Jets have lost 2/3 of their 20 best Corsi games

[broinwhyteridge]

Sorry buddy, but when people depend on "Luck" for their argument, its generally because they can't describe the complete picture. Probably the only reason we even hear of "Luck" in hockey is because its easier for players and coaches say to the media that "we were unlucky tonight" instead of "We just didn't have confidence" or "We just didn't feel like playing tonight" etc. They basically need excuses that some people will buy.

"Lucky" and "Unlucky" are go to excuses when players fail.
And for statisticians, when they use luck, its because their system is flawed.

Your example of getting hit by a truck at a light is a terrible example of "random chance" because there is a at least a second entity involved and there's no weight on individual chooses. And given this example I'd say that getting rammed at a light while stopped is probably because the light is green

To retort with an equally absurdity with is a person "lucky" that their wife divorced them? Or is it a series of events that can't be quantified neatly for a statistical analysis on the sidelines.

No idea if lucky or not but said divorced dude has a lot less stress probably. Money too.

 

[YWGinYYZ]

Good read regarding "luck" in hockey: http://www.arcticicehockey.com/2013/7/26/4513170/winnipeg-jets-nhl-stats-probabilities-luck

 

[SensibleGuy]

I can think of at least too occurances this season, both involving Bogo, that I would characterize as distinctly un-lucky...

 

[AriIQ]

You're getting into a silly word-argument.
luck doesnt exist, but probabilities do, and its considered "luck" when a certain probability you want to happens, happens.
Luck is just viewing a probability happening as positive or negative.

 

[Puckatron 3000]

Probably the only reason we even hear of "Luck" in hockey is because its easier for players and coaches say to the media that "we were unlucky tonight" instead of "We just didn't have confidence" or "We just didn't feel like playing tonight" etc. They basically need excuses that some people will buy.

I'll grant that luck is sometimes used inappropriately as an excuse.
Beyond that, I gotta join the "WTF?" pile-on.

Your example of getting hit by a truck at a light is a terrible example of "random chance" because there is a at least a second entity involved and there's no weight on individual chooses.

You'd have a point if there were no "second entities" in hockey. But there's TONS of other entities, all outside the control of the individual. Other players, refs, ice quality, equipment malfunctions, etc.

If humans had perfect understanding and control over physics, I'd buy your argument. But they don't. When the puck jumps over my stick due to an imperceptible wobble in the ice, is that the player's fault? Only if he could have accounted for that. But human eyes can't see an imperceptible wobble. Because it's like, imperceptible.

"Bad luck" means that the outcome was bad, despite the player doing the right thing (as limited by what is humanly possible).

You could also say that "bad luck" is when all the factors outside of our control (of which there are many in hockey) tend to go against you on a particular day.

Maybe your argument would hold weight for something like chess, which isn't subject the the laws of physics, and has no random elements. It doesn't fly for hockey.

 

[Skidooboy]

In another thread about stats , I think garret, or was it truck? Assured me that stats were perfectly predictable and that hockey wasn't at all a dynamic process.

And this has been my argument.
1. Stats work for baseball because the nature of the game is so static. 1 pitch 1 swing(or chance to swing), then pause and repeat.
Hockey is the opposite, play is continuous and dynamic. Corsi Fenwik etc only tell a part of the story, about 1/3 at best, and are not in any way exclusively definitive.

B) Possession doesn't equal points, it helps, but some guys can doodle around all might with the puck but never really produce much. Burmi is proof. So was Kovalev in his later years. Or Sundin.

• a player can have great stats but be a detriment to a team, Danny Heatley is great example. For Canada he's a great points producer internationally, except he gets all his points in round robin play against Latvia and Belarus. When it matters, Russia, US, etc , he not only disappears from the scoresheet, he actually becomes a detriment to his team, bad giveaways, poor d zone coverage, loses battles, etc. statistically he's at the top of the list, 5 on 5 I bet he's by far the best player in Canadian history for any stat you want to look at. Except for maybe Moose Watson.despite his great #s
I wouldn't want him on my team no matter what the stats say.

