Under John Tortorella, the Canucks get plenty of shots but few scoring chances

There are many reasons why the Canucks are on the verge of being mathematically eliminated from the playoffs, but the biggest can be summed up in three words: they can’t score. The team is 28th in the league in goals-per-game, ahead of only the woeful Florida Panthers and Buffalo Sabres. Their leading goalscorer is Ryan Kesler, with 23, and they’re likely to finish without a single 30-goal scorer for the first time in a non-lockout year since the 2007-08 season.

That season, they at least had three players with 20+ goals. Chris Higgins needs three goals in the Canucks’ final six games to reach 20 or Kesler will be the only Canuck above that mark.

The one bright spot? They’re currently on pace for 193 goals, which would be one better than their franchise low 192 in 1998-99. They need 14 goals in the final six games to avoid a historic low.

Many suggest the issue is personnel, that the Canucks lack legitimate snipers to finish chances, but the scoring struggles of normally reliable forwards like Alex Burrows and Daniel Sedin make me question that assessment. The Canucks could certainly use more talented goalscorers, but that doesn’t explain the struggles of the scorers they do have. For the moment, let’s set that aside and look elsewhere.

With Alain Vigneault back in town, it makes sense to look at coaching. Has John Tortorella’s coaching style and the offensive system he has put in place hurt the Canucks’ ability to score goals?

Tortorella’s system seems to focus on dump-ins to gain the zone with a hard forecheck to gain control on the puck. There has seemed to be an emphasis on shots from the outside, particularly the point, to create rebound opportunities for forwards going hard to the net. So, how does this impact their ability to create quality scoring chances?

In general, I tend to avoid discussing shot quality, as it’s a long and not particularly rewarding rabbit hole to go down. Shooting percentage tends to swing wildly from season-to-season to the point that it has next to no predictive power. A tiny fraction of NHL players can sustain higher-than-average shooting percentages and acquiring players based on their ability to put a higher percentage of pucks in the net tends to end in tears.

It’s important to note that no one is denying that shot quality exists — some shots are certainly of a higher quality than others, either because of the skill of the shooter or the location of the shot — but that shot quality is not particularly predictive.

The puck possession metrics that are based on shot attempts tend to be a lot better at predicting success. The most useful for this purpose is Fenwick Close, which measures goals, shots, and missed shots when the score is close, removing the oddities that occur when one team is leading by a significant amount.

What’s interesting is that the Canucks fare very well by this measure, with a 51.62 Fenwick Close percentage, ninth in the NHL. This would suggest that the Canucks are extraordinarily unlucky to not be in a playoff position right now, that their advantage in possession should have resulted in them outscoring their opponents far more frequently.

From this, you could make an argument that the Canucks are not as bad as their record indicates and that regression should see them scoring far more goals next season. I actually think there is some truth to this: there’s no way that Burrows can be as cursed next season as he was this year, for instance. But I am always going to be interested in outliers and it seems to simple to point just to one statistic when it is so far from where the Canucks currently sit in the standings.

So let’s look at shot quality just a little bit to see if it might have some explanatory power for the Canucks’ inability to score goals.

Thanks to the Super Shot Search tool, we can look at where the Canucks’ shots are coming from on the ice in comparison to their opponents. I looked at this earlier this season and noted that the Canucks take a higher than average number of shots both from beyond 30 feet and from within 10 feet, suggesting that Tortorella’s emphasis on shots from the point to create rebounds has led to chances in front.

The tool also allows us to narrow down the shot totals to just those within the “home plate” area in front of the net, ie. scoring chances. Puck possession statistics like Corsi have been shown to have a strong correlation to scoring chances, to the point that most advanced stat mavens have given up on the time-intensive work of tracking scoring chance data. So, if the Canucks Corsi doesn’t match up well with scoring chances, there’s a possibility that Tortorella’s systems have led to lower quality shots.

Before we tackle the numbers, a quick caveat: shot location data across the league is not particularly reliable. I’m counting on the volume of data to overcome that problem, but there’s every chance that it can’t, so let’s take this with a suitably-sized grain of salt.

These numbers are only at even-strength, so the Canucks woeful power play and recently struggling penalty kill are not included. They’re also from every game situation, so they’re not just from when the score is close.

The Canucks have out-shot their opponents 1880-1820 at even strength, controlling 50.8% of the shots. This is a direct match to their Corsi%. They’ve been out-scored, however, 147-136 at even strength, a percentage of 48.1%. As indicated above, this could simply be a result of bad bounces, but it’s interesting to note that the Canucks shots have come from an average distance of 38.1 feet, while their opponents have come from an average of 35 feet.

When we look at scoring chances, however, it’s a completely different picture.

The Canucks have been out-chanced 870-794 at even strength, controlling just 47.7% of the scoring chances. They have given up far more scoring chances than they have created, with a large percentage of their shots coming from outside the scoring chance area.

I’m not well-versed enough in statistics to determine whether the difference between the Canucks’ 50.8 Corsi% and their 47.7 Scoring Chance % — let’s shorten it to Chance% — is statistically significant, but it strikes me as incredibly interesting.

