Corsi, shot quality, and the Toronto Maple Leafs

TheDevilMadeMe

Registered User
Aug 28, 2006
52,271
6,981
Brooklyn
Here's where Leafs fans are missing the point: NO TEAM has been able to sustain the Leafs' shooting percentage from last season in modern history. You can hem and haw about how not all shots are created equal, however unless you think the Leafs are the best team the NHL has seen at creating scoring chances in a long time (hint: they're probably not), expect their shooting percentage to go down to more reasonable levels.

For some comparison, the Pittsburgh Penguins have consistently been the team with the best shooting percentage in the NHL (makes sense with Crosby and Malkin up front), and even they're at like a 9.5% average over the last five seasons.

What's "modern history?" Is it since 2006? Or since the late 70s/early 80s?
 

HarrisonFord

President of the Drew Doughty Fan Club
Jul 20, 2011
21,918
1,844
Toronto
Here's where Leafs fans are missing the point: NO TEAM has been able to sustain the Leafs' shooting percentage from last season in modern history. You can hem and haw about how not all shots are created equal, however unless you think the Leafs are the best team the NHL has seen at creating scoring chances in a long time (hint: they're probably not), expect their shooting percentage to go down to more reasonable levels.

For some comparison, the Pittsburgh Penguins have consistently been the team with the best shooting percentage in the NHL (makes sense with Crosby and Malkin up front), and even they're at like a 9.5% average over the last five seasons.

Have you considered magic?
 

TheDevilMadeMe

Registered User
Aug 28, 2006
52,271
6,981
Brooklyn
The last 10 years at least. I have a spreadsheet somewhere with all the team shooting percentages.

Shooting percentages in the 70s and 80s were obviously higher than they were today.

Meh, last 10 years a a pretty weak sample size, considering the claim is that the Leafs play a unique style. If the Leafs were shooting better than the 80s Oilers or 90s Penguins, I'd agree that they would be likely to regress. But I don't think any team in the last 10 years was particularly known as a good shooting team.
 

JoelWarlord

Ex-Noob616
May 7, 2012
6,111
9,351
Halifax
Meh, last 10 years a a pretty weak sample size, considering the claim is that the Leafs play a unique style. If the Leafs were shooting better than the 80s Oilers or 90s Penguins, I'd agree that they would be likely to regress. But I don't think any team in the last 10 years was particularly known as a good shooting team.

2009-10 Capitals?
 

Ari91

Registered User
Nov 24, 2010
9,900
30
Toronto
The Habs had the majority of the possession in last years series against the Sens. The difference was that Anderson was amazing.

The Habs showed in 2010 that a hot goalie can overcome bad possession, the problem is that a hot goalie won't last forever so being outshot all the time is not a good idea.

Even if this strategy of Carlyle's actually works in the regular season, I'd argue that it'll be less effective in the playoffs when teams spend more time 5 on 5. Also, instead of getting outshot badly by a team like Minnesota, you're getting outshot badly by a team like Pittsburgh. Look what happened in Game 7, the 2 goals to tie the game were created from those harmless shots from the outside that Carlyle lets happen.

I don't think any coach would say that puck possession is bad. So far into this season Carlyle already sounds like a broken record talking about the team needing to improve their puck possession and limit scoring chances against. It's not that Carlyle allows or wants his team to get badly out-possessed and rely on opportunistic scoring. That isn't the strategy. The strategy is always possession but what happens when you don't have possession? This is where I think the Leafs really take the system for granted. The most efficient way of playing in your defensive zone is being able to quickly get the puck out. If their lives depended on them clearing their zone quickly, every single player in a Leafs jersey would drop down dead on the ice. They are ridiculously horrible in that area of the game.

When they get stuck in their own end, they play very conservatively. They won't chase the puck, they'll get in their formation and start blocking lanes, they'll block shots and they'll look to clear bodies that are in their goalie's way from making a save. At least that's my theory about why they get stuck in their own end too often.
 