 

[Galbatross]

No idea if lucky or not but said divorced dude has a lot less stress probably. Money too.

Maybe.

My point is that that results aren't random. They follow the previous events to the conclusion.

 

[boanst]

Luck, variance, randomness whatever you want to call it, exists in sport

 

[Jimby]

Good and bad luck is a synonym in the great majority of instances, for good and bad judgment.
-- John Chatfield

 

[Puckatron 3000]

Good and bad luck is a synonym in the great majority of instances, for good and bad judgment.
-- John Chatfield

John Chatfield is a d*ck.
-- Puckatron 3000

 

[pucka lucka]

Luck in this context is variation in outcomes of seemingly random events. Good or bad is just the perspective of the perceiver. Luck is just a convenient colloquial term to describe events players and coaches can't control. do ppl think this a serious flaw in advanced stats? Like really?

 

[theamazingchris]

Luck in this context is variation in outcomes of seemingly random events. Good or bad is just the perspective of the perceiver. Luck is just a convenient colloquial term to describe events players and coaches can't control. do ppl think this a serious flaw in advanced stats? Like really?

I mean, it is a "flaw". But understanding how chance affects results, and controlling for it, is a huge part of statistics!

 

[theamazingchris]

In another thread about stats , I think garret, or was it truck? Assured me that stats were perfectly predictable and that hockey wasn't at all a dynamic process.

And this has been my argument.
1. Stats work for baseball because the nature of the game is so static. 1 pitch 1 swing(or chance to swing), then pause and repeat.
Hockey is the opposite, play is continuous and dynamic. Corsi Fenwik etc only tell a part of the story, about 1/3 at best, and are not in any way exclusively definitive.

B) Possession doesn't equal points, it helps, but some guys can doodle around all might with the puck but never really produce much. Burmi is proof. So was Kovalev in his later years. Or Sundin.

• a player can have great stats but be a detriment to a team, Danny Heatley is great example. For Canada he's a great points producer internationally, except he gets all his points in round robin play against Latvia and Belarus. When it matters, Russia, US, etc , he not only disappears from the scoresheet, he actually becomes a detriment to his team, bad giveaways, poor d zone coverage, loses battles, etc. statistically he's at the top of the list, 5 on 5 I bet he's by far the best player in Canadian history for any stat you want to look at. Except for maybe Moose Watson.despite his great #s
I wouldn't want him on my team no matter what the stats say.

They actually make up 1/2, not 1/3. They are predictive, but they are almost certainly not as much so as baseball stats. (I assume. I don't know jack about baseball stats.) It goes both ways. The baby can't go out with the bathwater. Also, possession doesn't equal points, no, but it is the best predictor of future points, so... yeah.

 

[garret9]

In another thread about stats , I think garret, or was it truck? Assured me that stats were perfectly predictable and that hockey wasn't at all a dynamic process.

And this has been my argument.
1. Stats work for baseball because the nature of the game is so static. 1 pitch 1 swing(or chance to swing), then pause and repeat.
Hockey is the opposite, play is continuous and dynamic. Corsi Fenwik etc only tell a part of the story, about 1/3 at best, and are not in any way exclusively definitive.

B) Possession doesn't equal points, it helps, but some guys can doodle around all might with the puck but never really produce much. Burmi is proof. So was Kovalev in his later years. Or Sundin.

• a player can have great stats but be a detriment to a team, Danny Heatley is great example. For Canada he's a great points producer internationally, except he gets all his points in round robin play against Latvia and Belarus. When it matters, Russia, US, etc , he not only disappears from the scoresheet, he actually becomes a detriment to his team, bad giveaways, poor d zone coverage, loses battles, etc. statistically he's at the top of the list, 5 on 5 I bet he's by far the best player in Canadian history for any stat you want to look at. Except for maybe Moose Watson.despite his great #s
I wouldn't want him on my team no matter what the stats say.

