Goals Per Game In The NHL

I have written a few posts, such as this one and this one which argue that the NHL's push for higher scoring games as a marketing tool is a bad idea. Many lower scoring teams have good attendance, while higher scoring teams have attendance problems. Scoring rate is not a significant determining factor in overall NHL revenues. Nevertheless, some people remain unconvinced. One commenter named Taylor remains convinced the NHL needs 25-40% more scoring and is willing to make radical changes to the game to make it happen.

I maintain that the NHL needs to make sure the games a good fair hockey games. That is far more important than the scoring rate. There is no easy "knob" for the NHL to turn to set the scoring rate. They have tried very hard to increase scoring in recent years with little success. In the meantime they have made some questionable changes to the game.

One change made to the game that was designed to increase the scoring rate, which has had little success, that is extremely questionable is the trapezoid behind the net. There is a trapezoid region behind the goal where a goalie is allowed to play the puck. If he is behind the goal and plays the puck outside this region he gets a penalty. The logic behind this rule is that it would make it harder to defend against the dumpin. There are some regions on the ice where the puck can be dumped where the goalie cannot field it. This rule has led to little if any change in goal scoring rates (it is impossible to isolate its change since the NHL instituted many rule changes at once). Aside from the fact it's an arbitrary stupid rule that adds an unnecessary complication to the game (on the rare instance where it is called); it will successfully lead to more dumpins. I don't think anyone ever suggested that the NHL would be more exciting if only they dumped the puck in more often.

The most successful method to change scoring rates in the NHL recently has been calling more penalties. The NHL instituted an obstruction crackdown when they relaunched the league after the lockout. One unavoidable problem that came with the increase in penalties was the increase in phantom penalties. Penalties were called when no rules were broken. If the puck was in a crowd and a player fell down a penalty was called. This was clearly unfair. A player got a penalty without actually breaking any rules. This situation needs to be avoided to have fair games. It is an unavoidable circumstance of calling more penalties. More penalties mean more false positive penalties. Players have adjusted to the new rule interpretation and referees have tried to adjust their calls to a sustainable level where they avoid false positive penalty calls. Scoring is slumping as a result. More penalties mean more goals. It doesn't matter if the penalties are deserved or not. If the NHL were to use this as a method to increase scoring, it would be more honest for each game to be a series of power plays. The game begins with the home team playing 5 on 4 hockey for two minutes. After that, the road team gets a two minute five on four power play. They swap power play opportunities all game. Scoring would be higher. There would be no need to call phantom penalties to keep scoring rates up. However, it would not be hockey as we know it.

It is informative to look at historical scoring rates. Here is a graph of goals per game in the NHL as a function of season (click on it to make it larger):

This graph goes back to the 1940/41 season. It is possible to extend the graph even further into the past, but the game was different enough that its interpretation becomes more complicated the further back into history we go. I thin this is far enough to capture the important points. I made this graph relatively quickly in microsoft excel and it extrapolates a value for the missing 2004/05 season, which was lost due to lockout. This makes the rise in scoring after the lockout look less sharp then it was.

Looking at the graph, we see an initial spike in scoring during World War II. The quality of NHL players was down because many were away at war. It is a well proven fact that scoring goes up in a lower quality league with the same rules. The AHL is always higher scoring than the NHL. The ECHL is always higher scoring than the AHL. The NHL had many players in its ranks who were the equivalent of AHL and ECHL players today, so scoring was higher. One lesson the NHL learns from this is to increase scoring all they need are worse players in the league. Maybe a war or some other artificial restriction to players playing in the league would help accomplish this. Of course, it would make the hockey played significantly worse.

Predictably, scoring slumped following the war. It reached its lowest levels in the mid-50's. Strangely, this is seen by historians as a "golden age" for the NHL. This is the heart of the original six years. Scoring may have been low, but hockey was great. There was no force pushing to increase scoring and attendance soared to never before seen levels.

In the 1960's scoring started to rise a little. The first reason for this is the advent of the slapshot. This modernized offences and gave a bit of an edge to the offensive players who had mastered it.

