The methodology is posted. Its a weighted average of market implied power rankings, where the recency of the market implied rating impacts the weight by the equation = 1/(weeks ago + 0.2).
The coefficient there was the result of back testing optimization, which is going to largely consist of games where teams have a less drastic week over week change in quality than OSU does this week.
Put another way, the 0.83 weight coefficient (1/(1.0+2)) that is put on OSU's previous week market implied rating is almost certainly too high, given that coefficient was found, by and large, by optimizing the normal cases.
"Lol public" as a conclusion seems pretty dubious.
Thanks a lot for the solid explanation. I should have done some digging on my own before posting. Much appreciated.
I am less and less convinced an 8 team playoff would be a good thing. Do we really need or want to see Michigan State or Arizona or the likes play again this year?
I am less and less convinced an 8 team playoff would be a good thing. Do we really need or want to see Michigan State or Arizona or the likes play again this year?