Let’s take a quick look at the Chance% for the Canucks’ Pacific Division rivals to see how it compares.

Team Fenwick Close ES Chances Chance%
San Jose Sharks 55.33 987 56.4
Los Angeles Kings 56.38 826 55.3
Anaheim Ducks 50.06 950 53.1
Vancouver Canucks 51.62 794 48.1
Calgary Flames 46.57 776 47.6
Edmonton Oilers 43.67 815 46.9
Phoenix Coyotes 49.35 727 46.2

 

That’s not pretty. The Canucks are a lot closer to the Flames and Oilers than they are to the Sharks, Kings, and Ducks. It’s also interesting to note how much better the Ducks’ Chance% is than their Fenwick Close, which might help explain why they’re at the top of the standings despite mediocre possession numbers.

The Coyotes, however, are on the other end of the scale, in a playoff position despite giving up far more scoring chances than they get. I definitely don’t like their chances in the playoffs, but it’s an indicator that we should definitely be careful not to read too much into these numbers.

Those scoring chance numbers for the Canucks are ugly, though, and suggest to me that there is a problem with the Canucks’ style of play. It’s not that the Canucks are bad at shooting the puck — or, at the very least, it’s not just that — but that the team does not create as many scoring chances as they give up.

Tags: ,

6 comments

  1. Chinook
    April 1, 2014

    Interesting analysis, to be able to quantify spectators’ impression. Good work Daniel.

    VA:F [1.9.16_1159]
    Rating: +2 (from 2 votes)
  2. Locky
    April 2, 2014

    Good stuff.

    A couple of musings on this topic:
    -An impression I have gathered of both the AV Canucks and this year’s Canucks is that this team tends to take less shots from the ‘slot’. As you say, it’s a point shot to rebound strategy, not a strategy focused on finding open players in this ‘home plate’ area. Also there is less tendency to fire shots off the rush – I wonder if this is quantifiable or just confirmation bias on my part.

    -One of the more recently repeated phrases in hockey analysis is ‘dirty goals’ i.e. rebounds and stick hacking in front of the net. I wonder if this is actually an advantage or not. Certainly this year would suggest that searching for those type of goals ONLY is not advantageous. But it seems ‘home-plate’ scoring chances more generally are, even though the subset that is rebound goals is not on its own.

    VA:F [1.9.16_1159]
    Rating: +2 (from 2 votes)
  3. iain
    April 2, 2014

    shooting from the outside is going to be a help if you have guys close in who can nab a rebound. far too often this season, the canucks have seemed unable to put someone on the edge of the crease or just beyond when they need to – there just seems to be a constant lack of hockey awareness about where to go, where to get to an opening, and, picking up an earlier theme, when they do so, they do it all at half-speed. it’s all just too depressing to think about for very long.

    but of course, far more significant in the canucks’ downfall is the total annihilation of their powerplay. if you fail to score on the power plays the refs give you, you are going to lose a lot of games.

    VA:F [1.9.16_1159]
    Rating: +2 (from 2 votes)
  4. DJ
    April 2, 2014

    This analysis certainly passes the eye test. All too often this season, the Canucks were getting shots, but not dangerous shots — shots from well out with no one in front, just wasting them at the goalie’s crest, resulting in a freeze and a face-off.

    I think it speaks to the de-emphasis on skilled play this year (i.e., Tort’s system). It takes skill to get good shots from the slot — think about all those goals created by Sedinery. Not so much if you’re just trying to throw them on net and hope for a rebound.

    More than anything, I want Torts gone before he can get his hands on Horvat and Shinkaruk (plus whatever skill/confidence Jensen has left).

    VA:F [1.9.16_1159]
    Rating: +2 (from 2 votes)
  5. Mt
    April 2, 2014

    I’d just like to try to strike down the “bad bounces” possibility. It seems to me that the sample size is too large for that to be possible within any kind or reasonable possibility. For that to happen to a team over a season has to indicate a a phenomenon that results from the actions, not from luck. Theoretically this could be caused by “bad bounces” but only in the sense that it is theoretically possible to flip a coin and come up tails a hundred times in a row. However once you get past flipping 5 or 10 or even 20 times, the sample size gets to big for luck to have any reasonable chance of playing out that way. For a player over part of a season or even a season, sure. But for a team over a season–no way. Just saying.

    VA:F [1.9.16_1159]
    Rating: 0 (from 0 votes)
  6. shoes
    April 2, 2014

    Torts may make a fine playoff style coach, but he will not get a western team to the playoffs with his style. As to the Yotes numbers going against the grain…..I don’t think so, because I think Dallas will bump them out of the 8th slot. Leaving them just in front of the Canucks. So the numbers never lied there either.

    The Torts experiement was a failure and nothing to do but take a better draft pick and move on to a new coach. Randy Carlyle may be on the market and he should have replaced Crawford after the Bertuzzi thing, instead of keeping the lame duck Crow around.

    VA:F [1.9.16_1159]
    Rating: +1 (from 1 vote)