3 of the top 4 Corsi teams in the league are in the Final Four.

[Filthy Dangles]

Corgi is whatever at this point. It has value, people on both extremes in here overrate/underrate it's importance.

One issue with it is the phenomenon (can't recall the name) is where a team might start try to game and win Corsi battle in and of itself, which is just a measure and not the actual goal. Hurricanes seem to be in this boat some and might explain why they underperform in terms of expected winning vs their corsi so often.

People get too caught up on the %'s themselves and think teams are actively gaming this rubbe chuck battle. The real key is can a team play a brand of hockey like the Panthers who don't artifically pump rubber @ the net. Their real key is suppressing shots and playing a suffocating brand of hockey and have the talent to score.

 

[Drew4u]

Corsi isn't meaningful. It's just one of the classic ways human beings over-complicate things. We take basic stats and keep adjusting them because we think it's better.

There's a lot of luck and randomness in the sport so why try to come up with all these advanced stat models that "in theory" are supposed to be better. I have news for you, it's a bunch of b.s. It's because younger generations think that everything in the past was wrong and needs to be changed. A combination of the eye-test and basic stats is all anyone needs.

And yes statistics has as much bias as one's own personal thoughts. There's almost nothing objective (free of bias) in this world. To think these advanced stats are better than the basic ones is completely delusional. You know what is objectively true though, advanced stats ARE more complicated.

If advanced stats were available back in the 90s I'm sure I could find a bunch of "proof" that Gretzky was an above-average player at best. Complete waste of time. You can't predict the weather, the stock market, and you can't predict sports games.

 

[JaegerDice]

Corsi isn't meaningful. It's just one of the classic ways human beings over-complicate things. We take basic stats and keep adjusting them because we think it's better.

There's a lot of luck and randomness in the sport so why try to come up with all these advanced stat models that "in theory" are supposed to be better. I have news for you, it's a bunch of b.s. It's because younger generations think that everything in the past was wrong and needs to be changed. A combination of the eye-test and basic stats is all anyone needs.

And yes statistics has as much bias as one's own personal thoughts. There's almost nothing objective (free of bias) in this world. To think these advanced stats are better than the basic ones is completely delusional. You know what is objectively true though, advanced stats ARE more complicated.

If advanced stats were available back in the 90s I'm sure I could find a bunch of "proof" that Gretzky was an above-average player at best. Complete waste of time. You can't predict the weather, the stock market, and you can't predict sports games.

This is the most 'Ok boomer' post I have ever read on these forums.

The game of hockey has changed very little over time. Understanding of the game and what matters has changed. Once upon a time assists weren't tracked. Once upon a time shots weren't tracked. I'm sure back then, there were a bunch of greybeards shouting about how it's all meaningless noise and all you need to follow is goals!

Now we track a bunch more. Nobody is adding complication, we're simply looking deeper into what has existed the whole time. Turning cliches like 'Just gotta keep shooting, keep getting chances, the puck will go in eventually' into measurable trends.

Also, somebody should probably inform Warren Buffet all the millions of dollars he made predicting stocks are invalid, and tell every meteorologist that understands how pressure systems work that it's all magic.

 

[mcdingdong]

5 on 5
HDCF and HDCA need to sort, moneypuck doesn’t translate sorted links.
HDCF none in the final 4
HDCA least, none in final 4.

NHL Team Stats

Stats and Graphs For All NHL Teams

moneypuck.com

Try HDCF versus HDCA as a ratio or percentage. It's more informative than raw totals of either used in isolation as you've tried (reason for this shouldn't need to be explained), and unsurprisingly 3 of the top 7 in HDCF%, 3 of the top 5 in SCF%, and THE top 3 in xGF% 5v5 are in the final 4 teams.

 

[Golden_Jet]

This is the most 'Ok boomer' post I have ever read on these forums.

