Explaining a subtle change to the PairWise

If you follow the PairWise Rankings, you probably know that it’s based on the publicly published NCAA men’s ice hockey committee’s selection criteria.

That criteria, however, isn’t published until the second half of the season even though the committee goes over any potential changes in the offseason. So even though some of the details may emerge before then, it’s never really set in stone until we see that list of criteria.

This week, we at USCHO saw this season’s championship handbook posted on the NCAA’s website (you can find it on this page) and it caused us to tweak the coding that makes the PairWise Rankings.

We’ll get to that in a minute, but let’s start a few weeks ago, when we first learned of the change in the selection criteria that included any team with a Ratings Percentage Index of .500 or greater as a team under consideration, as opposed to the previous limit of the top 25 teams in the RPI.

At the same time, we noticed that last year’s championship handbook (PDF available here, external link) was missing some language that had been there in previous seasons.

Up to that point, the comparison of the teams’ record against teams under comparison counted only if both teams had played at least 10 of those games. That apparently changed last season, when no such restriction was in place, according to the handbook.

So at the same time that we changed the PairWise to include all teams with an RPI of .500 or better, we also removed the 10 games limitation.

Fast-forward to this week, when this season’s handbook appeared. Well, the 10-games language is back:

Results versus teams under consideration (“teams under consideration” defined as those teams with an RPI of .500 or better). This category is used only if the two teams being compared have played a minimum of ten games versus “teams under consideration.”

So we put the 10-game minimum back in our PairWise Rankings, and that should be the end of the tweaking. Let’s hope.

It’s tough to say what the changes will mean because, in reality, the selection system matters only after every regular season and conference postseason game is in the database. By then, many of the teams involved will have played the requisite 10 games.

But with the expansion of the pool of teams under consideration, it wouldn’t be a stretch to have a few comparisons impacted by that change.

Enough to alter which 11 teams earn at-large bids to the tournament? You never know.