• Help Support TNDeer:

O-H-I-O Number 1 All Time

My post must have set off alarm bells. How did you get on that so fast!?! Call me impressed!
 
Congrats to Ohio State for this honor, but I had my resident math geek take a look at the article, and he quickly pointed out the flaw in their methodology: using their system, the AP weighted all weekly polls identically, while in reality the later in the season a poll is taken the more accurate it should be, since those later polls are based on more data points (games). Of course, the only poll that really matters each season is the final poll.

A better and simpler way to do it would have been to take the final top 10 teams for each of the last 100 years. Give 10 pts for every #1 down to 1 point for every #10, and sum the total. That would have resulted in a equally weighted result of the most accurate poll for each year. Unfortunately it looks like the AP let a communications major try his hand at stats. :?


DISCLAIMER: My "resident math geek" is a Bama grad and current grad student. He is completing an emphasis is statistical analysis, fwiw.
 
Interesting list........
 

Attachments

  • Screenshot_2016-08-02-22-58-34.png
    Screenshot_2016-08-02-22-58-34.png
    22.8 KB · Views: 1,333
Zulu":3id679zl said:
Congrats to Ohio State for this honor, but I had my resident math geek take a look at the article, and he quickly pointed out the flaw in their methodology: using their system, the AP weighted all weekly polls identically, while in reality the later in the season a poll is taken the more accurate it should be, since those later polls are based on more data points (games). Of course, the only poll that really matters each season is the final poll.

A better and simpler way to do it would have been to take the final top 10 teams for each of the last 100 years. Give 10 pts for every #1 down to 1 point for every #10, and sum the total. That would have resulted in a equally weighted result of the most accurate poll for each year. Unfortunately it looks like the AP let a communications major try his hand at stats. :?


DISCLAIMER: My "resident math geek" is a Bama grad and current grad student. He is completing an emphasis is statistical analysis, fwiw.

Now let's not get carried away and bring in that rocket science stuff. :D
 
Zulu":2tvuh5nd said:
Congrats to Ohio State for this honor, but I had my resident math geek take a look at the article, and he quickly pointed out the flaw in their methodology: using their system, the AP weighted all weekly polls identically, while in reality the later in the season a poll is taken the more accurate it should be, since those later polls are based on more data points (games). Of course, the only poll that really matters each season is the final poll.

A better and simpler way to do it would have been to take the final top 10 teams for each of the last 100 years. Give 10 pts for every #1 down to 1 point for every #10, and sum the total. That would have resulted in a equally weighted result of the most accurate poll for each year. Unfortunately it looks like the AP let a communications major try his hand at stats. :?


DISCLAIMER: My "resident math geek" is a Bama grad and current grad student. He is completing an emphasis is statistical analysis, fwiw.

I don't think this would work since FSU has only been playing football for 68 years.
 

Latest posts

Back
Top