The game of hockey has changed very little over time. Understanding of the game and what matters has changed. Once upon a time assists weren't tracked. Once upon a time shots weren't tracked. I'm sure back then, there were a bunch of greybeards shouting about how it's all meaningless noise and all you need to follow is goals!

Now we track a bunch more. Nobody is adding complication, we're simply looking deeper into what has existed the whole time. Turning cliches like 'Just gotta keep shooting, keep getting chances, the puck will go in eventually' into measurable trends.

Also, somebody should probably inform Warren Buffet all the millions of dollars he made predicting stocks are invalid, and tell every meteorologist that understands how pressure systems work that it's all magic.

Corsi isn’t it, if you want advanced stats , that would be near the bottom for teams, when they get much more useful info from Sportlogiq.

Now if you go back to the start of the cap, for Corsi, you will get more information, than the very small sample size you chose.

 

[JaegerDice]

Corsi isn’t it, if you want advanced stats , that would be near the bottom for teams, when they get much more useful info from Sportlogiq.

Now if you go back to the start of the cap, for Corsi, you will get more information, than the very small sample size you chose.

Do you actually subscribe to Sportlogiq? Or subscribe to any analytics site that does pay for their data? Cause your position somewhat misrepresents the data they provide.

First of all, Sportlogic tracks Corsi and has their own Expected Goals For model that, like every xGF model, weighs Corsi heavily.

In addition to traditional stats like shots, shot attempts, goals, shot type, shot distance, shot angle, etc, SportLogiq tracks a lot of microstats. Or, as Sportlogiq brands them, 'Possession driving plays'.

Possession driving plays are plays that they have found to correlate to... wait for it... higher Corsi.

Possession driving plays include controlled exits, controlled entries, puck battles won, etc.

They are very explicit that each of these individual stats are not predictive of future goals or wins. They are predictive of higher possession. Ie, higher corsi. A higher Corsi DOES predict future goals or wins at a measurable level. So they've essentially given more context to the final output.

They also track other microstats that do not predict greater possession, but do predict higher expected goals. Slot chances, inner slot chances, passes into the inner slot, etc. Again, these stats do not individual predict greater future goals or wins, they contribute to higher expected goals within their own Expected Goals model. If you find SportLogiq's xGF model the most reliable, then these additions are valuable. If not, they're just more 'color' to the overall picture.

As far as your request for a greater sample, I can go back to 2008 (the year the NHL made on ice events trackable) and give you the CF% of the Final Four in each season. If you want a full regression analysis of every team, in every season going back to 2008, that's far more work.

 

[Golden_Jet]

Do you actually subscribe to Sportlogiq? Or subscribe to any analytics site that does pay for their data?

No, NHL teams are paying about 200k for it. Out of my price range lol.

Possession driving plays are plays that they have found to correlate to... wait for it... higher Corsi.

Possession driving plays include controlled exits, controlled entries, puck battles won, etc.

Corsi is shot attempts, we’ve had players say, they just direct a shot at net sometimes to get a higher corsi number.

As far as your request for a greater sample, I can go back to 2008 (the year the NHL made on ice events trackable) and give you the CF% of the Final Four in each season.

Ya sure, bigger the sample size the better.

If you want a full regression analysis of every team, in every season going back to 2008, that's far more work.

Ya don’t bother.

 

[JaegerDice]

No, NHL teams are paying about 200k for it. Out of my price range lol.

Corsi is shot attempts, we’ve had players say, they just direct a shot at net sometimes to get a higher corsi number.



Ya sure, bigger the sample size the better.


Ya don’t bother.

Ok, can I get a link to a player admitting this? Cause I feel like it would be a pretty big scandal for a player to come out and say 'yeah, I made a lesser play than I could have to boost a stat most fans have never heard of'.

I'll put together a quick rundown of the final four for each year since 2008.

 

[Do Make Say Think]

Dubas and Chayka are the two worst GM's of the last decade. Eye test matters the most. Using advanced data only works if you can use it as a tool to reinforce what a smart hockey mind is watching.

