Around the League 2017-18

[Jimby]

There wasn't really correlation with corsi and winning last regular season and playoffs too. To see a relationship between corsi for % and winning may just be "motivated reasoning".

Let’s wait and see what happens over a few months. Game to game it won’t tell you a whole lot because of such a small period. See the same trends over a longer period and teams on the positive end will likely be towards the top while the teams on the other end will likely be out of playoff spots.

 

[YWGinYYZ]

@Guffman - are players statisticians? Are they concerned about finding predictive patterns?

I'm more interested in the fact that all NHL teams have some form of advanced stats tracking internally than some anecdotal quotes from players that, quite honestly, probably have zero understanding of why people are looking at these numbers.

Ignorance is bliss, or it should be for them. Your arguments / quotes are an appeal to authority.

 

[Aavco Cup]

BTW, many players think Corsi is absolute GARBAGE. If players view it with such disdain, why is it looked at with such reverence by non-players? Here is more to that point from the article:

If there is one stat I could take away, it would be Corsi,” Seth Jones said. “Because I don’t think it gives a full representation of the game. You can be caught in situations that are not your fault. You will have four shots against and you have nothing do with that situation.”

In some cases, Corsi is even a running joke on the ice.

“I’ve never heard any player say ‘Oh man, I had a great Corsi last night. I crushed it!’” Oshie said. “We’ll joke about it in practice or in the summer league. To me, I don’t think it means that much.”

There was also some outright disdain.

“I think Corsi is garbage. It’s kind of a waste of time,” Eberle said. “I know stats. I pay attention to team puck possession – you know which teams are good at it. But I also think plus/minus is a waste of time, too. There was one year I had nine empty-net minuses.”

It has relevance because it has been shown to be a very good predictor of future results. It needs about 20 games or so until what it tells you becomes significant.

One game corsi stats don't do much other than describe what happened

 

[Aavco Cup]

Tank on?

 

[Guffman]

It has relevance because it has been shown to be a very good predictor of future results. It needs about 20 games or so until what it tells you becomes significant.

One game corsi stats don't do much other than describe what happened

Like Jimby just said:

There wasn't really correlation with corsi and winning last regular season and playoffs too. To see a relationship between corsi for % and winning may just be "motivated reasoning".

 

[Guffman]

@Guffman - are players statisticians? Are they concerned about finding predictive patterns?

I'm more interested in the fact that all NHL teams have some form of advanced stats tracking internally than some anecdotal quotes from players that, quite honestly, probably have zero understanding of why people are looking at these numbers.

Ignorance is bliss, or it should be for them. Your arguments / quotes are an appeal to authority.

Oh, sure...NHL teams do, as you say, have some form of internally tracked advanced stats that they guard religiously. It is also something that needs to be further refined.

While the publicly available advanced stats might be considered to be building blocks for future, more-refined and useful information, it doesn’t change that they are still relatively crude in its current state.

Ten years from now, I imagine people will scoff at these current metrics, much like players do today.

Anyways, I think we’re getting off-topic.

I am disappointed no one commented at my post about Drew Stafford scoring a sweet goal on Pavelec last night!

 

[Aavco Cup]

Like Jimby just said:

There wasn't really correlation with corsi and winning last regular season and playoffs too. To see a relationship between corsi for % and winning may just be "motivated reasoning".

There has always been a correlation between corsi and winning. Nobody has ever claimed that the correlation is 100%

 

[YWGinYYZ]

To see a relationship between corsi for % and winning may just be "motivated reasoning".

Psychoanalysis now?

Why would someone have motivated reasoning when it comes to Corsi (or other more advanced stats)?

There has always been a correlation between corsi and winning. Nobody has ever claimed that the correlation is 100%

Very correct. By its nature, correlation comes with a margin of error.

 

[nobody important]

I am disappointed no one commented at my post about Drew Stafford scoring a sweet goal on Pavelec last night!

Sorry.

How the hell did we let these guys get away? It's Lee Stempniak all over again.

Was that OK?

 

[Jimby]

That margin of error is in the coin flip range. Last season 6 teams with CF% above 50 missed the playoffs while 6 teams below 50 made the playoffs. In the playoffs themselves 6 of the teams with the highest corsi in the playoffs were bounced in the first round while 5 with the lowest moved on to thee next round(s). One of those went on to win the cup. Pretty hard to see any correlation at all.

Psychoanalysis now?

Why would someone have motivated reasoning when it comes to Corsi (or other more advanced stats)?



Very correct. By its nature, correlation comes with a margin of error.

There has always been a correlation between corsi and winning. Nobody has ever claimed that the correlation is 100%

 

[Duke749]

Do people not understand what “small sample” means or do they just pretend it doesn’t actually exists?

 

[Aavco Cup]

That margin of error is in the coin flip range. Last season 6 teams with CF% above 50 missed the playoffs while 6 teams below 50 made the playoffs. In the playoffs themselves 6 of the teams with the highest corsi in the playoffs were bounced in the first round while 5 with the lowest moved on to thee next round(s). One of those went on to win the cup. Pretty hard to see any correlation at all.

I think you are tackling this the wrong way. I would suggest the most effective way to predict success would be a rolling 20 game sample. What a team does and how they play in october or november don't mean much in april. I also think playoffs bring a very different dynamic (playing the same team up to seven times in a row) that change the nature of predictability.

