6 of the bottom 10 teams in CF% 5on5 have a record above .500.

[Iapyi]

i put more emphasis on the WGAFAA% and even more on the BCHPH% these are far far more important then the CF%

 

[KingsFan7824]

You're aware the NHL counts points received during those "gimmicks" though...

Sure, but unless the remaining 20% is crazy horrible, 80% of a league can't be above .500, and they're not.

3v3 and SO's are random, and there just to get more goals in the game. The Kings had the most OT wins last year, and missed the playoffs by 10 points. Detroit was 9-0 in the SO, and didn't come close to the playoffs. If you want to look at which teams are at the traditional .500, you look at what they're doing in the larger 60 minute sample size. For the final standings, all the points get counted up.

 

[Zippy316]

You're starting to notice a bit of a trend away from CF% lately.

There's teams like Ottawa and Pittsburgh and the Rangers who have essentially been built around countering that style of play. Devils and Maple Leafs are headed that way as well.

 

[KingsFan7824]

Okay - I don't think the OP was referring to that, and don't really see how it's relevant to... anything.

I think kingsfan7824 was just trying to correct the OP.

Technically, 80%(or whatever high number) of the league is above .500. The real standings are hidden underneath the extra points teams get for winning after regulation. Well, not the real standings. The real standings are the real standings. Winning in regulation, OT, and the SO are all different though. The larger sample size of 60 minutes matters more in terms of perception, not so much reality, than 5 minutes or a few breakaways.

 

[NJ DevLolz]

NJ is playing for quality of opportunity, not quantity. With Schneider rounding back into form we "don't mind" giving up shots as much as last year, Cory is seeing and saving a lot.

We were riding a PDO wave at the beginning of the season. A lot of our bottom-6 was getting great puck luck. That seems to have regressed a bit. NJ has done most of their scoring in transition or from within 5 feet of the cage, very high percentage opportunities.

The whole quality vs quantity thing is actually true with New Jersey vs. teams who were actually lucky af like the Avs a few years ago and Carlyle's Leafs. NJ's high danger scoring chances for % is 53.4, good for 8th in the league.

 

[Bounces R Way]

I think a total time of possession stat and a 50/50 puck battle differential stat would both be more useful than CF%.

 

[Pookie]

You're starting to notice a bit of a trend away from CF% lately.

There's teams like Ottawa and Pittsburgh and the Rangers who have essentially been built around countering that style of play. Devils and Maple Leafs are headed that way as well.

This was always a trend though

31 teams don't play the same style, every game and every shift. Teams send scouts. They adapt. They change it up.

What is a trend is that people are finally waking up to that reality and knowledge that "keep away" isn't the only way to play this game.

 

[Pookie]

I think a total time of possession stat and a 50/50 puck battle differential stat would both be more useful than CF%.

What would total time of possession tell you though?

In a 60/40 game, which is a pretty good possession time ratio, one team has the puck for 36 mins. The other for 24.

This isn't a run out the clock game like soccer or football and you can get up and down then ice in under 10 seconds.

24 mins is a long time.

 

[Zippy316]

This was always a trend though

31 teams don't play the same style, every game and every shift. Teams send scouts. They adapt. They change it up.

What is a trend is that people are finally waking up to that reality and knowledge that "keep away" isn't the only way to play this game.

Back in the late 2000s and early 2010s, CF% was the trend though. It was a clear indicator of the good team and the bad teams because there were a lot of veteran laden teams who played structured, more old school hockey.

There's a lot more emphasis on speed and skill lately which is also bucking the trend.

 

[Pookie]

Um please prove that.

It's actually true.

What Corsi enthusiasts make up to counter that are score effects.

The theory that shots taken while trailing are less significant than shots taken while ahead. Shot are weighted equal if tied and more important if leading.

But if shots are shots, the team losing should be scoring more.

Score effects are nothing more than a fudge factor to keep the hypothesis alive.

 

[Pookie]

Back in the late 2000s and early 2010s, CF% was the trend though. It was a clear indicator of the good team and the bad teams because there were a lot of veteran laden teams who played structured, more old school hockey.

There's a lot more emphasis on speed and skill lately which is also bucking the trend.

I think you will find though that good teams tend to have good metrics. If you are counting shots that turn into goals, the top of the league likely leads that category as a function of actually being at the top of the league.

I don't think you can take an existing team and say, ok boys, go out there and possess that puck. Either you have talent or don't.

 

[Zippy316]

I think you will find though that good teams tend to have good metrics. If you are counting shots that turn into goals, the top of the league likely leads that category as a function of actually being at the top of the league.

I don't think you can take an existing team and say, ok boys, go out there and possess that puck. Either you have talent or don't.

That's true.

