A (recent) history of goaltending
... and today's one-shot difference
The following was put together after a few days of research, and the results may have to be digested in chunks given this is a really, really long post. Perhaps a new record. For this, James apologizes.
It's fair to say that this introduction is a trend familiar to most hockey fans, but bear with me on a quick walk down memory lane.
Rewind the NHL back to the 1950s, hockey's post-World War II era, and you would find goal scoring at pretty much the same rate it is right now — the Dead Puck Era's forefather. From 1950 to 1970, teams scored somewhere between 2.5 and 3.0 goals per game, and that generation's goaltending greats posted all kinds of gaudy numbers (Terry Sawchuck's 103 career shutouts come to mind).
Beginning in 1971, however, as the league ballooned, more pucks started going in, and by 1979, the end of the WHA era, we were at 3.5 goals per team per game. Along came the Edmonton Oilers, who ushered in hockey's highest-scoring era, which saw goals hit the 4.0 per game mark from 1982 to 1986.
Eight years later, in 1994, we were back at 3.25 and sliding, so much so that the 1998 season fell into the same 2.65 range we've seen for just about the past decade. I'd bet good money that that's where things will sit this coming season, and likely for the near future.
That's all really just a very wordy way of saying what stats man Gabe Desjardins has put together here, a nice graphical compilation of the rise and fall of goals in hockey.
Now, at a very basic level, there have always been two explanations for what happened, for where those goals "went":
- Teams started playing more defensive hockey, the result of talent dilution and better coaching systems
- The goaltending got better
Wouldn't that then mean the majority of the goal's disappearing act could be blamed on the 'tenders? Or is it more complicated than that?
Oh, the things that keep you up at night...
The pertinent year with regards to the rest of this post is 1982-83.
The Boston Bruins were a powerhouse, leading the league with 110 points, and Wayne Gretzky was his high-scoring self, notching 71 goals and 196 points in 80 games with the Oilers. Steve Larmer was the rookie of the year with Chicago, Pete Peeters won the Vezina with Boston, and Edmonton was swept by the Islanders in the Stanley Cup final.
1982-83 was also the year detailed shots on goal statistics were kept, so that's where we begin.
The shrinking goals against average
Statistics do not include empty-net situations
No surprises there, really. Goal scoring began to bottom out just before the last lockout, and we saw enormous year-over-year declines from 1992-95, which included the rise of Dominic Hasek and the New Jersey Devils, among other defensive influences.
It should also be noted the rise in number of total games played coincided with these declines as well.
But what about save percentages? Were goalies simply seeing fewer shots and, therefore, allowing fewer goals? And how did fewer goals against affect the number of shutouts posted per 60 minutes of play?
The Shutout Era
These days, an .875 save percentage is almost unheard of. Dan Cloutier posted an .860 in limited action last season, which made him one of the worst netminders in the league, but 24 years ago, it wouldn't have been all that awful.
Predictably, save percentages have followed an inverse path to goals against average, with the average figure cresting over the .900 mark for the first time during the lockout-shortened campaign.
On a related note, and as the next chart will attest, the past 12 years have also been the era of the shutout, as goose eggs are far, far more likely than in the mid-80s. In 1986-87, for example, you would see a shutout every 61 games, on average — and there were only 28 in total that year.
More recently, it's been down to as common as one every 13 games (and people like Brian Boucher have put together strings of five in a row). Last season, there were 148.
Few surprises, perhaps, so far. But what I really wanted to know from all of this was if netminders were facing fewer shots these days when compared to the goal-scoring heyday, and if so, how many fewer? Was a considerable amount of the decline in goals attributable to a lack of shots?
And, to follow that logic, would we increase goal scoring in the NHL by somehow getting more pucks on goal?
Saves and shots against
(per goalie per 60 minutes)
From the very, very peak of NHL goal scoring, 1985-86, the year Gretzky had 215 points, to last season, goaltenders faced an average of 1.5 fewer shots against per game and made 0.4 fewer saves.
Not much of a difference, really.
Looking at the past 24 NHL seasons, the highest shots on goal against ever reached was 30.72 in 1985-86, and the lowest was 26.97 in 1997-98.
Here's the swing represented in a graph:
Are goaltenders facing fewer shots? Yes, quite clearly. But what we don't know, at least from this data, is if that's a result of better netminding (fewer rebounds, etc.) or more defensive hockey. Anecdotally, it's certainly true that coaching systems and the like have improved immeasurably in the past 25 years, and it would appear 1997-98 was the pinnacle of clutch-and-grab.
But in the past two seasons, with more power plays and a push by the league to create more offence, we've seen shots on goal levels return to within one per game of the absolute peak of shot production in the NHL. The average goals against average from the heyday to the postlockout period, however, is nearly a full goal per game lower.
Goaltenders now are making relatively the same number of saves they did from 1982 to 1993, around 26.5 per game. The past two seasons, they've simply faced one fewer shot per game — and that swing alone is enough to drop save percentages by 3 per cent.
And in the periods where shots on goal fell lower, to below 28 against per 60 minutes of play, goal scoring dropped dramatically.
Now, it's no surprise that shots on goal and goals scored are related, as more offensive hockey would certainly produce both. The year shots on goal dropped dramatically (1997-98), shutouts rose considerably and Hasek won his second consecutive Hart Trophy.
Shooting percentages have fallen from a high of 12.5 per cent to a low of close to 9 per cent over the past 25 years, but what's really amazing is that, throughout an era that has seen a dramatic change in goals scored, netminders have continuously saved almost the exact same amount of rubber.
A 26-save performance was as average in Gretzky's 215-point season as it was last season.
But does it all come down to a one-shot difference? And if goaltenders were to, somehow, face an all-time high number of shots per game, let's say 31-plus, would that 'solve' the league's goal-scoring problem?