National Hockey League teams have not always publicized their attendance data. Team-by-team attendance data is not even available in the current edition of the NHL guide - the league considers this data to be under a team's purview and many teams just aren't interested in publishing this information.
However, the data is indeed historically interesting - we all know that the California Seals did not do well at the box office, but is anyone aware that they averaged just 4,960 fans for their inaugural season? Does anyone know that the Chicago Blackhawks averaged in the low 6,000's in the mid-1950's, at a time when the Toronto Maple Leafs were averaging twice as many fans, near 13,000?
Graphing the data makes easier to visualize, especially when it is assembled in two ways;
This data is not yet complete; it will be filled in as more research is finished.
Data published on the ESPN website, presumably compiled from electronic game sheet data that ESPN has access to. This encompasses data from 1993-94 to present.
NHL attendance data compiled by David Stewart-Candy and published on the Hockey Research Association website (with a tip of the hat to Marc Foster, who put the archive together). David's data is mostly from the post-expansion era, with some pre-expansion data found in NHL media guides. David lists these as his sources:
Team media guides for the post-expansion era.
NHL guides for pre-expansion data.
The Hockey News, to fill in gaps.
Hockey Summary Project (HSP) data. The HSP is a group of people who are researching and digitizing NHL game summaries. As part of this task, they have compiled NHL attendance data for many of the games they have entered.
Research compiled from various newspapers (NY Times, Toronto Star, Christian Science Monitor, Chicago Tribune, Boston Globe, etc.) by Carl T. Young Jr., who graciously sent me copies of his research
Additional research by me (Ralph Slate), using newspaper reports of games. This is done to fill in the HSP data, since often they did not compile the attendance for every single game in a season.
Since most of this data was not compiled by me personally, I can't guarantee its perfection. I have spot-checked many of the HSP game attendance figures, and they have generally been pretty accurate, but I have found occasional mistakes (usually transposed digits), so keep that in mind.
Data released by NHL teams does not always match the totals that are derived from the published attendance in game summaries. In most cases, they are close, but since they were compiled by two separate organizations, they can be quite different - in some cases the sources differ by several thousand. In this situation, the data published by the team is considered to be more authoritative than the data from boxscores, and the data released by the league is considered more authoritative than the data released by a team.
From the 1920s to the 1960s, the National Hockey League sporadically released attendance via newspaper
articles. The totals often do not match either the team totals or the totals derived from the games. In this
case I usually used the NHL figures.
Occasionally, I simply could not find the attendance for every single game that a team played in a season. Sometimes, there were just one or two games which eluded me. However, I think that it makes more sense to publish a good estimate, based on 90% completion rate, than to report N/A. In this case, I have noted that the data is estimated with an asterisk.
While going through old newspapers, it appears to me as though the numbers are not particularly precise. For example, the Boston Globe frequently reported "13,000" in the 1940's and 1950's, and also reported "13,900" even though the capacity was 13,909. Please take them with a grain of salt