If anything disproves the usefulness of the cfrc, it should be this. I used the cfrc to make my picks for my family's bowl pool. This is a confidence pool, so the game most confident you give 34 points to the game you think is most a toss-up you give 1 point to.
Using the CFRC, I used the differences in the rankings between the two teams to determine the confidence point.
The biggest difference was between MSU and TTU so, I picked Tech with 34 points, all the way down to the BCS champ game where there was only a 1 rank difference between Alabama and Texas.
So far the CFRC has gotten the 33 point (Fresno and Wyoming), 32 point (Nevada and SMU), 28 point (WVU and FSU) and 24 point (Houston and USAFA) along with 11 others for an overall record of 11-14 and a point total of 215. This ranks 18th out of 28 in my family and 225,666th in all of espn (better than only 32.9% of the other entries)
This all points to the fact that too many rankings make too much noise in the system
In the jiMpossibleandsarcasMike pool - where I made my own picks I am 14-11 and in the 90th percentile! Obviously my rankings are better than the CFRC and I should be consulted when you are looking for the relative strengths of particular teams!
Happy New Year!
Subscribe to:
Post Comments (Atom)
The CFRC does exactly what is computer rankings are intending do, in general. It ranks how teams actually performed, not predict how teams will perform. CFRC is still the best ranking system in the history of ranking sytems in any sport, on any planet. And if anyone has anything to say about it, I will punch them in the nose.
ReplyDeleteIn other words, I think it is fun.
I should have somehow tracked the rankings from when I made my picks, then I could have determined which ranking was the "best"
ReplyDelete