The reason Dubas and Chayka sucked isn't because of analytics though.

Dubas sucked because everybody told him he was a genius (when all he is is very lucky) and Chayka was just, straight-up, a total scumbag.

 

[Golden_Jet]

Ok, can I get a link to a player admitting this? Cause I feel like it would be a pretty big scandal for a player to come out and say 'yeah, I made a lesser play than I could have to boost a stat most fans have never heard of'.

I'll put together a quick rundown of the final four for each year since 2008.

I’ve only heard a few retired players say it, don’t think I’ve heard a current player say it.

 

[Gaylord Q Tinkledink]

So, what youre saying is teams that tend to carry the play more, because they have more talent/system go further in the playoffs ?

 

[dgibb10]

Last season the 2nd best and 3rd best teams met in the Final.

If there's a rematch this year, it will again be a meeting between the 2nd best and 3rd best team.

Crazy how predictive this stat continues to be.

Expected goals too

 

[dgibb10]

Corsi isn't meaningful. It's just one of the classic ways human beings over-complicate things. We take basic stats and keep adjusting them because we think it's better.

There's a lot of luck and randomness in the sport so why try to come up with all these advanced stat models that "in theory" are supposed to be better. I have news for you, it's a bunch of b.s. It's because younger generations think that everything in the past was wrong and needs to be changed. A combination of the eye-test and basic stats is all anyone needs.

And yes statistics has as much bias as one's own personal thoughts. There's almost nothing objective (free of bias) in this world. To think these advanced stats are better than the basic ones is completely delusional. You know what is objectively true though, advanced stats ARE more complicated.

If advanced stats were available back in the 90s I'm sure I could find a bunch of "proof" that Gretzky was an above-average player at best. Complete waste of time. You can't predict the weather, the stock market, and you can't predict sports games.

Corsi isn't complicated.

It's counting how many times a team attempts a shot.

 

[dgibb10]

So a perimeter team launching pucks into shin pads all night would be considered potentially an elite corsi team? I don't buy it.

What are these teams ranking by high danger chances for and against? I bet those are all top 5

If only we had a stat that was corsi but didn't include blocked shots (cough fenwick cough)

 

[Edmonton East]

Nah, people like yourselves that blindly cherry pick "advanced" stats without context when they support your argument are the problem and a detriment to the analytics community. The same type of people that fanboyed over the Chayka Coyotes for YEARS.

Let's take this silly thread as a great example. It's pathetically misleading. Sure, 3 of top 4 Corsi teams are in finals....who would have thought generally having more shot attempts is better than less??? Duh. But does it matter THAT much? Is it the end all be all? Is it that predictive? Nope.

The Canes are far and away the top Corsi team. They are a fraud and not remotely competitive. Is that pro your silly argument? How about that the #3 and #5 teams in Corsi lost in the 1st round? Or that the #6, 7, 8, 10, and 11 teams MISSED THE PLAYOFFS ALL TOGETHER??

This thread is dumb. Objective data my ass. Context is everything.

Going further, my VERY advanced calculations (because I'm sooooo smart) tell me a whopping 42% of the top 12 Corsi teams missed the playoffs entirely this season. And, again, the team that is far and away #1 is getting its teeth kicked in yet again. Plus the #29 team in Corsi was wayyyy more competitive vs the Panthers than the #1 team.

Last year: 33%
2022: 33% (Cup winner #22 in Corsi. ECF/WCF #3 in Corsi torched #1, #22 beat #10)

When do the Corsi bros award the Carolina Hurricanes their dynasty Cup this year? I wanna be at the ceremony. 3 straight years of leading the league in Corsi by a mile and they haven't even sniffed a Cup finals appearance. Corsi on its own doesn't mean jackshit. It simply provides a minimum threshold teams GENERALLY need to be above to have a shot at the Cup (unless you have an elite goalie outlier and/or get impact roster players late via acquisition or return from injury). And it's ridiculous people needed some "advanced" stat to tell them that generally having more shot attempts than the other team is better than less. What a revelation. Yet people pat themselves on the back like they are geniuses.

