I'm interrupting a series of posts on win thresholds to continue a discussion we had in one of the comment threads about goalie equipment. A lot of people make a big deal about the size of goalie equipment in the NHL, and the NHL is currently working on yet another attempt to "fix" this problem. Many see large goalie pads as the main reason for the low scoring rate, and feel that if more restrictions were put on equipment then the league would become significantly more high-scoring. I've seen some fans go so far as to argue that the main difference between the 1980s scoring environment and the scoring environment today is that goalies wear much larger equipment.
I think those claims are completely unfounded. The reasons for the rapid improvement in goalie numbers over the last two decades have been improved goalie technique and better defensive play. Goalie equipment size has been a minor factor.
The main proof for this is the way that the goalie crop turned over in the late '80s and early '90s. The old guys were phased out of the game, a new wave took over, and there was a rapid increase in the league-average save percentage. All of this happened before goalies started increasing the size of their equipment.
Patrick Roy is generally credited with popularizing the butterfly style. Roy played his first NHL game on February 23, 1985. At that time goalies spent most of their time on their feet, they relied on their limbs rather than their body to make saves, and they made skate saves and stacked the pads and hugged the post with their arms and did all the other traditional things goalies had been doing for years. Let's look at what happened to that group of netminders.
There were 66 goalies who played a game in that same 1984-85 season, including Roy. Their average age was 25.7 years old. Just 25 of them (38%, a little over one-third) were still in the NHL in the 1989-90 season, a mere five years later. Only 8 of those 25 were still in the NHL a decade later in the 1994-95 season. Only 2 of of those 8 performed at a high level compared to league average in the 1990s, Roy and John Vanbiesbrouck. Roy was the only one of them that won a Vezina Trophy in the 1990s, and he won his last award in 1992.
Let's fast forward a decade to that shortened 1994-95 season, and do the same analysis for that group of goalies. Remember, this is still before the equipment got huge. Here's a picture of Patrick Roy in action against the Quebec Nordiques during the 1993-94 season, for example. He has hardly any thigh-rise on his pads and his pants, chest protector and jersey all fit fairly snugly.
In 1994-95, 68 goalies played in the NHL, with an average age of 26.3. Jump ahead 5 years to 1999-00, as we did for the prior group, and we find that 44 of them are still in the league (65%, or about two-thirds). Not only that, but a dozen of them were still playing 11 years later when the league came back from the lockout for the 2005-06 season. Six out of the 9 Vezinas awarded from 2000-2009 went to goalies who were active in 1994-95, and that group included many of the top goalies of the 2000s.
Having established that the goalies from 1994-95 had much more longevity than the goalies from 1984-85, let's look at the average save percentages for those two seasons. I'll also throw in the average for 2003-04, the season with both the lowest average goals per game and the highest average save percentage of the so-called Dead Puck Era.
Power plays per game were at about the same level for all three seasons, 7.9 per game in 1984-85, 8.6 in 1994-95, and 8.3 in 2003-04. Some of the increase in save percentage would have been from defensive play, but most of it was because the goalies were better. League-wide defensive play also improved from 1994-95 to 2003-04, yet the league average save percentage only went up by .010, despite goalie equipment getting much larger. It is obvious that the more significant change happened between 1985 and 1995, not between 1995 and 2004.
We can also get a sense of the changing dynamic from 1985 to 1995 by comparing the save percentages of the goalies who were active in both periods. I chose to look at three year averages to avoid small sample sizes, including the year before and the year after for each goalie (i.e. 1984-86 and 1994-1996). There were 8 goalies in the group, but I dropped Roy since he only played 1 game in 1984-85 and therefore would really only have one season count. Roy played the modern technique anyway so his progress isn't really meaningful to what we want to track, which is how the older goalies adapted their games to a changing league.
There is a substantial improvement in the numbers. It is important to note, though, that this improvement was mostly being driven by the younger goalies in the group. The goalies who saw their numbers jump the most were Ken Wregget (.871 to .900), Tom Barrasso (.886 to .898), and especially John Vanbiesbrouck (.882 to .915), all of whom were 21 years old or younger in 1984-85. Not only were these goalies young and still not at their prime in the mid-'80s, but they already used some modern techniques or were able to adapt to the changing game.
Let's look at the older group of goalies, which includes Andy Moog, Grant Fuhr, Don Beaupre, and Kelly Hrudey. These goalies did not change their styles as much over the same period:
Despite seeing their save percentage numbers rise as a result of better defensive play in front of them, this group of four lost a ton of ground to the rest of the league during this period. They went from +.015 compared to league average in the mid-'80s to .002 below league average in the mid-'90s.
Let's compare that to the 1995 group. We want to look at the goalies who were still playing a decade later, so I picked out the 12 that played in the post-lockout NHL. I decided to compare their 1994-96 results to their numbers from 2002-04, since we want to look at seasons with no equipment restrictions.
That's just a slight increase. To be fair, we should remove Dominik Hasek, who is older than the rest of the group and is skewing the numbers with a mostly age-related decline from .926 to .914. Without the Dominator the group goes from .904 to .911, an increase of +.007. In a more defensive league with larger equipment, the increase in numbers is exactly the same as the increase we saw from the standup goalies from 1985-1995. In both cases, the league improved defensively over the period. Larger equipment would have had some small effect, but certainly not the game-changing impact that some would have you believe.
Finally, recent years have also showed us that there is not much of a relationship between equipment size and goalie play. After the lockout in 2005 there were new restrictions placed on goalie leg pads (reduced from 12" to 11") as well as glove and blocker sizes. Yet today the average save percentage in all game situations is back up to the same levels as it was in the early 2000s, despite the removal of most of the clutching and grabbing.
Just to recap, the league average save percentage went up by .037 from 1985 to 2004. Most of that improvement (.027) was already made by 1995, which was before the league-wide increase in goalie equipment size. The rest of the improvement coincided with an increasingly defensive league. When we look at the standup goalies who played in both 1985 and 1995, and compare their performance to goalies who played in both 1995 and 2006, we find that both groups increased their numbers by a similar amount even though only the latter group benefitted from huge equipment.
If I had to estimate and rank the factors that led to the change in league save percentage between 1985 and today, I would rank them in this order:
1. Improved goaltending technique (~.015-.020)
2. Improved defensive play (~.010-.015)
3. Goalie equipment size (~.005-.010)
The post-lockout crackdown on goalie equipment was still probably a good idea, but the continued focus on equipment size is in my view excessive and not something that will yield significant rewards. If the league wants to increase scoring, there are better alternatives to pursue.