The NCAA made some big changes to all D3 tournaments. Theyve changed the ratio from 1:6.5 to 1:6 which means next years d3 hockey tournament will have 14 teams and is just a couple of teams away from a 15 team tournament. There will now also a single at large pool for bid, ditching the old Pool B and C bids. Lastly and arguably the biggest change is that the NCAA power index for the selection of at large berths for all team sports championships. Unless "NCAA Power Index" is another word for pairwise I think we might see the end of the pairwise rankings
I’ve been confused on that last point. Through the whole process of proposing a change to the selection process for all team sports (only Hockey was using PairWise, the rest were still using the old “committee evaluates the criteria” process), they referred to the proposal as “selection criteria database,” and only at the penultimate hurdle of passing the DIII Championships Committee did they refer to “NCAA Power Index,” so I’m unsure if they’re the same thing. Or did they plan to do their own thing and then at the last hurdle pivot to the NPI in use for D1 women’s hockey? Someone smarter than me will have to decipher that. Maybe Russell has a better grasp on this than I do?
If the “selection criteria database” being considered earlier and the “NPI” that was adopted as the selection process are the same thing, an explanation from a Championships Committee report earlier in the process seemed to describe something similar to the PairWise with a few differences:
-It is a calculation based on Winning Percentage and Strength of Schedule. Each individual sport Committee can set the “dial” on how those 2 are weighted (ex. 25-75, 33-67, 100-0, 0-100, etc) before this upcoming season, tweak it the following season, and then tweak it further at set intervals thereafter (i.e. not every year). The changes will always be made in the off-season, so teams know going in what they need to do (though I could imagine coaches saying it’s difficult deciding who to schedule for non-conference if you don’t know how big strength of schedule will be).
-If it’s the same NPI as used in DI Women’s Hockey, the strength of schedule metric is a bit different: whereas the RPI/PairWise used Opponents Winning Percentage and Opponents’ Opponents Winning Percentage, the NPI used in DI Women’s Hockey calculates strength of schedule by the average NPI rating of opponents. I wish we still had NUProf around to get into the weeds of how that works, since your opponents’ NPI affects your NPI, but your NPI affects your opponents’ NPI, so is this like the KRACH where you just repeatedly run the calculation using the previous values until eventually the values stop changing between rounds of calculation?
-In the initial description of the “selection database,” they also said the individual sport committees will have the freedom to tweak the “dials” on weighting winning percentage based on Home/Away and OT/Regulation (like the PairWise currently lessens the impact of OT results, Home wins, and Road losses, but increases the impact of Home losses and Road wins). No idea if that aspect of it carried forward.
-DI women’s hockey’s NPI applies a Quality Win Bonus (like in the RPI/PairWise) for any win or OT loss against a team with an NPI above 51.5%. The previous descriptions of the “selection criteria database” by the DIII Championships Committee made no reference to QWBs at all.
Bottom line: there are a few different ways this can go, but none of them stray TOO far from what the PairWise does. My biggest concern is whether the weights to the criteria decided by the individual sport committees will be published before the season so places like USCHO can provide us those numbers throughout the season like the currently do with the PairWise.
