x

x

x

x

x

x

x

x

x

rails - interesting analysis. i agree with you that there is nothing collusive about agreeing to all go 10-0. i have tried to get one of my conferences to do this ages ago. for the mathematically challenged, the logic is really very simple. when you play a game, as rails pointed out, rpi is 25% who wins, 50% is your opponents w/l, and then 25% is a more complex opponents opponents win loss. essentially, who wins or loses can vary. but when you play the game, 75% of the contribution to rpi is already set in stone. and two thirds of that is your opponents record.

many coaches incorrectly believe scheduling opponents by rpi is the way to go. but, only 25% of your opponents is their win loss, which is two thirds of that same opponents contribution to your rpi. so rpi is really a poor measure of how an opponent will impact your rpi. you are much better off to schedule by win loss. this is precisely why the optimal schedule for a less than exceptional team is to play top teams in the bottom conferences. you end up playing relatively easy teams (doesn't have to be the best team, maybe 2nd or 3rd is fine), who make large contributions to your rpi.

from a conference standpoint, the conference doesn't care who wins or loses within conference. its largely (not entirely but close) a wash. so, you really care about that other 75% - of which, two thirds is win loss. it is technically possible to put together a schedule better than 12 10-0s against easy opponents. but it is very, very, very difficult and i doubt it happens more than once every ten thousand conference seasons.

on a final note, much of that last 25% of your rpi - opponents opponents record - comes from the WL of your conference. you play 16 games against conference teams (more, because of CT, but lets keep this simple). they play 16 games against conference teams. so the majority of your opponents have  your conference mates as the majority of their opponents. the WL of the teams in conference is fixed at .500 for in conference games, so putting up a high WL out of conference is the way to take advantage of the situation.

so, from just about any angle, the best thing your conference can do if you want to be super strong is to schedule easy, and win 9 or 10 games every season - even if that means on a down year, you play 10 simmies out of conference.

Gillispie Says in the first paragraph:  75% of the contribution to rpi is already set in stone. and two thirds of that is your opponents record.

Indeed and add to that.  65% of the two-thirds (42% of your RPI) is how your conference mates fare in OOC.  That is because conference play is zero sum.  That is something that one can influence.  Add the 42% to the 25% of the formula for W/L%, that's a whopping 67% that one can either directly control or influence through the conference mate's OOC schedule.  The schedule of the 10 OOC opponenets or the OWP or OOWP of your conference mate's OOC schedule is insignificant.

You and I are on the same page.  But it is much better for a team to play a 10-0 team that has an RPI of 100 than it is to play a 5-5 team with an RPI of 80.   

Interesting stuff, rails. I'm pretty sure your hypothesis will end up being proven. Even in other seasons when the non-con performance hasn't been this spotless (well, almost spotless ... coughkylecough), I've watched as our SOS has gotten better and rpi lead widened. There also wasn't any kind of collaborative effort to schedule like this. I never did before getting to the ACC, but the conference is so brutal, I felt like it was the only strategy that made sense. We've had several times where a team that finished 4th or 5th in their own division went on to win it all.

Hopefully I'll remember to check back ...

man i have been preaching this for ages and have been laughed out of conferences more times than i can remember. i love being proven right. thanks rails

Posted by joehuskie84 on 12/10/2011 3:31:00 PM (view original):
man i have been preaching this for ages and have been laughed out of conferences more times than i can remember. i love being proven right. thanks rails

Thanks Joe.