Scoring began a steady increase from the late 60's into the 1980's. This was a combination of two factors. Expansion led to more marginal players in the NHL and that increased scoring and the slapshot became more and more widespread. By the 1980's even a stay at home defenceman had a relatively good slapshot. It took some time for goaltending to catch up, but it did. The advent of the butterfly goalie style was effective in combating the slapshot and associated tipins and deflections since a goalie could block most of the lower part of the net where most of these goals were being scored.

The NHL expanded again throughout the 1990's but so did its talent pool. There were many players coming from places in Europe and USA to play significant roles in the league. This influx of players from outside Canada was able to exceed the rate of expansion and keep the talent levels at a high level. Better defensive strategies were widely introduced (traps, left wing locks etc.) to keep players from finding the open ice to score. By this time all goalies had learned to play the more effective butterfly techniques or other hybrids (like those played by Dominik Hasek) and the stand-up goalie had ceased to exist.

There was a minor peak in scoring from the obstruction crackdown after the lockout occurred, but it was far smaller than the World War II and expansion/slapshot peaks. Scoring now is at roughly the level that it was in the 1960's and higher than it was in the 1950's. These were decades of never before seen NHL attendance and revenues. The successes of the 1960's led to the expansion of the league. At no point then did an idea exist that hockey had to be higher scoring to succeed. The leaders of the NHL thought hockey was succeeding and tried to put it into more markets. The problem right now is that hockey has been placed into markets that are at best lukewarm to it. For whatever reason, the NHL marketing machine has decided that more scoring would help sell the game (despite the contradictory historical supporting data). Many things have been tried to increase scoring in recent years and they have had marginal if any success.

If we take Taylor's numbers from his comments that the NHL would be better with 25-40% more scoring, we need radical changes to achieve that. Making the goals much larger could achieve it, while compromising the integrity of the game. Perhaps changes that force a large percentage of the NHL talent out of the league (perhaps to play in Europe) might achieve it. The changes need to be much larger than we saw in the NHL relaunch after the lockout. Right now, there are 5.58 goals per game this season. A 25% increase in scoring would increase that number to about 7 goals per game. That level of scoring has only existed during World War II and during the 1980's. It is unreasonable high to be sustainable if we consider the NHL's historical trends. A 40% increase in scoring would mean 7.8 goals per game. This is nearly the peak in NHL scoring in its historical data over the past 70 years. That level of scoring is unsustainable over any long period of time.

There is a historical ebb and flow of offensive, defensive and goaltending techniques in NHL history. One makes an advance and in time the other make their advances to catch up. There may be an advance in offensive techniques on the horizon. This advance is not one that can be obtained from rule changes. It will happen from creativity of coaches and players. Scoring rates in the NHL are not so low. They are closer to the historical averages than they were in the 1980's (which was an anomalously high scoring decade) and has set the scoring expectations of some fans. Some of the best years in hockey history (the original six years) have had the current level of scoring or less. There is no desperate need to increase scoring to create a good hockey product.

There is no historical data to suggest that higher scoring hockey is more popular than lower scoring (in fact the opposite may be true). Scoring rates today are close to (but slightly below) the NHL's longterm average. Many low scoring teams do well in attendance while higher scoring teams struggle. The idea that the NHL needs to raise scoring for marketing purposes is misguided at best and possibly tremendously stupid. The level of scoring rates suggested (25-40% increases) will require significant changes. They cannot be obtained with minor tinkering to the game. The new game that would be produced will not be hockey as we know it. Likely it will be worse than hockey as we know it because the changes would not be driven by what is good for the game, but instead by what is going to increase scoring. You cannot make radical changes to hockey unless what is good for the game is your primary concern and even then you need to tread extremely carefully.

NOTE: Al Strachan has just written an article that touts some rule changes designed to increase scoring (that won't improve the calibre of hockey as far as I can tell) and likely won't make overall difference to the scoring rates anyway. Perhaps, he should have read this first.