For the fifth straight season, the NCAA men’s ice hockey committee has altered the selection criteria for the national tournament.
This season, the list of teams under consideration — those that are compared against each other to form the basis of the PairWise Rankings — has been expanded to all teams with a Ratings Percentage Index score of .500 or better, NCAA associate director of championships Kristin Fasbender said.
In the previous four seasons, the teams under consideration were limited to the top 25 teams in the RPI.
As of Wednesday, the change adds nine schools to the PairWise Rankings, which mimic the selection process used by the committee to select the at-large teams for the NCAA tournament.
Also, recent NCAA championship manuals show that the committee last year quietly removed the line that required both teams to have played at least 10 games against teams under consideration for that comparison to be factored into the overall comparison between schools.
Previous changes to the criteria included a requirement that teams have a winning percentage of .500 or better for at-large consideration; the addition and then removal of a bonus points system for quality non-conference victories; and alterations to the formula for calculating the RPI.
It’s believed that this year’s tweak was made at least in part to remove what could be considered an arbitrary limit of teams under consideration.
The PairWise Rankings replicate the committee’s process in comparing all of the teams under consideration against each other based on four criteria: RPI, record against common opponents, head-to-head competition and record against other teams under consideration.
For each comparison won, a team receives one point, except for the head-to-head comparison, for which each team receives one point for each game victory in the teams’ series. The final order reflected in the PairWise Rankings is based on the number of comparisons won. If two teams win the same number of comparisons, the team with the higher RPI is ranked higher, breaking the tie.
This year’s change makes the number of teams under consideration dependent on performance instead of a fixed figure.
Meanwhile, the women’s ice hockey committee has gone the opposite direction as the men’s committee.
According to the 2011 championship handbook, one of the criteria is record against teams in the top 12 of the RPI; that comparison formerly was made against teams with a RPI of .500 or better.
As of Wednesday, that change eliminated six teams from the PairWise Rankings.
Changes in detail
Here are the recent changes the men’s committee has made to the selection criteria, starting in 2003:
2003: Changed the RPI equation to 25-50-25 from 35-50-15 (all RPI equations listed as winning percentage-opponents’ winning percentage-opponents’ opponents’ winning percentage); added undefined bonus points to a team’s RPI for winning home, road and neutral-site non-conference games against teams in the top 15 of the RPI; adjusted a team’s RPI to prevent losing points for winning conference postseason games against weaker opponents; defined the pool of teams under consideration as those with an RPI of .500 or better.
2007: Changed the RPI equation to 25-21-54; changed the bonus points system to count only for road non-conference wins against teams in the top 15 of the RPI; added all games to the RPI adjustment for winning games against weaker opponents; redefined the pool of teams under consideration as those in the top 25 of the RPI; added a condition that the comparison of record against teams under consideration would be used only if both teams have played 10 or more games against such teams.
2008: Eliminated the bonus points.
2009: Added a requirement that to be eligible for at-large selection, a team must have a winning percentage of .500 or greater.
2010: Eliminated the condition that the comparison of record against teams under consideration would be used only if both teams have played 10 or more games against such teams.
2011: Redefined the pool of teams under consideration as those at .500 or better in the RPI.