TheDevilMadeMe

Registered User
Aug 28, 2006
52,271
6,981
Brooklyn
I don't think any coach would say that puck possession is bad. So far into this season Carlyle already sounds like a broken record talking about the team needing to improve their puck possession and limit scoring chances against. It's not that Carlyle allows or wants his team to get badly out-possessed and rely on opportunistic scoring. That isn't the strategy. The strategy is always possession but what happens when you don't have possession? This is where I think the Leafs really take the system for granted. The most efficient way of playing in your defensive zone is being able to quickly get the puck out. If their lives depended on them clearing their zone quickly, every single player in a Leafs jersey would drop down dead on the ice. They are ridiculously horrible in that area of the game.

When they get stuck in their own end, they play very conservatively. They won't chase the puck, they'll get in their formation and start blocking lanes, they'll block shots and they'll look to clear bodies that are in their goalie's way from making a save. At least that's my theory about why they get stuck in their own end too often.

No math behind this (so why am I on this board?), but I would imagine that the Leaf's offensive performance is more sustainable than their defensive performance. Just from watching hockey, I think a quick-strike counterattack IS a viable way to play offense if you have the right horses, but getting stuck in your own zone will eventually lead to trouble.

Or maybe not, the Julien-era Bruins sure seem to do well defensively with allowing lots of shots.
 

TieClark

Registered User
Jun 14, 2011
4,112
0
As mentioned, the Leafs collapse and simply block any prime chances but by doing this, the other team can maintain possession for extended periods. I don't think this a problem if you're able to block the chances consistently, especially when you're confident in your goalie.

The real problem right now with the Leafs is the breakout. They seem to have a ton of trouble breaking the puck out.... almost stuck between a board play and a transition team. Some guys are flying up trying to stretch the D out, some guys are just chipping it off the boards. It's an area they need to work on for sure if they want to be successful.
 

Master_Of_Districts

Registered User
Apr 9, 2007
1,744
4
Black Ruthenia
Hard to tell what happened with the 2009-10 Capitals - did they regress to the mean or did they suffer from injuries and Ovechkin's cheeseburger days?

That team was both very good (dominant possession team at even strength with a strong powerplay) and very lucky (shot ~10% at even strength despite having league average numbers (~8%) in that respect in both the following and preceding seasons).

By 2010-11, they were no longer controlling the play to the same degree at even strength (although they were still relatively good), and their shooting luck evaporated. Hence the regression.
 

Master_Of_Districts

Registered User
Apr 9, 2007
1,744
4
Black Ruthenia
Regarding the Leafs specifically, it's possible that they possess above average shooting talent. Maybe not to the extent that the raw numbers would suggest, but above average nonetheless.

Even though the spread in talent is much narrower than the spread in observed results (about 1/3 as wide after 82 games; about 1/4 as wide after 48), it's clear that some teams have more shooting talent than others.

The talent standard deviation is about half a percent (a touch lower, but close enough) for EV shooting percentage at the team level, and the shape of the distribution is close to normal (slightly right-skewed, but, again, close enough).

So, in a typical year, the best shooting team in the league might have a true talent EV shooting percentage of 9.0% or therebouts.

(Of course, it's a virtual given that some team will exceed that mark in terms of results, but that's not what we're talking about.)

Teams with poor possession numbers also tend to go a bit better in this regard, so that's another factor.

I dunno - I'm just kind of rambling here, but I suppose my point is that about half of the teams in the league, by definition, should possess a true talent EV shooting percentage greater than 8.0%, with about five or so exceeding 8.5%. That's what the data says, at least.

So maybe the reasonable expectation is for the Leafs to shoot between 8.0% and 9.0% in the long run. That's a broad range, but there's a lot of uncertainty with this type of thing.
 
Last edited:

Burke the Legend

Registered User
Feb 22, 2012
8,317
2,850
As mentioned, the Leafs collapse and simply block any prime chances but by doing this, the other team can maintain possession for extended periods. I don't think this a problem if you're able to block the chances consistently, especially when you're confident in your goalie.

