(PWR) PairWise Rankings - 2009

[sagard]

Some have been waiting to see what would happen due to this scenario. When a storied program such as WI beats UND to end the season, and plays a boatload of TUCs and misses the tournament due to the math track team's decision making. I haven't done the math yet for Minnesota, but they had to be right there too with the old RPI.

I think there will be some "discussions" about this in the summer meetings.

I agree there should be discussions. RPI has it's flaws, but tweeking the formula to reduce negative impact wins is dumb. UW and SLU paid the price. If you can't live with negative impact wins during playoffs, discard the result as in years past.

The Gophers are out from all draws unless KRACH was used. Then UM, UW, BC and Lowell are in. Miami, OSU, Cornell and Princeton are out.

I do think the RPI formula needs to be changed back. You don't measure SOS using primarily your Opponents' Opponents Record, you use your opponents record.

I said it at the time, but the WCHA coaches really dropped the ball allowing that RPI formula change to occur. They simply gave away a major competitive advantage. This year UW paid the price. Next year could be anyone, but the WCHA is likely to pay the price.

[nodakvindy]

The biggest problem is that for some reason people think it should be impossible for the RPI to drop when you win a game. That's not an invalid result at all. The problem would be if somehow a win hurt your RPI and a loss helped it. But in the cases when a win still drops your RPI, a loss would have dropped it much more. The RPI, like the PWR, are supposed to be looked at only at the end of the season, when you can view things in totality.

I think a move back to the old RPI is appropriate, but I also think both the TUC and COP should be the RPIs for those games rather than the actual record, with the TUC limit dropped to 8 and a COP limit of at least 2 opponents added.