Re: New WCHA is dead
I'm partially in agreement about creation of the NCHC over Pairwise concerns but it's more of a result that a specific reason.
It was a financial resources decision. The schools that created the NCHC have a history of prioritizing hockey over other sports. And they followed that up with significant monetary support. When the B1G was announced, that left an even number of rich vs less rich schools. The rich schools were seeing they were losing the pull they were used to in the WCHA and they didn't want to be controlled by schools that weren't as interested in increased travel costs, bigger budgets, and more expenses.
Generally, yes you can say that those less rich schools are going to traditionally be lower in the pairwise. Year-to-year, that's usually the case, although there are obviously going to be exceptions each year as teams fluctuate and do well.
I agree with this. The above example about Bemidji moving from 12 to 17 (or whatever it is) by flipping one win to a loss......That's going to happen with almost any school in the 10-20 range. The problem isn't a loss against a lower team. The RPI calculation doesn't distinguish between these 2 things:
1) Win at home at #20 in the RPI, loss at home against #59 in the RPI or...
2) Loss at home against #20, win against #59.
The real issue there is that margins are soooo tight in that range of the RPI. Basically, wherever you are right now, if your RPI is between .52 and .535, and you win your next 2 games, your RPI will rise to about .53 or .545. So, there really isn't much conclusion which can be formed about that.
Now, in general, the WCHA teams right now will start the conference season lower in the PWR rankings, because of the usual sub-.500 non-con record. So, there is 'drag' on the top schools because the lower schools don't have as good of results. But, it's not the losses, per se, which pull the Mankato's and Bemidji's down. It's the schedule in the first place.
And, that is a discussion for the 'accounting guys', because the reality is that all the tiny tweeks in the RPI calc are more evidence that it's a poor system. As has been said on this site numerous times, the best system would be a modified KRACH (I call it the Bradley-Terry method), with the further parameter of a nationwide home-ice advantage. And, you can find that under KASA on the RPI (the school) website.
However, as stated above, the move by NCHC wasn't because of PWR. It was because of being the schools which emphasize hockey to the extent they do...