The real problem right now with the Leafs is the breakout. They seem to have a ton of trouble breaking the puck out.... almost stuck between a board play and a transition team. Some guys are flying up trying to stretch the D out, some guys are just chipping it off the boards. It's an area they need to work on for sure if they want to be successful.

I wonder if there's a correlation between collapsing and poor breakout...hmmm :sarcasm:
 

TieClark

Registered User
Jun 14, 2011
4,112
0
I wonder if there's a correlation between collapsing and poor breakout...hmmm :sarcasm:
The Bruins do it very well, they collapse but are very quick at breaking the puck out. Just need to have the right system in place to allow for quick and consistent breakout options.
 

BoredBrandonPridham

Registered User
Aug 9, 2011
7,573
4,061
As mentioned in this thread, a team with a great Shot Differential can end up in the bottom of the standings due to poor goal-tending and poor special teams.

As we saw last season a team with a poor Shot Differential can end up in the top of the standings due to great goal-tending and great special teams. But, why aren't great goal-tending and great special teams considered sustainable? They certainly have been for almost 60 games.

Also, don't fall victim to the gambler's fallacy that just because Leafs have had "high shooting %" for the last 60 games that they are due to fall down to *below reasonable* shooting percentage in the future.

The Leafs play a strong transition game, does that not mean that the leafs are going to play with forwards high in breakout position? This would mean lower D coverage down low (take more shots) and better rush chances up the ice (higher shooting %)?

Just as I type this, Kessel scores the hat-trick on the Transition 2-on-1. Luck?
 

BoredBrandonPridham

Registered User
Aug 9, 2011
7,573
4,061
Also, why not use Goal Differential as a proxy for Quality Shots For v.s. Quality Shots Against, then you might actually be talking about something that matters.
 

saskganesh

Registered User
Jun 19, 2006
2,368
12
the Annex
I see the Leafs won again. According to the stats, they were again outshot, they lost more faceoffs, and blocked twice as many shots.

All of that is interesting, but doesn't really tell me how the game played out or why the Leafs won as they did.
 

Community

44 is Rielly good
Oct 30, 2010
6,774
1,683
The Darkest Timeline
I see the Leafs won again. According to the stats, they were again outshot, they lost more faceoffs, and blocked twice as many shots.

All of that is interesting, but doesn't really tell me how the game played out or why the Leafs won as they did.

Well first off the shots were pretty even, 2 more shots for Anaheim and that would be because the leafs tried to "shut it down" in the last ~3 minutes to hold the lead. Faceoffs were once again horrible, and the blocked shots were 21-14 (not twice as many).

Poor 1st, Great 2nd where they scored 3 unanswered, and then an okay 3rd where they scored another goal and did a pretty good job defensively to hold the lead.
 

MVP of West Hollywd

Registered User
Oct 28, 2008
3,527
976
For the Corsi experts, what are you opinions on Corsi itself being affected by luck? For example I see some years like Leafs 4th in 09-10 or NJD 2nd in 12-13 in CF% that look like possible 'Corsi luck' years considering the weaker performance in surrounding seasons

For example the Ducks were 22nd in CF% in 12-13 and are 3rd so far this year. What if the Ducks or a team with a similar improvement, had similar scoring chances for/against both years - but in the year with the high Corsi, suddenly were taking a lot more low quality shots than the other team. Now their Corsi goes up, their shooting % goes down, but the quality of play hasn't really changed. The "unsustainable" part of their play in 12-13 (shooting %) is in reality, that they were getting more scoring chances than the other team without also at the same time, taking more shots than them by equal magnitude. What the "shot quality is unsustainable" evidence shows is that over the long run total shots and scoring chances end up correlated. That's why you can't consistently out shooting % other teams, because when you're outplaying them enough to do, at least in the modern NHL you always end up taking more low quality shots and thus total shots than them too. In the long run root cause of both low quality shots and high quality shots is the same, in other words, possession and outplaying the other team. But in the short run it's obviously the case that a team can get more scoring chances, but take less shots than the other, if the opponent is blasting poor perimeter shots only

