So let's broadly look at the numbers: (I'll admit I'm a closet analytic and gladly let Cheese, Iguana or someone smarter than I run the hard numbers)
- RPI is 25% W/L
- 50% OWP
- 25% OOWP
I am assuming 28 games prior to the selection of NT teams, not 26. Roughly 65% of a team's schedule (18 of those games) are conference games.
The prevailing thought for most is that how they schedule their out of conference (OOC) games is the most important factor in their overall sos by season's end. It is not. How your conference mates schedule their OOC games is the biggest factor in one's SOS!
And it's only the 10 OOC games that matter because after OOC play every conference is .500. When Duke's win is UNC's loss for example, Maryland will still count both Dukes and UNCs OOC wins as their OWP, but it wouldn't really matter after that because at the end of the season conference play is a zero sum game. I do acknowledge that if one division does much better than the other half, it will impact teams in those divisions differently.
12/11/2011 11:17 AM (edited)