NHL Power Rankings

[djsFlames]

Ha. The team should get "unsustainable" printed on the side of their helmets just as a jab at all the critics that keep claiming some kind of fall off is imminent.

The only power rankings that carry any weight are the standings. That'll be evident to these wanna be experts when the Flames take their 43% corsi into a home playoff position in April.

[formulate]

Originally Posted by HotHotHeat

Except that Corsi isn't talked about as 'simply another stat'. It's all these ranking guys want to talk about.

I'm just going to start talking about intangibles whenever Corsi is brought up.

So then stop reading them. What does taking a staunch opposite view on the statistics accomplish? The truth is where it almost always is: somewhere in the middle.

I too wish bloggers would stop relying so heavily on possession metrics; perhaps there would be less strong opposition to them. I will continue to see these stats as I always have: a part of the game (like special teams or team speed) that contributes to a successful team.

[Hells Bells]

Drance is a tool.

How anyone can rank the Coyotes, Flyers or Stars ahead of the Flames in any way, shape or form right now is insane. Obviously other questionable ones ahead of the Flames but those three should put you out of a job.

Honestly, does he watch hockey?

[CanadaMatt]

Originally Posted by djsFlames

At last! We made the Super 16.

http://www.nhl.com/ice/news.htm?id=741373&cmpid=nhl-fb

Took them long enough to recognize, them haters.

I'll be the first to admit that I have not paid much attention to The Wild this year...have they gone on a 12 game winning streak to justfiy the 6th place ranking at 12-9-0?

It sticks out a little that they are 3 games above .500 and yet ranked higher than a team 10 games above .500 (NYI and Ducks)

...anyone?

[djsFlames]

Originally Posted by CanadaMatt

I'll be the first to admit that I have not paid much attention to The Wild this year...have they gone on a 12 game winning streak to justfiy the 6th place ranking at 12-9-0

It sticks out a little that they are 3 games above .500 and yet ranked higher than a team 10 games above .500

...anyone?

Suter (who NHL writers have pegged ahead of Gio for Norris contention....)
Kuemper and the goaltending - they've only allowed 55 goals.

Otherwise, I got nothing. That team isn't doing anything special.

Just going to maintain that those rankings are a joke.

[Blaster86]

Originally Posted by djsFlames

Suter (who NHL writers have pegged ahead of Gio for Norris contention....)
Kuemper and the goaltending - they've only allowed 55 goals.

Otherwise, I got nothing. That team isn't doing anything special.

Just going to maintain that those rankings are a joke.

More or less it. Minnesota are a bubble team much the same as last year. They're also like the Canucks in that when they lose they tend to do it in magnificent fashion (blowing big leads or getting absolutely shelled).

[Hemi-Cuda]

Originally Posted by StrykerSteve

I'll help you out.

What the hell is a Corsi For percentage? Stats like faceoff percentage, possession time, goal differential, plus/minus, all are easily understood just from their names. But WTF is a corsi or a fenwick? They sound so stupid just saying them, and whenever I see them mentioned in any article I immediately give it a pass

[troutman]

We got to fight the powers that be
Lemme hear you say
Fight the power

[Enoch Root]

Originally Posted by formulate

Yes, they do, in pretty much every situation. They also have a better PK than us and are better in the faceoff circle. Does that mean you will now disregard PK% and Faceoff win% as statistics?

It's simply another statistic to keep track of, with a shown correlation to success in the win column. Like special teams or faceoffs, having worse numbers doesn't mean you can't have a good team.

Correct.

The problem is that too many people want to draw conclusions from it (and it alone).

What should happen is a review of several stats to see if they are suggesting the same thing - then a conclusion.

But for some reason, people want to put tremendous stock in Corsi and/or Fenwick. Even though with the case of the Flames, goal differential is saying pretty much the exact opposite. (and goal differential is more strongly correlated with, and more in line with, their record).

[troutman]

Is the data for Corsi/Fenwick etc gathered reliably from game to game, from arena to arena?

[Textcritic]

Originally Posted by troutman

Is the data for Corsi/Fenwick etc gathered reliably from game to game, from arena to arena?

I'm sure there are discrepancies, but not to the extent that the figures are in any way invalidated. The margin of error is likely in the factor of tenths or hundredths of percentage points—nothing at all to make much of any difference.

