Just wanted to post a quick link to an article by Bill James, a statistician, advocating that quantitative analysts effectively go on strike by refusing to allow the BCS to incorporate their computer rankings in its algorithm. Among other things, he points out that the BCS limits the computer rankings' inputs to such an absurd degree that it's impossible for them to make unique and meaningful contributions.
More interesting is James' assertion that participation in the BCS actually hurts the reputation of the field of quantitative analysis, since rightly or wrongly, when the BCS malfunctions, the knee-jerk reaction from practically everybody is to blame the computers. He says statisticians should refuse to continue playing whipping boy for the BCS. The BCS can hang itself with its own rope.
Wednesday, January 7, 2009
Subscribe to:
Post Comments (Atom)
Your computer rankings are just as bad based on you having tam over ark this year. You should fix it or stop posting these ranks on massey. Massey comp is used in the bcs so your part of the problem. I have kept up with ranks on massey for over 10 years and really wished he would drop bad polls and computer ranks, but he added instead. so even that shows you are part of the problem when your poll is this off.
ReplyDelete