In the case of the Leafs, it could be that they're creating less scoring chances than the other team on average and getting lucky. But it could also be that they're creating more than the other team on average and the "luck" is that the other teams are getting more low quality shots, in a way that is unsustainable if the Leafs keep playing well enough to win the high quality shot game - in other words, the Leafs total shot count starts to mirror their high quality shots more and their Corsi goes up

My gut says the Leafs are indeed a weak team who collapses, but nonetheless. The fundamental problem for Corsi is that it has to use total shots 5 on 5 as a proxy for scoring chances. And while in the long run the evidence shows this is fair, it's absolutely possible in the short run that total shots and scoring chances are not lining up perfectly. That by 'luck' a team is having more or less low quality shots than their scoring chances suggest they should have. Then their Corsi as I suspect it was for the Leafs in 09-10, ends up the wrong and lucky one. The Avs for example could be a team who is really playing that well, but by chance the opponent is shooting way more poor shots than expected for their scoring chances/quality of play and likewise the Avs are shooting less low quality shots. Or something. Corsi % is clearly affected by low quality shots in one way or the other, despite these shots having a small possibility of affecting the game.
 

MVP of West Hollywd

Registered User
Oct 28, 2008
3,527
976
Two reasons

1) the stats community has come up with predictive models not involving shot quality

2) too many variables go into "what's a quality shot" to easily represent numerically

So instead of trying to get to the bottom of how to represent "shot quality," the tendency is to say "over a long enough timeline all teams will give up the same percentage of quality shots," even though a scatterplot/heatmap or, Hell, a pair of eyes can show that this is unambiguously wrong.

It's like those ancient representations of the solar system that had planetary orbits that reversed themselves. It worked most of the time, so the people who believed in them fought against anything challenging the orthodoxy of the model, even though the model was obviously illogical.

I don't think the Corsi guys believe no shot quality exists, since eyes or data showing how many goals are in front of the net, makes that obvious.

What their studies imply, is that the teams that get more shots period, get more good shots. Over the long haul there isn't teams that consistently jack up more low quality shots per scoring chance that is average, or the opposite, take less low quality shots per scoring chance than is average. It goes together, whatever causes a team to end up with more total shots, causes them to get more scoring chances - thus no difference in shooting %
 

Rants Mulliniks

Registered User
Jun 22, 2008
23,058
6,116
I see the Leafs won again. According to the stats, they were again outshot, they lost more faceoffs, and blocked twice as many shots.

All of that is interesting, but doesn't really tell me how the game played out or why the Leafs won as they did.

If you go here http://scores.espn.go.com/nhl/gamecast?gameId=400484372 you'll see yet again how the Leafs keep the other team to low quality scoring areas while having a large percentage of their own shots for in high scoring areas.

Conversely, if you go to the Chicago game, they allowed Chicago tons of chances from high quality areas and had few there themselves.

This keeps playing out over and over again to largely the same results.
 

Rants Mulliniks

Registered User
Jun 22, 2008
23,058
6,116
I don't think the Corsi guys believe no shot quality exists, since eyes or data showing how many goals are in front of the net, makes that obvious.

What their studies imply, is that the teams that get more shots period, get more good shots. Over the long haul there isn't teams that consistently jack up more low quality shots per scoring chance that is average, or the opposite, take less low quality shots per scoring chance than is average. It goes together, whatever causes a team to end up with more total shots, causes them to get more scoring chances - thus no difference in shooting %

One would suggest that the season where Toronto finished 29th overall despite being among the shot leaders (coupled with having watched every game) disputes what you are stating (unless you think any shot on goal is a "scoring chance".....Jason Blake says otherwise).
 

