Advanced Metrics: The top PDO dynasties of all-time

For those of you who aren't familiar with PDO, it is just Save Percentage plus Shooting Percentage [more precisely: PDO=1000*(G/SF+SV/SA)]. What's interesting about it is that it trends very heavily – at the individual level or the team level – to its mean of 1000. But it is possible to assemble a team that plays above 1000. For example, in recent times, Jacques Lemaire has put together defensively-minded teams that limit opponent shooting percentage. For several years, Dominik Hasek played so well that he limited opponent shooting percentage irrespective of the defensive system in front of him. Such examples are few and far between.
Here are the only teams since the 1967 expansion to break a 1020 PDO for four or more consecutive seasons:

I'm missing data for 1979-80 and 1980-81, so the Islanders are over a nine-year period, not 11 years. We essentially see three distinct groups in this data:

1) Post-Expansion

In the NHL's inaugural 12-team season, the "original" six teams played substantially more games among themselves than against the expansion six. The league changed to a balanced schedule in the second season (1968-69), which, combined with any number of failures on the part of the expansion teams, led to more than a half-decade of good teams beating up teams that barely deserved to be in the league. Here are the PDOs for the "original" six teams for the first nine post-expansion seasons:

In many ways, this was unavoidable. It is simply impossible to double the number of teams in a top professional league and maintain the level of play without finding a massive source of previously untapped talent. The NHL did not do that, and so eight California-Pittsburgh games a year were a surefire path to bankruptcy. It was better to have Montreal come in even if it guaranteed a loss. And it did – Montreal, Boston, Chicago and New York each had PDOs significantly above 1000 through 1973; none of them ever lost a playoff series to any of the other eight teams in that time period.

2) Dynasties

The expansion advantage may have faded for the established teams by the mid-1970s, but it didn’t create parity across the league. Over a 20-year period, the Philadelphia Flyers, Montreal Canadiens, New York Islanders and Edmonton Oilers successively hit six-to-nine year peaks where they were in the Stanley Cup semi-finals or finals every single year. These teams simply managed to horde talent in a way that’s impossible today and were so much better than their opponents that they really could achieve a higher-than-average shooting percentage or save percentage.

The Pittsburgh Penguins and Colorado Avalanche were very much the heirs of the 1970s and 1980s dynasties, as were the dominant New Jersey and Detroit teams of the mid-90s through early 2000s. But their PDO numbers were simply not as high as the great teams of the past. Talent levels in the modern NHL show nowhere near the variation that we saw prior to 1990 – in 1967, there were 12 teams, six of which were stocked with career minor-leaguers; in 1976, there were 32 professional teams stocked almost exclusively with Canadian players. Today, there are 75% more people in Canada, yet Canadians fill up approximately 15 NHL teams. The 1981-90 Edmonton Oilers went 98-33 in the playoffs; it is essentially impossible to be that much better than the other teams in the league today.

3) Great Goaltending

How often does a team lead the league in save percentage in five straight seasons? It’s possible that it could happen under certain defensive systems, but really, it takes a once-in-a-generation talent like Dominik Hasek to make it happen. The 1993-99 Buffalo Sabres generally had shooting percentages that were average or above-average, but their save percentages were always outliers. This type of team was more common when the league talent level was more diluted – today, there are simply too many good goaltenders for any one to consistently stand out above all of the others.