I would also suggest that even the correlation you are suggesting here would likely be stronger in previous years.

 

[bustamente]

Anyone think that a coaching chance is a few weeks away if the Rangers continue their losing ways

 

[Mortimer Snerd]

Jets fans wringing their hands over our CF% should look at the Oilers. 66% in last night's game which they lost 6 to 1. At no time did they look like a team that was dominating possession yet the stats say they "should" have won...

So far in the NHL this year there is no correlation between corsi and winning. Jets are not alone.

Corsi is all about large amounts of data. That is what the entire concept it is built on. It may be descriptive in small samples - or not. 1 game is a very small sample. 5 games is still a small sample. Take another look after 20 games, or 50.

This season has been a little crazy so far. NJD, VGN and AVS among the league leaders?

 

[Duke749]

Anyone think that a coaching chance is a few weeks away if the Rangers continue their losing ways

Their fans certainly hope so.

 

[Mortimer Snerd]

I think they'll be disincentivized by losing their roster spot when the coaching staff pull the rug out from under them if they're not playing the systems requested.

I highly doubt that anyone is trying to "game" their Corsi stats.

If anyone is actually trying to game their Corsi it would likely only be those players who completely fail to understand the concept. They may actually believe that 'good Corsi = good play' instead of 'good play = good Corsi'. I think most players are smarter than that.

 

[Mortimer Snerd]

BTW, many players think Corsi is absolute GARBAGE. If players view it with such disdain, why is it looked at with such reverence by non-players? Here is more to that point from the article:

If there is one stat I could take away, it would be Corsi,” Seth Jones said. “Because I don’t think it gives a full representation of the game. You can be caught in situations that are not your fault. You will have four shots against and you have nothing do with that situation.”

In some cases, Corsi is even a running joke on the ice.

“I’ve never heard any player say ‘Oh man, I had a great Corsi last night. I crushed it!’” Oshie said. “We’ll joke about it in practice or in the summer league. To me, I don’t think it means that much.”

There was also some outright disdain.

“I think Corsi is garbage. It’s kind of a waste of time,” Eberle said. “I know stats. I pay attention to team puck possession – you know which teams are good at it. But I also think plus/minus is a waste of time, too. There was one year I had nine empty-net minuses.”

OK, I stand corrected. A lot of players are not smart enough to grasp the concept.

 

[Mortimer Snerd]

Do people not understand what “small sample” means or do they just pretend it doesn’t actually exists?

Yes.

 

[Whileee]

The real question is whether some coaches misunderstand Corsi and are coaching their players to just fire randomly towards the net whenever they can, thinking that high shot attempts will inevitably lead to a positive goal differential. There is more nuance in hockey than the brute correlation between Corsi and future success. I expect that some teams are coaching to different analytical concepts.

 

[Mortimer Snerd]

Anyone think that a coaching chance is a few weeks away if the Rangers continue their losing ways

Could be less than that if they continue.
Are the Rangers vulnerable to a bad trade?

 

[DeepFrickinValue]

I'm not high on the oilers because
1. McLellan's go to move is "moar MCDAVID". Powerplay, penalty kill, down a goal, etc., Double duty all night for the guy.

2. Draisatl is not a centre. Forget the boards all summer after that giant contract, but he is McDavid's winger. They terrify me together, but that's over 20 million dollars of eggs in one basket.

3. It's the oilers. They would kill the old Jets, they sucked for ten years, prevented talent from going to other teams, rushed their first overalls into the NHL (hampering development), and yet the day they scored McDavid the fans became cocky and entitled, the coach is an Adams candidate, and they're a cup favorite??

i have pity for the oilers.

draisatl out for extended time/concussion will be huge. with only mcdavid to focus on, he may be injured next with player continually following up on hits and double coverage.

$20 million on 2 players is crazy. jets are getting 4 permium players (ehlers, sheif, little, and wheeler) for $24 million or so. what would you rather have?

 

[Aavco Cup]

The real question is whether some coaches misunderstand Corsi and are coaching their players to just fire randomly towards the net whenever they can, thinking that high shot attempts will inevitably lead to a positive goal differential. There is more nuance in hockey than the brute correlation between Corsi and future success. I expect that some teams are coaching to different analytical concepts.

This has been going on for decades with coaching. The term they use is "get pucks to the net" pucks get tipped, goalie's get screened, rebounds happen etc.

More good than bad happens.

 

[Eyeseeing]

Was alot of fun seeing our guys get the W and then watching the Oilers get their poop pushed in. They've surpassed the Leafs for team I most get a kick out of losing. Their players/fans take it so hard

I like Edmonton fans, most are very chill.
Toronto on the other hand should have a minimum age requirement to post.
Many Toronto posters articulate like typical 14 year olds.

 

[Aavco Cup]

Are we seeing a "Vegas" effect as teams travel in to face the Golden Knights for the first time?

Bruins trailing 2-0

 

[Eyeseeing]

Are we seeing a "Vegas" effect as teams travel in to face the Golden Knights for the first time?

Bruins trailing 2-0

Vegas looks good to me.
Guessing the Vegas effort could be a thing.
Gallant is a good coach