[Calgary4LIfe]

What I would be more interested in is scoring chances. Flames have indeed been outplayed in some games and won, but they have also outplayed other teams and lost.

Didn't Kent Wilson used to track scoring chances for and against? I would much rather see those numbers. I bet they correlate better with the 'eye test'.

[PsYcNeT]

I haven't seen anything in depth, but I've heard that shot totals tallied by (non-professional) viewers or spectators at games can be quite varied from the "official" totals you see published.

Not to mention, "scoring chances" is one of the most subjective stats ever created.

[troutman]

Originally Posted by Textcritic

I'm sure there are discrepancies, but not to the extent that the figures are in any way invalidated. The margin of error is likely in the factor of tenths or hundredths of percentage points—nothing at all to make much of any difference.

Thanks.

I was thinking shots are counted differently from arena to arena. IIRC, DET is known for having super inflated shot totals.

[EldrickOnIce]

Originally Posted by PsYcNeT

I haven't seen anything in depth, but I've heard that shot totals tallied by (non-professional) viewers or spectators at games can be quite varied from the "official" totals you see published.

Not to mention, "scoring chances" is one of the most subjective stats ever created.

Corsi is probably the easiest count ever. While shots on goal can be subjective, shots attempts at goal is pretty easy.
That said, I do not know if there is any specific criteria. For example, when you ice the puck, is that a corsi for event?

[PsYcNeT]

Originally Posted by EldrickOnIce

Corsi is probably the easiest count ever. While shots on goal can be subjective, shots attempts at goal is pretty easy.
That said, I do not know if there is any specific criteria. For example, when you ice the puck, is that a corsi for event?

Sorry I should have been more clear.

My main gripe is with PDO, because of how unreliable shot tracking is.

[GGG]

Originally Posted by troutman

Is the data for Corsi/Fenwick etc gathered reliably from game to game, from arena to arena?

I would say yes as both Corsei and Fenwick are just using shots attempted and shots on goal as proxies for posession.

[GGG]

Originally Posted by PsYcNeT

Sorry I should have been more clear.

My main gripe is with PDO, because of how unreliable shot tracking is.

PDO is good if you change the regression from PDO regresses to 1 to PDO regresses to a range of .99 to 1.01 and each year there will be teams at around 1.02.

The flames 1.025 is unlikely to be sustained over the course of a season and certainly won't be sustained year over year. Just go through the old PDO numbers and you find usually there is 1 team in the 1.02 range at the end of the season. Occasionally there is an outlier at the 1.025 but that team doesn't repeat that performance.

So you don't actually need to measure shot quality to make the statement that the flames 11.2% shooting is unsustainable. Now it might be if you believe that this flames team is historically good and generating shot quality.

[TheDebaser]

I would like to see someone carry those ideas from that goaltending article and run with them. I'd like to see what percentage of shots the flames are creating where the goalie isn't set or is screened, and I'd like to see what percentage of shots the flames are allowing where the goalie isn't set or screened.

[J epworth]

Originally Posted by Calgary4LIfe

What I would be more interested in is scoring chances. Flames have indeed been outplayed in some games and won, but they have also outplayed other teams and lost.

Didn't Kent Wilson used to track scoring chances for and against? I would much rather see those numbers. I bet they correlate better with the 'eye test'.

Flames nation still does, Ryan Pike just had a good article today there looking at how scoring chances occur. He has starting tracking zone entries, looking at how the puck entered the offensive zone during the game (either carry-in or dump). He has found a trend that:

Calgary converts carry-ins to scoring chances at a rate of about 27.1%, while dump-ins convert at about 8.1% - indicating that carrying in the puck is more than THREE TIMES more likely to result in a scoring chance than a dump-in.

That's over a data size of

Over the first 25 games of the 2014-15 season, the Calgary Flames have generated 359 even-strength scoring chances. (For the curious, that's about 14 chances per game.)

Well, most of their entries are dump-ins - 1,296 even-strength dump-ins versus 800 even-strength carries.

It's data like this which Burke talks about when he says we do lots of in-house analytics to help get a larger picture of how goals occur and who is being successful in scoring for the Flames. Not just straight up counting stats, but using stats to find trends in how goals occur, what is successful.

Here's the whole article, it's a good read:
http://flamesnation.ca/2014/12/1/whe...nces-come-from