Erndog

Registered User
Jul 17, 2007
4,092
1,525
Forgive my ignorance (still learning the advance stats stuff) but how do the advanced stats guys differentiate Kessels snipe on the 2 on 1 vs, say, a random bad angle shot from someone like Bodie or something?
 

TieClark

Registered User
Jun 14, 2011
4,112
0
Forgive my ignorance (still learning the advance stats stuff) but how do the advanced stats guys differentiate Kessels snipe on the 2 on 1 vs, say, a random bad angle shot from someone like Bodie or something?
They don't, and that is why it's flawed. It's the entire reason the Leafs look terrible in advanced statistics... they treat low quality shots as equal to high quality shots and with the Leafs system, they allow a lot of low quality shots against and generate most of their offence off the rush which means low quantity, but high quality of shots for them.
 

Master_Of_Districts

Registered User
Apr 9, 2007
1,744
4
Black Ruthenia
For the Corsi experts, what are you opinions on Corsi itself being affected by luck? For example I see some years like Leafs 4th in 09-10 or NJD 2nd in 12-13 in CF% that look like possible 'Corsi luck' years considering the weaker performance in surrounding seasons

For example the Ducks were 22nd in CF% in 12-13 and are 3rd so far this year. What if the Ducks or a team with a similar improvement, had similar scoring chances for/against both years - but in the year with the high Corsi, suddenly were taking a lot more low quality shots than the other team. Now their Corsi goes up, their shooting % goes down, but the quality of play hasn't really changed. The "unsustainable" part of their play in 12-13 (shooting %) is in reality, that they were getting more scoring chances than the other team without also at the same time, taking more shots than them by equal magnitude. What the "shot quality is unsustainable" evidence shows is that over the long run total shots and scoring chances end up correlated. That's why you can't consistently out shooting % other teams, because when you're outplaying them enough to do, at least in the modern NHL you always end up taking more low quality shots and thus total shots than them too. In the long run root cause of both low quality shots and high quality shots is the same, in other words, possession and outplaying the other team. But in the short run it's obviously the case that a team can get more scoring chances, but take less shots than the other, if the opponent is blasting poor perimeter shots only

In the case of the Leafs, it could be that they're creating less scoring chances than the other team on average and getting lucky. But it could also be that they're creating more than the other team on average and the "luck" is that the other teams are getting more low quality shots, in a way that is unsustainable if the Leafs keep playing well enough to win the high quality shot game - in other words, the Leafs total shot count starts to mirror their high quality shots more and their Corsi goes up

My gut says the Leafs are indeed a weak team who collapses, but nonetheless. The fundamental problem for Corsi is that it has to use total shots 5 on 5 as a proxy for scoring chances. And while in the long run the evidence shows this is fair, it's absolutely possible in the short run that total shots and scoring chances are not lining up perfectly. That by 'luck' a team is having more or less low quality shots than their scoring chances suggest they should have. Then their Corsi as I suspect it was for the Leafs in 09-10, ends up the wrong and lucky one. The Avs for example could be a team who is really playing that well, but by chance the opponent is shooting way more poor shots than expected for their scoring chances/quality of play and likewise the Avs are shooting less low quality shots. Or something. Corsi % is clearly affected by low quality shots in one way or the other, despite these shots having a small possibility of affecting the game.

Certainly, we observe some movement among teams from season-to-season.

But within any given season, the split-half reliability of Corsi is quite high - on the order of 0.80 - 0.90. By comparison, the split half reliability of goal differential or points percentage is only about 0.40 - 0.50.

Technically, there is some luck component to Corsi - a team with a true talent Corsi percentage of 60% isn't likely to obtain exactly 60% of the Corsi events over a single season. Even if that team played the same way all of the time.

Luck/randomness is part of the reason why the split half reliability is less than 1, but not the entire explanation. If luck were the only factor, we would expect to observe a split-half reliability of about 0.95. Thus, injuries, lack of uniformity in strength of schedule, and differences in underlying play have a causal role as well.
 

Ad

Upcoming events

Ad

Ad