Helmet Team Ranking Exercise

[FearlessF]

your thread, your choice

[utee94]

probably bearable mixed 50/50 with Tito's

Pretty much anything is better mixed 50/50 with Tito's.

[FearlessF]

Ed Zachery

I did some Tito's mixing last night during the 2nd half of the Oregon debacle

Budweiser wasn't strong enough to deaden the pain

[FearlessF]

What If College Football Hadn’t Wasted Decades On Polls And Just Used A Stinking Playoff?

https://fivethirtyeight.com/features/what-if-college-football-hadnt-wasted-decades-on-polls-and-just-used-a-stinking-playoff/

[OrangeAfroMan]

I was going to get this back going after the season.  Thanks for saving me the trouble of finding it.

Good to see Florida would've led the nation in playoff appearances 

[ELA]

Not to be a homer, but I don't see any way the committee would have put in Alabama over MSU in 2013.

In their methodology they are ignoring how the AP poll, in the CFP era seemed to adjust to match what the committee was doing.

[OrangeAfroMan]

Their method went 16 for 16, so I don't see a reason to just say "nuh uh" and have them change their methodology.

[ELA]

Yes, but their methodology ignores that the AP voters have changed how they've voted in the CFP era.

They wanted to assemble the data as quickly as possible, and that's fine.  And it's a stats blog, they want to find a stats based methodology.  But they also could have gotten together and actually broke it down year by year and probably done a better job.  I don't think a single person here thinks in a million years a 1 loss P5 champ would ever be omitted unless all 5 P5 champs had 1 or fewer losses.  Maybe in their limited sample size, their model fits.  But there are several different methodologies you could use with such a limited sample size and also get a perfect fit.

[OrangeAfroMan]

I think only going back to 1989 was lazy.

[FearlessF]

they also explained only going back to 89

merely for consistency

yes, they could have put much more work into it

[OrangeAfroMan]

If I wanted to go through and nitpick with what I feel would've happened....

1990 Miami and FSU should have much higher percentages.
91 Washington should have the highest chance.
1993 WV should be 4th or worse.  ND (2nd) beat FSU (1st) h2h.
1995 N'western would have no shot.  I know 12% isn't much, but it's still too high vs those 3 teams above it.
1998 OSU should be 2nd.
That 99 Alabama team could beat anybody.  But why is VT so low?  They finished 1st in total O and D that year.
2001 Florida should be 2nd over Oregon.
How in the holy hell is 2002 Miami 4th?
2003 LSU/USC are even...but USC would've been a healthy favorite if they had played, no?
This system seriously would've had Florida a heavy favorite to win it, not knowing the outcome vs OSU?  Wow (2008).
As with '95, 2009 Cincinnati had a zero % chance of winning a playoff with those 3 teams - NONE.
2010- Arkansas was the 2nd-best team, really?  47th in scoring D.  C'mon, man.

[ELA]

I guess brand strength is similar, but I have no idea the methodology for this

[ELA]

I guess brand strength is similar, but I have no idea the methodology for this

So basically they asked 200+ recruits, if they were the #1 recruit in the country, and had offers from every single school, to rank them.
http://picksixpreviews.com/how_to_win_in_recruiting.html
Also, not a surprise is the correlation between brand strength and revenue

[ELA]

But also it makes no sense.  How is Alabama that low?  My guess is that it's skewed by the few who want to "stick it to them" and put them all the way at the bottom or something.