(12-06-2016 01:54 PM)MplsBison Wrote: (12-06-2016 12:37 PM)quo vadis Wrote: But nobody has claimed that possibility because there's no evidence for that.
It's self-evident. All algorithms must weight the input parameters. The weightings are arbitrary. There is no "correct" set of weightings.
You exactly exposed that this is correct in your post: when the programmer first starts building the classifier, it will spit out a nonsensical answer at the beginning. It might say the Sun Belt is the #1 conference and the SEC is the last conference.
So the weights have to be tuned, in order to get a sensible answer. Ah ... but when does the output stop being nonsensical and start being sensible?? Arbitrary.
Just as I'm certain you'll clip this part of my post out of any quoted response, I'm certain that you're fighting this because you yourself are the author of an algorithm to rank conferences. Your algorithm is biased, all the same.
How difficult is this to grasp:
Like any system, including the various formulas, mine is "biased" in favor of some performance results (e.g., winning games against highly ranked teams) and against other performance results (e.g., losing games against FCS teams). Because In My Opinion, a strong conference will have teams that win games against good teams from other conferences, and will not lose games to weak teams from other conferences.
But it is not biased against any conference. The formula is applied uniformly to all conferences for the entirety of the season, there is no line of code that gives bonus points to the SEC merely
because it is the SEC, etc., nor were any variables added or weights assigned with an eye to possibly helping one conference over another.
It's like If I was judging two contestants in a beautiful body pageant, and my attribute criteria for what is a beautiful female body includes large firm breasts, long legs, a thin tight waist, and little excess body weight, and two of the contestants were Angie Jolie and Rosie O'Donnel. Angie would get my vote, but only because she has the attributes i believe define 'body beauty', not because she is "Angie Jolie". If the two women switched names, then "Rosie O'Donnel" would be the winner. In that case, at the personal level, my criteria is unbiased. It is biased only at the attribute level, like my college football system.