With 70 NHL games played as of the morning of October 14, we are well into the “Why is scoring up?” season of hockey narratives. This will shortly be followed by “When will the coaches tighten it up and ruin the fun?” season, and I’ve seen the first hints of that along with the fall colours. Later on, we’ll all confidently discuss how the coaches ruined the fun of hockey and maybe we should make the nets bigger to compensate for this horrible over-coaching.
We had a particularly strong “Coach ruined the fun.” season last year in Leafs Land because the Leafs happened to go on a really hot run early, winning almost as many games as they have this year, and then they sagged back to a much more modest points pace. There are many people absolutely convinced Mike Babcock destroyed the Leafs in the name of...you know, I’ve never been sure — work ethic, maybe.
This year is starting out eerily similar. So, why is scoring up, though?
The short answer is: It isn’t.
This graph from a few days ago shows the goals scored in all games this season vs last to that same date. Scoring is largely equivalent. That could be the end of the story, but I wanted to dig into why everyone thinks it’s up, why they thought the same last year, and what really drives the narrative cycle of mean coaches spoiling all the fun.
Maybe it’s fair to ask people who think scoring is up just what they mean by up? Up over what? nobody decided to compare this year’s opening to last year’s opening to make that chart, but our memories hold all of last season, not just the first 50 or so games.
You will recall that at the start of last season, the NHL was enacting a crackdown on faceoff infractions and slashing. The result was a lot of early power plays, but we could tell that pace tailed off. One thing that’s obvious to anyone who watches hockey is that it’s easier to score on the power play than at even strength, so the more power plays there are overall in the league, the higher the scoring.
Hockey Reference has a very handy League Averages page which we can use to see this in action. The PPO (Power Play Opportunities) column is what we want to look at. Right now, it’s 3.46, and last year’s season average is 3.04 with 2.99 the year before that. So there’s more power plays right now in this season that last year’s average, but last year’s crackdown didn’t end up adding power plays overall by the time the season was over.
Scoring did go up last year. You can see that in the GAA column (finally a use for that), but the save percentage was almost the same as recent years. The overall shot rate had to be higher, driving more actual goals out of a similar save percentage. I’ll compare this year’s shot rates to last year’s in a moment.
I found last year’s Hockey Reference page in the Wayback Machine for October 25, and it shows that the PPO was at 3.93. The crackdown meant more early power plays than we’re having this season. So scoring last year should have been really high at first. The real question we should be asking is: Why isn’t the scoring this year even lower than it is?
How do you score a goal anyway? A dumb question, but the answer is that a player shoots the puck and it either goes in or it doesn’t. We can discuss how often this happens without ever using the word goalie since save percentage is just shooting percentage from their point of view. Let’s forget goalies, the equipment changes and the size of the net, and just look at shooting.
I chose to use Fenwick for this calculation. Fenwick, or all shots (Corsi) minus the ones that are blocked, is useful to compare the pace of play in one year over the other without having to worry if the difference is a variance in block rates like you might get with Corsi. Shots on Goal, a subset of Fenwick, varies because of the rate of missed shots, so Fenwick is a little more stable. To keep things consistent, I’m using Fenwick Shooting Percentage, which is just the percentage of unblocked shots that were goals.
For this season, there have been 70 games played to the end of October 13. For last year, there were 76 games played to the end of October 14. I’ve split each set of games into Power Play and Even Strength, using Natural Stat Trick’s data. Note that even strength includes four-on-four or three-on-three and power play leaves out the goalie pulled situations.
Before we look at goals, let’s look at shot rates, or the Fenwick For per 60 minutes for each game state in each year:
- 2017-2018 Even Strength: 44.00
- 2018-2019 Even Strength: 42.38
- 2017-2018 Power Play: 77.10
- 2018-2019 Power Play: 77.98
So while the power play shooting is a little more brisk this season, remember you’re seeing less than one more shot per hour of power-play time. These numbers are all essentially static shot rates from one year over the other.
Now to goals per game played year over year:
- 2017-2018 Even Strength: 4.34
- 2018-2019 Even Strength: 4.61
- 2017-2018 Power Play: 1.24 (Shorthanded: 0.26)
- 2018-2019 Power Play: 1.46 (Shorthanded: 0.20)
The rate of goals per game at even strength is up a little, but remember there has been more even-strength minutes this season. The real change is in the increased power play scoring in decreased minutes. We can expect from the above that the Fenwick Shooting Percentage should show a increase on the power play.
- 2017-2018 Even Strength: 6.28
- 2018-2019 Even Strength: 6.60
- 2017-2018 Power Play: 8.84 (Shorthanded: 8.81)
- 2018-2019 Power Play: 10.00 (Shorthanded: 7.37)
That’s not a typo, there have been 102 power-play goals on 1020 unblocked shots league-wide so far. That 10% shooting percentage on the power play is the sole reason scoring isn’t down over last year, since the power play opportunities have not matched last year’s.
Why is the shooting percentage so much higher? Well, it’s easy to assume that with so much similarity in all the other numbers that something has changed, but there are only 784 and 1030 minutes of power play time being considered here, so it could just be ordinary variance.
If it is a change from something other than just randomness, then there are two main ways it could be happening. First, the shooters or goalies could have changed in overall effectiveness. Second, this season, the teams with the better power plays might be getting a higher proportion of the power play opportunities. The reality is, it’s likely some of column A, some of B and some of C. Or maybe it’s just the Leafs’ fault for that absurd power play crew they put on the ice.
As we move deeper into the season, we should expect the number of power plays to decline like they did last year. If they stayed at the rate they’re at right now, it would be the highest since 2010-2011. If you go back in time from there on the Hockey Reference page, you can see that from 1980 to 2010, there were a lot more power plays, and then they’ve declined fairly steadily since 2010. This coincides very nicely with the dawn of new goalie techniques, the rise of the big-bodied goalies, and helps exaggerate the evidence that scoring rates in the NHL are driven by the goalies’ equipment, size and ability.
If we want more goals, we could just go back to a 1990s whistle-blowing pace. Whether that would be warranted by player behaviour is a good question, and a very difficult one to answer with anything but guesses. Maybe players pay more attention to their high-speed game now and don’t commit as many offences. Or maybe the referees and the league don’t want to be seen as deciding games with the whistle and they’ve just stopped calling as many.
Whatever the root causes are of changes in power play opportunities, the reality is that a lot of the “tightening up from the coach” that everyone is sure they can see is just fewer chances on the man advantage. The overall decline in scoring in recent decades is absolutely due to goalie changes, but not just due to that, because the power play rate is partly at fault.
So is scoring up? Not really, it’s just not down as much as we should expect, and as the season goes on, it will be interesting to see if that holds or not.
By the way, the Leafs are shooting a FF/60 of 85.34 on the power play, with a Fsh% of 24.24. The Leafs power play is much better than league average, but that shooting percentage is not going to hold forever.