Changes To NCAA Tournament Criteria

The Men’s Ice Hockey Committee has once again made changes to the selection criteria for the NCAA tournament. That includes a big shift in the Ratings Percentage Index (RPI), arguably the most important component of the selection process, which summarizes in the form of the PairWise Rankings. Changes — and analysis — follow.

Automatic Qualification Recommendation: That the following receive automatic qualification for the 2007 NCAA Division I Men’s Ice Hockey Championship: Atlantic Hockey Association, Central Collegiate Hockey Association, College Hockey America, ECAC Hockey League, Hockey East Association and the Western Collegiate Hockey Association.

Analysis: College Hockey America retains its autobid despite dropping down to five teams with Air Force’s departure. This has been expected for months, ever since the decision to permit conferences with fewer than six teams to maintain an autobid under limited conditions. The decision keeps the CHA viable.

Rating Percentage Index (RPI) Recommendation: That the Rating Percentage Index (RPI) for men’s ice hockey be modified as follows: (a) won-lost record remains 25 percent; (b) opponents’ winning percentage from 50 to 21 percent; and (c) and opponents’ opponents’ winning percentage from 25 to 54 percent.

Analysis: The published rationale is that “[o]ver the last couple of years, the committee observed that the current RPI discourages teams from playing certain opponents because playing and winning against these opponents would significantly lower their RPI. … These negative games (games that are won, but causes [sic] the RPI to decrease) may cause teams and conferences to reduce the number of games played outside of their leagues or to severely limit the ability of some teams to get quality non-conference games on their schedule.”

This change may only substitute one problem for another. Under the old system, teams which tended to hurt opponents’ RPI did so because they themselves had poor records, whether against good competition (Northeastern, Michigan Tech) or not (Robert Morris, American International). Now, with opponents’ opponents’ winning percentage taking up the lion’s share of the RPI, teams from College Hockey America and Atlantic Hockey are much more likely to boost opponents’ RPIs even if they themselves have records well below .500.

The decision went on to say, “[T]he committee believes that the new formula can reduce the number of negative-impact games, while at the same time not change the order of teams in the RPI. The modification simply reduces the number of negative-impact games and rewards teams for competing. Had the recommended RPI been used in 2005-06, the number of negative games would have been reduced from 117 to four.”

Selection Criteria Recommendation: That the following modifications be made to the selection criteria:

(1) That when reviewing the category of “teams under consideration,” automatic qualifiers are not automatically included in this category. Further, “teams under consideration” will be redefined as those teams that finish in the top 25 of the final RPI.

Analysis: This one is aimed squarely at the situation which arose during last season’s Atlantic Hockey playoffs, in which a Bentley victory over Holy Cross in the conference championship would have wreaked havoc on the RPI due to the fact that Bentley would have automatically become a Team Under Consideration, thereby adding its games to opponents’ TUC results. With the right outcomes in other games, the result would have been Holy Cross making the NCAA tournament as an at-large bid, which the committee obviously did not take well.

(2) That the category of “teams under consideration” will be used only if the two teams being compared have played a minimum of ten games each versus “teams under consideration.”

Analysis: This is a tweak aimed at schools with weaker schedules, which could hypothetically play very few Teams Under Consideration but manage to win a couple of those games and end up with, say, a 2-1-0 record for a .667 TUC win percentage. Since the TUC criterion uses only percentages, not total wins, the old system could have produced nonsensical results based on a handful of games. That will no longer be the case.

(3) That any win which results in a decrease in a team’s RPI will not be included. Currently, this policy applies to conference postseason tournaments. This would extend the application to the regular season as well.

Analysis: This one seems to have the least impact, given the drastic change to the RPI weighting formula above. If the point of changing the RPI weights to favor opponents’ opponents record was to eliminate the “negative win,” why also add this change? Regardless, it should have minimal effect since so few games will produce the “negative win” outcome under the new RPI weighting.

(4) That the bonus points for the RPI will be redefined as follows:

(a) A “home” game is defined as one that takes place in the facility that the institution uses primarily for “home” games.

(b) A “neutral” game is defined as one that does not take place in an institution’s primary facility.

(c) Bonus points are awarded only for wins on the road (in the opponent’s primary facility) and will be used for interconference games only.

Analysis: The “bonus point” system has never served a clear purpose, despite the notion that it encourages teams to play top-tier competition outside their own conferences. Top-caliber teams schedule one another for plenty of reasons apart from a paltry boost in the RPI, and the bonus system has unnecessarily put extra weight on a handful of nonconference games, many of them in tournaments where the teams in question cannot control who their opponents will be.