But I think there's little difference between a team like Carolina that dominates the puck but doesn't generate many high quality chances and the Leafs/Rangers who give up more shots but get quality chances with skilled players who can finish.

A team like the Lightning that are outscoring opponents with ease and still dominating stands out. Same goes with a team at the bottom of the standings with poor metrics. But a majority of teams are in that range where there is little separating them.

 

[Pookie]

That's true.

But I think there's little difference between a team like Carolina that dominates the puck but doesn't generate many high quality chances and the Leafs/Rangers who give up more shots but get quality chances with skilled players who can finish.

A team like the Lightning that are outscoring opponents with ease and still dominating stands out. Same goes with a team at the bottom of the standings with poor metrics. But a majority of teams are in that range where there is little separating them.

That's one thing missing in the metrics, is there an statistically significant difference in the range of data? When you ask for a t test, it goes silent.

I think there is very little difference across the spectrum with the exception of a few favourites.

I also think that the recognition of the fact that this game is not auto play is a good thing.

Shot quality metrics are promising though. As are player health metrics. My guess is that stuff will be proprietary and not published on blogs for us to discuss.

 

[Syckle78]

So both are wrong?

Sure if you want to act like hf is one singular opinion and not made up of thousands of different people.

 

[Zippy316]

That's one thing missing in the metrics, is there an statistically significant difference in the range of data? When you ask for a t test, it goes silent.

I think there is very little difference across the spectrum with the exception of a few favourites.

I also think that the recognition of the fact that this game is not auto play is a good thing.

Shot quality metrics are promising though. As are player health metrics. My guess is that stuff will be proprietary and not published on blogs for us to discuss.

I'd wager that almost every team is zeroing in on certain metrics they value.

There's shot data like Micah McCurdy has where you can map out shot locations and the type of shots. I'd imagine teams are investing heavily in that data as it shows you a ton more than Corsi ever would. Something like this:

 

[Pookie]

I'd wager that almost every team is zeroing in on certain metrics they value.

There's shot data like Micah McCurdy has where you can map out shot locations and the type of shots. I'd imagine teams are investing heavily in that data as it shows you a ton more than Corsi ever would. Something like this:

Indeed.

McCurdy interestly ran a study that looked at various Corsi and Fenwick measures and the ability to predict wins. The r2 was really low... like 18% and only for a handful of games in a season.

 

[Zippy316]

Indeed.

McCurdy interestly ran a study that looked at various Corsi and Fenwick measures and the ability to predict wins. The r2 was really low... like 18% and only for a handful of games in a season.

Do you have a link for that study or is behind his paywall?

 

[Pookie]

Do you have a link for that study or is behind his paywall?

Predictivity and wins is referenced near the end of the article

Adjusted Possession Measures

 

[SladeWilson23]

The whole quality vs quantity thing is actually true with New Jersey vs. teams who were actually lucky af like the Avs a few years ago and Carlyle's Leafs. NJ's high danger scoring chances for % is 53.4, good for 8th in the league.

This to me doesn't mean much. As I said, every team in the league averages between 10 and 13 high danger chances for and against per 60 minutes. There isn't much separation at all regarding high danger chances. The best way to create separation is through shot attempt volume. Always willing to concede shots from the outside is just not a good way to win games. Teams should be trying to prevent all shots, and not just the high danger ones.

 

[rent free]

215-198

That's the combined record of teams getting outshot in games. It's a winning percentage of .521 which to me doesn't say a whole lot.

Still higher than .500 no?

 

[bobholly39]

Before colorado's loss yesterday, only 6 teams across the whole NHL had a record below 500.

 

[SladeWilson23]

Still higher than .500 no?

Yea, but not enough to really make any conclusions.

 

[zeke]

Top 10 worst teams in CF% 5on5 (From worst to best):
1. Anaheim: 11-11-6 45.61%
2. Minnesota: 13-11-3 46.54%
3. New Jersey: 16-7-4 46.55%
4. Arizona: 7-18-5 47.03%
5. NY Rangers: 15-10-2 47.76%
6. Colorado: 12-12-2 47.81%
7. Vancouver: 14-10-4 47.91%
8. Ottawa: 9-10-6 47.98%
9. Washington: 16-11-1 47.99%
10. Nashville: 18-7-3 48.42%

In fact, New Jersey, Washington, Nashville, and the Rangers are all doing very well in their conference. What gives?

not sure what the question here is.

1. first you should be score adjusting those numbers.
2. secondly, and obviously, is that the answer to this is most usually goaltending and special teams.
3. thirdly, we look at sustainability, and figure that the teams high in the standings with poor possession will likely fall down going forward, eventually.
4.there are only 7 teams "below .500" in the league, so obviously a bunch of the bottom ten possession teams will be "over .500".