2023 Pairwise Discussion

[Numbers]

November 24. As of right now, there is some discprepancy in the way that this site and the other big college hockey site are calculating RPI. I wonder if Ed Trefzger might chime in here and help us out.

 

[moose97]

There was a change in calculation of overtime percentage this year. Guessing some sites didn't make the change while others did?

 

[manurespreader]

I's like to hear why it's one way and not another. Seems like it was straightforward, mostly. And while I'm at it, does it really hurt you to win in OT at home vs a weak opponent?

 

[Numbers]

I's like to hear why it's one way and not another. Seems like it was straightforward, mostly. And while I'm at it, does it really hurt you to win in OT at home vs a weak opponent?

I think OT was changed from 55% win to 66% win for this year. In other words, a win in OT this year is closer to full value than it was last year. All conferences are doing 3v3 OT, so it was possible to make this change to the formula.

But I am still confused as to which site is doing what, as USCHO and CHN do not show the same pairwise (RPI is different). Ohio State and Michigan State right now have discrepancies of up to .005 in RPI, which is big.

 

[sezenack]

USCHO frequently is wrong about RPI and pairwise. CHN is definitely the right one

 

[Bronco Hockey Fan]

I miss when USCHO had the pairwise grids for every single team, with Greenish, Yellow and Red-Orange coded blocks to distinguish comparison points won and lost. Now that link just says nothing's available.

 

[Sol Diablo]

USCHO frequently is wrong about RPI and pairwise. CHN is definitely the right one

Dang! All this time I thought it was the site that had my team ranked the best!