bad team and they blew a huge lead against them at home. Much, much worse loss than Sparty.
12 USC (13)
13 Miami FL (15)
14 Notre Dame (14)
15 Auburn (20) Offensive line looks much better after reshuffling and the QB better with game experience. OTOH, it probably helps to not be playing Clemson.
16 Oklahoma St (17)
17 San Diego St (18)
18 Louisville (11)
19 UCF (24) The teams they kicked the **** out of are now going out and kicking the **** out of other teams.
20 South Florida (19)
21 Wisconsin (16)
22 Stanford (25)
23 Michigan State
24 NC State
25(T) Utah (22)
25(T) West Virginia

Out:

Oregon (21)
Florida (23)

* Except Maryland.

current theory on bama's 2nd half performance is that they were looking ahead, PAST arkansas and getting excited about hanging 100 on the vols in 2 weeks

HOT TAKE: the SEC has gotten a lot worse because LORD SABAN has totally wrecked the conference. Enough top recruits who aren't going to bama are going to other conferences that the whole thing is out of balance.

Washington is better than Wazzu but Wazzu's resume is >>>>>>>> Washington's, not sure how either of those is debatable.

Choose your flavor for which matters to you more

Is it even that hot of a hot take? It's not just the recruits, it is the coaches. Urban Meyer ran off to the big 10 and Les Miles was fired because he wasn't Saban. If it wasn't for Saban letting Kirby Smart off his blood contract the conference would be a complete mess.

That any knowledgeable fan would have Virginia even close to Top 25 is mind blowing to me, in a good way. They've been so bad that it even being respectable has been a foreign concept.

The schedule gets really hard at later, but I'll enjoy a few wins right now. At least bowl eligibility is within reach.

Duke's decent, thumping Boise on the blue turf is a very good win... losing to Indiana isn't a great look but it's not like they lost to Kentucky or something

Power ranking I feel fine about. Yeah it's weird to see OU lose to Iowa State (at home!) an only drop a bit, but meh. Model lines continue to correlated well with vegas lines and continue to do a tiny (inside the vig) bit better job predicting outcomes. Although I'm overdue for an new analysis on that topic.

Achievement is a more bit interesting and subject to intuition. Especially its behavior at the margins, both in terms of terrible teams and in terms of game outcomes.

Some examples..

Miami (Fl). If you exclude their opener against Bethune Cookman their rank 3rd in achievement. But they "only" beat Bethune Cookman by 28, and Bethune Cookman is fairly bad. Bethune-Cookman lost to FAU 45-0, for example. If Uconn (who are themselves terrible) and Bethune-Cookman played I'd have Uconn as a 28 point favorite. But that said, maybe Miami eased up, or used it as somewhat of a warm-up game, etc. How much that game should factor into their achievement rating is an interesting question, at least for me.

WSU vs. UW. I have UW at 6 and WSU at 16. The reason is basically WSU's games against Montana State and Boise State

WSU has two games with a rating < 0.900. Montana State (won by 31, 0.719) and Boise State (won by 3, 0.641)

UW has one such game. @Rutgers (won by 16, 0.823).

The achievement model considers the other games to be roughly on par. i.e. blowing out cal / colorado is rated similarly to a close with over USC at home or a 3 TD win over Oregon on the road. I'm not exactly saying that's the right path or w/e. Maybe achievement should be more focused on wins than margin of victory, I dunno.

But to dive a little deeper. UW beat Cal by 31. If they played today Bama would be a 28 point favorite over Cal at home. So UW did slightly better than Bama would be expected to do. So UW's rating for that game is slightly higher than Bama's current power rating.

WSU beat USC by 3. If they played today Bama would be favored by 12 over USC (per the model) so WSU did worse than Bama would be expected to do. Consequently, WSU gets a lower achievement rating for beating USC by 3 than UW did for beating Cal by 31.

Following? Not following at all? If yes, where am I going awry?

Hey now, we are 5-1 (but yes the worst 5-1 team in the country)

Oregon State HC Gary Andersen out

Well they're god awful

Goofball already nailed it- UW's resume is better because they have absolutely blown out the terrible teams they have played.

USC is not good and super injured because they have an Advocare salesman as head S&C coach.

Oregon sucks and started a true frosh QB.

Wazzu didn't look good in any other game.

If two teams are going to play in 5 weeks and you have the double-digit underdog ranked higher...you are doing something wrong.

Putting auburn at 17 seems way too low. I imagine in a neutral field they aRe a favorite over At least 10 teams you have ranked before them.

If you're going to start including MOV then you're basically just doing a straight (albeit bad) power ranking... losing by 1 is basically the same as winning by 1.

And, if you're just gonna rank teams by power, then might as well not even play them after a certain point, once we already know how good teams are.

Wazzu has beaten *THREE* teams better than the best team Washington has played... six games into the season that's a pretty stark contrast.

(To be fair, the resume is also better than Alabama's, so idk why they get a magic pass and Washington doesn't)

I agree that UW is probably 10 points better than Wazzu, and presumably you agree that Wazzu's resume on a w/l basis is substantially more impressive. At that point, you just pick & choose what you think the purpose of the ranking should be.

Goof, is there a better metric by which to measure early season games against patsies?

I feel like in the grand scheme of things you could use (YPP, Something like weighted YPP where you give more weight to the first half, some to the 3rd Q, and very little to the 4th Q just as two examples), it seems like total score between a good team and a bad team is just not all that good of an indicator of how good a team is. For example, most times a team wins 31-0, they will dominate a team much more thoroughly than winning 65-21 or something like that.