Muahahaha
Junior
Check it out. What do you think?http://teamrankings.com/ncf/27confratings.php3
-
Follow us on Twitter @buckeyeplanet and @bp_recruiting, like us on Facebook! Enjoy a post or article, recommend it to others! BP is only as strong as its community, and we only promote by word of mouth, so share away!
- Thread starter Muahahaha
- Start date
Buckinghorse
Will work for bpCash
I think that it must be a work in progress as it makes no sence now.
Last edited:
Upvote
0
1998_OSU_Grad
Junior
i have so many inappropriate things to call the developer of this... mildly retarded, idiot, assbag... those are the top things that come to mind.
if anyone is in the webpage developing business, don't hire this guy.... i'm guessing he meant for people to read it from bottom to top, and to disregard the column on the far left????
if i were a monkey, i'd throw poop at him.
if anyone is in the webpage developing business, don't hire this guy.... i'm guessing he meant for people to read it from bottom to top, and to disregard the column on the far left????
if i were a monkey, i'd throw poop at him.
Upvote
0
Actually, Mike Greenfield has a very good ratings page, see the main page at http://teamrankings.com/ncf/. He has been impressive in predicting Buckeye winning and losing margins this year. His strength of schedule rating is not the traditional kind that we understand and appears to be in a nonlinear unit. Just ignore the SOS page and go to the main page. Click Buckeyes and see their page.
He will not post rankings until the season is well underway, but when his basketball rankings come up, they also were scary good last year.
He will not post rankings until the season is well underway, but when his basketball rankings come up, they also were scary good last year.
Upvote
0
sandgk
Watson, Crick & A Twist
OK - chiming in FWIW,
Here is a little Bio from the site
That page is meant to tell you how the teams are ranked. But all it uses is what is known in the trade as "hand-waving." No substantial description of the underlying math - and I think I appreciate why Mike Greenfield is unclear about his precise methodology. He wants to protect the same.
The rating system was developed by him, I believe, as an aide for bettors.
What he is trying to do is apply an RPI formula to football.
That probably means his "predictor," which is what you as a user would be interested in where you to be betting on games, is highly accurate when right of bell curve (top of his rank) meets left of bell curve (bottom of his rank). It likely has far less value in the middle of the pack. Furthermore, it is likely more predictive when looking at team vs team than when pooled data tries to predict the outcome of a binned population of teams - i.e., conference vs. conference.
As evidence of this failing one only has to look at the relative strength ranking for the SEC versus the Big East in the original linked table. One need look no further to understand that the conference ratings tell us nothing of value. In effect, if this were a Ranking correlation study you would have a low R and R^2 coefficient for the data implemented in that fashion.
The team by team ratings do have value however.
Here is a little Bio from the site
That page is meant to tell you how the teams are ranked. But all it uses is what is known in the trade as "hand-waving." No substantial description of the underlying math - and I think I appreciate why Mike Greenfield is unclear about his precise methodology. He wants to protect the same.
The rating system was developed by him, I believe, as an aide for bettors.
What he is trying to do is apply an RPI formula to football.
That probably means his "predictor," which is what you as a user would be interested in where you to be betting on games, is highly accurate when right of bell curve (top of his rank) meets left of bell curve (bottom of his rank). It likely has far less value in the middle of the pack. Furthermore, it is likely more predictive when looking at team vs team than when pooled data tries to predict the outcome of a binned population of teams - i.e., conference vs. conference.
As evidence of this failing one only has to look at the relative strength ranking for the SEC versus the Big East in the original linked table. One need look no further to understand that the conference ratings tell us nothing of value. In effect, if this were a Ranking correlation study you would have a low R and R^2 coefficient for the data implemented in that fashion.
The team by team ratings do have value however.
Upvote
0