A way more interesting convo would be why certain teams' actual success consistently drastically underperform their Corsi ranking year after year after year. Looking at you Canes, my Devs, Pens, Flames, etc. I already have my answers, but they involve the sin of actually watching the teams play, understanding playoff hockey, and then supplementing that with statistical analysis. Hint: it isn't simply goaltending.

Nah, you and some other Corsi bros were doing the typical "eye test" mockery nonsense. As if those people are too stupid to understand big brain "analytics" like shot attempts. That's the problem. And this thread, including the title, framed the data in a way that was completely disingenuous and misleading...as I already proved. But people that are lazy and/or fake "smart" guys will buy it I guess.

It's adorable you still think you have the intellectual high ground here when you haven't refuted a single one of my statements. Thanks for proving my point(s). Again, this is exactly what's wrong with people that pretend to understand "analytics" and it gives the hockey analytics community as a whole a bad name (fairly so tbh). The arrogance at which some folks spew cherry picked stats or over glorify a single statistic without understanding the context or how it can be manipulated/misinterpreted is frustrating.

Still waiting for ya to explain why 42% of the top 12 Corsi teams missed the playoffs if it's SO PREDICTIVE. Or why the far and away #1 Corsi team pretty much never wins a Cup (or even rarely gets there) over the last decade. Or why the #29 team this year played wayyyy better vs Florida than the #1. Etc. Etc. Etc. Etc. But keep acting like it's some profound statement to say it's generally better to have more shot attempts than less. Again, duh.

Last 10 Cup winners reg season Corsi: 2, 22, 7, 9, 5, 10, 24, 16, 2, 2

So yea, again and for the last time because this is apparently an effort in futility: Corsi simply provides a minimum threshold teams (very) generally need to be above to have a shot at the Cup....which is pretty meaningless by itself because everyone knows you'd rather have more shot attempts than less, all things being equal. Example: I hand you NHL or NBA team A and team B over multiple seasons. All you can see is games played and shot attempts. Everyone picks the team with more shot attempts every time. This is common sense. But that doesn't mean the team you chose actually performed better overall....as literally shown by several examples I've pointed out as recently as this very season.

TL;DR - Corsi, like most stats, is pretty much meaningless by itself and it's ridiculous this is a thread. Folks masquerading as genius statisticians love to cherry pick / disingenuously present basic stats like this as if it's some profound, highly predictive big brain revelation....condescendingly mocking the "eye test" and actually watching games. It's annoying, but kinda funny at least.

Nah. Again, 2 issues:

1) The OP intentionally (or lazily) misled folks with the very title of this thread. Very convenient to ignore 42% of the top 12 Corsi teams didn't even make the playoffs. Or that far and away the #1 Corsi team is getting dog piled (AGAIN) by a team that nearly lost to the #29 team.
2) Folks like yourself keep acting like wanting more shot attempts than your opponent is like cracking nuclear fission.

Basic Corsi is some deep quantitative analysis??

Paging OP @JaegerDice since you are replying to folks. Enjoy the light reading above and defend your misrepresentation in the title and OP. As a reminder, you claim CORSI by itself is very predictive in winners and top playoff teams.

 

[dgibb10]

Going further, my VERY advanced calculations (because I'm sooooo smart) tell me a whopping 42% of the top 12 Corsi teams missed the playoffs entirely this season.

The top 12 corsi teams:

9/12 made playoffs

 

[dgibb10]

Corsi is shot attempts, we’ve had players say, they just direct a shot at net sometimes to get a higher corsi number.

Yes, hence why we can then look at the quality of shot attempts. It's called expected goals.

And any guys doing things to try and "cheat" these metrics, will be noticed once the sample size is large enough and they consistently underperform these metrics.

 

[Golden_Jet]

Expected goals tooView attachment 1041185

Sample size of one , is not useful, in predicting. Go back much further in time, to see if there is any correlation.

