Goals are fun. Goals are exciting. We all (except for goalies) want more goals. And this season in the NHL feels like we’re getting more goals. At least to me, a person who watches every Tampa Bay Lightning game.
But as I’ve learned from watching hockey, my eyes are filthy dirty fibbers. Every night I watch hockey. And every night I get an idea that turns out to be wrong when I check the numbers. So instead of trusting my idiot eyes, I decided to pull some data and see if goal scoring is actually increasing.
To start, I wanted to look at recent trends. Evolving Hockey has data going back to 2007-2008. The chart below shows goal scoring rates at both 5v5 and in all situations for all seasons in the data set.
Ok. So maybe the growth isn’t massive. But we do see a growth trend over the last three years after about five years of relative consistency. And we’re currently at the highest point in the last 12 years. With the season only being half over, the trend could change for 2019. But overall, what we’re seeing is in line with the way things were heading the last two seasons.
One of the issues with goals as a statistic is that they are particularly susceptible to fluctuations due to shooting percentage and save percentage, even in a sample as large as a full league season. So to dig a little deeper, the following chart is similar to the one above but shows expected goals. This will tell us if teams are generating better offense or just scoring more.
The trend here is more positive and more consistent than the goal trend. In the data set, teams have been getting better almost every year at generating more offense. That’s an encouraging sign for fans who want to see more scoring. It also suggests that the modest upward trend in goal scoring is a legitimate improvement in offensive play and not a fluke due to percentage fluctuations.
Looking at data since 07-08 is good enough to give a recent trend and gives us the chance to look at more robust measures like expected goals in better context by accounting for special teams. But to really understand where the league is headed in terms of goal scoring, we have to go back further. And that comes with some issues.
The best we can do when we look historically is to calculate the average amount of goals per game in the NHL each season. We also have issues with changing overtime rules and the silly shootout goal that teams get now.
Acknowledging those weaknesses, the chart below shows NHL goal scoring going back to 1950. The 1950 cutoff is somewhat arbitrary but I chose it because the data starts to get weird in the first half of the century due to significant differences in the game and the structure of the league. Data here is via Hockey Reference.
Looking in this context, the recent growth trend looks a little more meaningful even if we mentally account for the shootout goals. This season, the league is above its average in goal scoring relative to all the seasons since 1950 for the first time since 2005-2006, which was the year after the lockout and featured an emphasis on calling a tighter game leading to a temporary spike in goal scoring.
Prior to that, we have to go back to the late nineties to find a season above the average line. If the second half continues the way the first half started, this would be one of the better offensive seasons for the NHL in recent history.
Most things in sports are cyclical. Players get good at something, then they get better at something else to counteract that. And so on forever. We probably won’t see anything like the wild west shootouts of the 80s any time soon, if ever again, in the NHL. But we could see a run of years where the scoring gets above six goals per game.
For most of this century, goaltenders have changed the way the game is played through vast improvement in their technique. Maybe offense is finally starting to catch up to that by learning how to create more dangerous shots and converting on those chances.