NCAA Division I men's ice hockey tournament

Tournament format history

1948–1976

4 teams (one game series)

1977–1980

5–6 teams (one game series)

1981–1987

8 teams (two games, total goals first round at higher seed)

1988

12 teams (two games, total goals first two rounds at higher seed)

1989–1991

12 teams (best of three games first two rounds series at higher seed)

1992–2002

12 teams (divided into two regionals, East Regional and West Regional; six teams each)

2003–present

16 teams (divided into four regionals; four teams each)

At-large bids

Up until 1976, the NCAA tournament invited two teams from each of the two major regions: East and West. Initially, all teams were invited based upon their regular season performance with the NCAA selection committee occasionally deferring to an in-season or unofficial tournament to make their selection easier. Over time, as each of the two regions became dominated by single conferences, the selection committee would just choose the top two teams from each of the two leagues or, when held, the champion(s) and runners-up of the conference tournaments. During the first three decades of the national tournament other conferences did exist, however, most of these were either unofficial or contained teams that were largely regarded as inferior (MIAC for example) and were ignored by the selection committee. Things began to change in the 70s when several new programs joined the top level of play. With the new clubs being unable to join the existing programs, a new conference, the CCHA was formed.

For the first half of the decade, the NCAA tournament continued its recent pattern of only inviting two teams from both ECAC Hockey and the WCHA. However, after several years of petitioning by the new league, the selection committee finally changed the tournament format for the 1976 series. While it could continue to invite two teams from the more established leagues, the committee gave itself the ability to invite up to 4 additional teams to the tournament as it saw fit. Under this policy, the CCHA tournament champion was given a de facto automatic bid but the committee was reticent to include any additional teams in the field. Over a 4-year span, only once did the committee invite any other team, doing so in 1978 when the #1 ranked team in the nation, Boston University, lost in their conference semifinals and would not have been included in the tournament otherwise.

After failing to use most of the additional slots made available by the rule chance, the NCAA tournament was expanded into three full rounds in 1981. At the time, with just one large conference comprising all eastern teams, the ECAC subdivided itself into three regions. The conference's tournament champion as well as the two division champions from the other groups would receive automatic bids. For the west, the WCHA would continue to name two tournament co-champions who would each receive an automatic bid while the single CCHA champion would also receive an automatic bid. One at-large bid for each of the two regions was then available for the best remaining teams.

Within two years this cumbersome policy was abandoned and the NCAA tournament would only offer a single automatic bids to each of the three conferences with the rest of the field being made up of at-large bids. This policy had remained in effect since 1983 and has survived the addition of several other conferences thanks in part to the tournament expanding from eight to twelve to sixteen programs as of 2025.

Beginning in 1981, when at-large bids were first officially introduced, the selection of teams that were offered bids was based upon their national rankings in polls. These were primarily done by some combination of head coaches and sports writers and tended to favor more established programs. In the early 1990s, the selection committee began to try and compare teams objectively by instituting a new ranking system.^[8] Some of the initial rankings that resulted were decried as some of the tournament invitees possessed arguably subpar records and their inclusion pushed out teams that were widely regarded as superior, such as Brown in 1993. The system was constantly adjusted and modified and occasionally ignored until it became sole selection criteria around 2008.

Since then, at-large bids were offered to teams based upon their PairWise ranking which provided a single number for each program based upon several categories. The categories were altered and changed over time with each receiving different weights or priorities, however, by 2014 the system was largely seen by the committee as sufficient and went unchanged for several years afterwards. In 2024, the NCAA committee announced that it was abandoning the PairWise rankings for the Division III tournament in favor of the NCAA Power Index (NPI), the same system used for college basketball.^[9] A year later, the committee changed the selection process for the Division I tournament, fully adopting the NPI for all levels.^[10]