An Interesting Blind Resume Test (Please Stop Screaming About SEC Bias)

Landlord

Banned
So a user over on reddit had an interesting idea. He made a blind resume system for users to rank teams on.

The resumes showed a team's record, record vs Top 25, record vs Top 10, strength of schedule, margin of victory, and description of loss. no names, no conferences.

Almost 900 regular ole college football fans just like you and me filled out the survey, ranking the teams in the order they thought was fair based on blind resumes. Here were the results:

GPi6XYL.png


Now, obviously this is a very imperfect means of ranking. Even more obviously, there is no such thing as a perfect means of ranking. Even more obviously still, teams can't be ranked in a vacuum without influence of the general college football landscape, including ESPN. But, remarkably, 900 regular fans without knowledge of the names or conferences of the teams they were actually voting for, came up with a poll that was remarkably close to the poll of the CFB Playoff Selection Committee.

 
So.......we have discovered how the AP goes about ranking teams?

 
Last edited by a moderator:
Interesting. I'm interested in knowing if the wins against top 25 that were listed were ranked at time of playing or how they are ranked now.

 
What top 25 and top 10 rankings did they use for the wins against on the resumes? Doesn't that kind of eliminate the blind thing since these rankings supposedly have the SEC bias?

 
I still wouldn't use the teams' records vs. top-25 or top-10 opponents. All that does is transfer the AP or Coaches poll rankings into the new poll.

 
By conference:

3 ACC adds up to be +1 differential (+ 1/3 per school)

6 SEC adds up to be 0 differential

5 Big 12 adds up to be 0 differential

5 PAC adds up to be -3 differential (- 3/5 per school)

3 B1G adds up to be -4 differential (- 1 1/3 per school) (Thanks osu)

So the ACC is slightly under rated, and *gulp* the B1G is the most over rated

 
Last edited by a moderator:
So a user over on reddit had an interesting idea. He made a blind resume system for users to rank teams on.

The resumes showed a team's record, record vs Top 25, record vs Top 10, strength of schedule, margin of victory, and description of loss. no names, no conferences.

Almost 900 regular ole college football fans just like you and me filled out the survey, ranking the teams in the order they thought was fair based on blind resumes. Here were the results:

GPi6XYL.png


Now, obviously this is a very imperfect means of ranking. Even more obviously, there is no such thing as a perfect means of ranking. Even more obviously still, teams can't be ranked in a vacuum without influence of the general college football landscape, including ESPN. But, remarkably, 900 regular fans without knowledge of the names or conferences of the teams they were actually voting for, came up with a poll that was remarkably close to the poll of the CFB Playoff Selection Committee.

Thing is though folks wouldn't be complaining nearly as much if this was the committee's result. OM being in the playoff over Oregon was the tipping point. Ya things aren't over by any means and the committee said the difference between 4, 5, 6, and 7 was paper thin but this is all about picking a top4 the committee's top4 was not a good sign.

 
Back
Top