WTF? I'm 90% sure none of us have said that cos that is ridiculous. We all know stats arent 100% predictive. That be a ****ing weird world if it was. We know hockey is dynamic. Point is these numbers are still tell you things. Some of these things are important. We take consideration how important things are and take account of it.

1. No matter if it is 1/3 -which it is not, Corsi/Fenwick + natural variance actually account for about 75% of winning percentage, with natural variance capped at 30%-, it still is an important part of the game. That's the whole point. You should probably read the article Puckatron posted it would help things.

B. Possession is still predictive of wins. Also, these stats don't have ONLY a relationship puck possession, but also scoring chances. People are ignoring that the original work was involved with their relationships with scoring chances is why we use these numbers, puck possession is actually secondary. Corsi is (generally) considered a puck possession stat and Fenwick is (generally) considered a scoring chance stat since they both are a tad closer to that than the other; however, both still have a relationship with both. Burmistrove is actually an example of why it does work; other good things happened which help win games when Burmistrov was on the ice... Including scoring chances. This is in part why Kane scored at such high rates with Burmistrov. This comes back to the first point, that Corsi isnt everything. Corsi (and Fenwick) aren't everything, just they are a large input so they have a large consideration in evaluating individuals. This is why neither Truck or I would say Burmi was the Jets best player but an undervalued player. There is a difference. Burmi had his own faults, one being scoring.

* Your example is highly flawed. What about the players that scored against the tough teams but not the easy teams? If player's could choose when they score, it would be always, but it isn't like that. Besides, I bet that his Corsi wouldn't have looked so hot at 5v5 those poor games.

I once saw a quote that predates hockey adv statistics that seems on point here.
It basically said early resistance to new stats is usually because people expect it to cover and represent all things.
It is true in this case. The fact that it does not cover all things and tells you everything, 100% of the time, doesn't diminish what it can and does say.

 

[garret9]

It is highly evident that the main detractors are from people who do not understand those who use these metrics and their actual point of view.

I would highly suggest reading the article Puckatron posted near the beginning and the articles it links to. Then return and this discussion could be continued in a more constructive manner.

 

[Grind]

It is highly evident that the main detractors are from people who do not understand those who use these metrics and their actual point of view.

I would highly suggest reading the article Puckatron posted near the beginning and the articles it links to. Then return and this discussion could be continued in a more constructive manner.

If only the explanations were available in a highly consumable, aesthetically pleasing manner....

 

[lablite47]

It's pretty simple Corsi doesn't mean much when you have below average goaltending and a bunch of decent players who lack finishing skills.We have a fast hard working,gritty team right now thanks to coach Mo but we need to make a move to add some pure skill to our line up.

What I've been saying for a few years now.. don't need corsi stats to tell what my eyes see. Not enough skill and finish.

 

[Sweech]

I love the Tim Tebowers of these types of threads.

#JustWinz

 

[garret9]

Are the ones that saying Corsi doesn't work for Jets the ones ignoring the gm0ney post on the first page?

It worked for the Jets too!

 

[garret9]

Oh hey look everyone, it's scoring chances and Fenwick...

Well whatdyaknow...
Jets are similar in both.

 

[tacogeoff]

What I've been saying for a few years now.. don't need corsi stats to tell what my eyes see. Not enough skill and finish.

agreed. but they are interesting and a great thing for stat junkies.

 

[Joe Hallenback]

agreed. but they are interesting and a great thing for stat junkies.

"Mathematics is written for mathematicians."
-- Copernicus

 

[Mathmew Purrrr Oh]

"what the **** is differential calculus?" - Copernicus
"what the **** is this gravity business?" - Copernicus
"what the **** is Chipotle?" - Copernicus

 

[pucka lucka]

"Mathematics is written for mathematicians."
-- Copernicus

Is that supposed to mean something?

Btw Copernicus
Died: May 24, 1543, Frombork, Poland

Maybe you can find some good quotes from Aristotle. Maybe ether can explain luck?

 

[garret9]

"85% of quotes are made up and/or taken out of the context that they are intended, man"
- Jeus H. Christ, Superstar