tag:blogger.com,1999:blog-6331918571067993675.post3775810989529339809..comments2017-05-14T06:53:45.135-07:00Comments on Playoff Guru: The Battle for No. 4Samuel Chihttp://www.blogger.com/profile/07694013336972776957noreply@blogger.comBlogger2125tag:blogger.com,1999:blog-6331918571067993675.post-38639684824019000102008-10-22T10:29:00.000-07:002008-10-22T10:29:00.000-07:00Good point, anonymous. I agree.Good point, anonymous. I agree.Utenoreply@blogger.comtag:blogger.com,1999:blog-6331918571067993675.post-68003344438719168802008-10-20T12:23:00.000-07:002008-10-20T12:23:00.000-07:00Ok, perhaps others have already commented on this,...Ok, perhaps others have already commented on this, BUT ....<BR/><BR/>The guys who make up the BCS formula are kind of dumb!!<BR/><BR/>OK, several years ago, it was pointed out many times how wrong it is was to use the "ordinal" numbers for the Human polls. The human polls certainly don't have "equally spaced" teams, and it certainly possible, (in fact probable), that #4 can be much closer to #3 than to #5, or even that #5 could be closer to #10 than to the #4.<BR/><BR/>SO, to the BCS's credit, they changed to use the "points" and to normalize the points and use percentage of possible points, which makes lots more sense.<BR/><BR/>GOOD JOB BCS - you recognized the mistake of using Ordinal numbers and fixed it in a very simple and straightforward manner ...<BR/><BR/>NOW, we add a few more computers into the mix, and need to average those, so what does the BCS formula do? USES THE ORDINAL NUMBERS AGAIN!! IF the ordinal numbers were a mistake in the human polls, they are a mistake in the computer polls as well. And, I would think that it's not hard to have a normalized points system for the computers either. The answer is right there staring them in the face!!<BR/><BR/>Why would the "fix" the ordinal problem in the human polls, and then repeat it in the computer polls?Anonymousnoreply@blogger.com