Reply to thread

Message: <blockquote data-quote="drothgery" data-source="post: 1814811" data-attributes="member: 360">It may seem like I'm being inconsistent here (as just a few posts ago I noted that the BCS formula hasn't worked as well as the human pollsters over the last few years), but the computers do okay. They haven't done as well as the humans, at least at the top of the polls, but a lot of the biases that coaches and writers factor into their rankings tend to have a lot of foundation in reality (major conference schools are better; winning on the road means more; non-confernce games say more about strength of schedule than conference games; winning impressively is better than winning ugly, but winning by 70 is not more impressive than winning by 35). I think that given time, enough data, and no artificial limitations on what the computers can factor in (margin of victory can't be used by the BCS computers), computer rankings will eventually at least be better than the pollsters, even if they're not now. Having said that, even the result on the field isn't indicative sometimes. For a lot of reasons (most notably injuries, failure to be emotional ready for the game, freak weather conditions, or one coach just having another's number) an inferior team can win a game. And when one team isn't better than the other by much in terms of talent and gameplanning (which is usually the case in the NFL), the luck factor is often enough for the worse team to win.</blockquote>