re: BCS works __% of the time? 27%

One of the problems I have with these SoS rankings is the way they come about. Teams are normally losely linked to every other team by 3-4 games ( I actually wrote a graph application to plot it out ). Football shows us time and again that just because Team A beat Team B and Team B beats Team C, it does not mean that Team A will beat Team C. Its a flawed system, but alas, what can you do?

The other problem I have with the SoS components is when you get to a team that is a 'significantly' lower ranked team than another. For instanct, LSU should walk all over Washington to start the season, but is Washington better than say, Syracuse? Honestly, do the differences between those teams matter? In most SoS formuals, every team is ranked and things could theorhetically come down to the 'difference' in playing a Washington or a Syracuse team. I honestly would like to see a team's opponents broken down into teirs: elite, good, average, poor, very poor. Whatever. Auburn may not have gotten screwed in 2004 if it was done this way.