 

[Golden_Jet]

The top 12 corsi teams:

9/12 made playoffs
View attachment 1041190

The discussion I have seen everyone has been using 5 on 5.

 

[chaz4hockey]

One of these teams isn't like the others:

 

[dgibb10]

The discussion I have seen everyone has been using 5 on 5.

Why would I ignore the PP and PKs of teams.

On an individual player basis sure, some guys don't get PP time and some guys don't get PK time, it can skew stats.

But on a team level, why would I ignore it?

 

[dgibb10]

Sample size of one , is not useful, in predicting. Go back much further in time, to see if there is any correlation.

3/4 this year
3/4 last year
2/4 in 22-23
1/4 in 21-22
1/4 in 20-21 (of course weird since
2/4 in 19-20
2/4 in 18-19

About a 50% hit rate. Even ignoring the fact that 2 top 4 teams can get matched up earlier (eg 1st place Edmonton and 4th place LA this year, or 2nd place NJD and 1st place CAR in 22-23).

That implies a 71% chance of a top 4 team winning any given series (root of 0.5 needing 2 series to make the CF). Or 60% on a per game basis (more complicated binomial distribution stuff here Best of Seven Match )

Checks out as pretty accurate, understanding that these margins aren't massive, the teams they're up against are also generally quite good in expected goals because they are also playoff teams, and just overall volatility.


The big thing you need to evaluate separately is Goaltending of course, since expected goals does not involve goalie's (eg 2 Shesterkin carry jobs), and of course any injuries/deadline acquisitions that are not properly reflected in full season numbers.


Going too far back and you start getting much worse tracking data.

 

[Golden_Jet]

Why would I ignore the PP and PKs of teams.

On an individual player basis sure, some guys don't get PP time and some guys don't get PK time, it can skew stats.

But on a team level, why would I ignore it?

I was just saying from what I’ve read,

Regardless if 5 on 5 or all strengths, for last 2 years, it was the same 4 teams under either,
Not in any order,

Carolina
Edmonton
Florida
Colorado

Note: one team last year fell from 4th to 5th at 5 on 5. Otherwise apparently on top 4 really.

Would need to look further back in time, to see if any trend though.

 

[Golden_Jet]

3/4 this year
3/4 last year
2/4 in 22-23
1/4 in 21-22
1/4 in 20-21 (of course weird since
2/4 in 19-20
2/4 in 18-19

About a 50% hit rate. Even ignoring the fact that 2 top 4 teams can get matched up earlier (eg 1st place Edmonton and 4th place LA this year, or 2nd place NJD and 1st place CAR in 22-23).

That implies a 71% chance of a top 4 team winning any given series (root of 0.5 needing 2 series to make the CF). Or 60% on a per game basis (more complicated binomial distribution stuff here Best of Seven Match )

Checks out as pretty accurate, understanding that these margins aren't massive, the teams they're up against are also generally quite good in expected goals because they are also playoff teams, and just overall volatility.


The big thing you need to evaluate separately is Goaltending of course, since expected goals does not involve goalie's (eg 2 Shesterkin carry jobs), and of course any injuries/deadline acquisitions that are not properly reflected in full season numbers.


Going too far back and you start getting much worse tracking data.

Thanks for checking further back.

 

[Edmonton East]

Why would I ignore the PP and PKs of teams.

On an individual player basis sure, some guys don't get PP time and some guys don't get PK time, it can skew stats.

But on a team level, why would I ignore it?

The top 12 corsi teams:

9/12 made playoffs
View attachment 1041190

You really need to pause and research statistics (gain a better understanding) before commenting so confidently. This isn't the first time. You do drive-bys with blatantly incorrect data all the time. It's starting to almost look intentional at this point.

Like, in this very thread you are advocating for CORSI (which is SHOT ATTEMPTS) including special teams and non 5v5 play....I guess the good thing is all of us disagreeing in this thread on other things can probably unite on this issue.