Curious Statistical Case of Carolina Hurricanes

[lomiller1]

Its been 4 years.

In a row.

Carolina’s sh% going back to the start of the 2015 season is 7.08%. The median sh% is 7.51% so there is nothing particularly odd about the Hurricanes. There are in the bottom 5 in the NHL but someone has to be at the bottom. Yes they have a lack of shooting talent, but that won’t cause them to have a sh% 3 points below a typical NHL team.


The biggest thing aa teams sh% tells you are the Jack Adams candidates for the following year because those teams almost always take a big jump in the standings the following year. Take a look at the 3 teams with a sh% below 7 last year and their projected point totals this year.

Buffalo +39 points
Montreal +22 points
Calgary +24 points

The thee worst sh% from 2016/17 and their change in points last year
Colorado +47
LA +12
NJ +27
2015/16

2015/16 to 2016/17
Toronto +36
Buffalo -3
Chicago +6

2014/15 to 2015/16
Arizona +22
Carolina +15
Chicago +1

 

[Bounces R Way]

It really didnt come off that way considering you said "there are way too many variables to ever think you have hockey scientifically fully explained."

Not a single advanced stat advocate has stated that or even implied that with their responses in this thread. And considering the lack of warranted stimuli for such a comment, your initial statement comes off as having an ax to grind/having an agenda



And to your last point, every stat is fairly useless without context besides goals, which is why I always find it humorous when people completely disregard some of them.

Well I certainly didn't mean to come off that way. The post I originally quoted mentioned there might be concerns with the validity of xGF, how it might not account for quality the way we suppose it does. I agreed and stated why I agreed. Having a stat named expected goals for kind of implies that it does have hockey scientifically explained, what with scoring a goal being the entire point of the game and it supposing to have the answers on how much we can expect to score.

 

[Maukkis]

Yes, each attempt adds a given value based on likelihood of going in.

Aren't there models which account for the player's individual SH% based on their last ~400 shots and apply that to the equation too?

Not that it makes the stat perfect, but I think there is more to existing models than shot type and location. And there should be.

 

[Kamiccolo]

This was addressed after the 11/21/18 game against Toronto when the Maple Leafs broadcast team (and subsequently their fans) accused whomever counts shots at PNC of being a bit heavy handed. I took a look at our home shot count versus away shot count. Here's that data with updates from the games since then.

First, the home games:

Date	Opponent	CAR Shots	OPP Shots
10/4/18	NYI	46	20
10/7/18	NYR	40	24
10/9/18	VAN	33	25
10/20/18	COL	43	22
10/26/18	SJS	41	23
10/28/18	NYI	39	20
10/30/18	BOS	44	31
11/10/18	DET	52	32
11/12/18	CHI	40	35
11/17/18	CBJ	31	24
11/18/18	NJD	25	34
11/21/18	TOR	45	32
11/23/18	FLA	39	35
11/30/18	ANA	34	29
12/11/18	TOR	30	29

[TBODY] [/TBODY]

And away

Date	Opponent	CAR Shots	OPP Shots
10/5/18	CBJ	35	32
10/13/18	MIN	57	23
10/14/18	WPG	43	26
10/16/18	TBL	40	27
10/22/18	DET	39	21
11/2/18	ARI	51	25
11/3/18	VGK	34	25
11/6/18	STL	39	20
11/8/18	CHI	34	38
11/24/18	NYI	27	19
11/27/18	MTL	22	49
12/2/18	LAK	34	35
12/5/18	SJS	40	23
12/7/18	ANA	39	19

[TBODY] [/TBODY]

And now the averages:

CAR Average Shots (Home) - 38.8 shots/game
CAR Average Shots (Away) - 38.14 shots/game

CAR Home/Away Shot Differential - 0.66 shots

OPP Average Shots Allowed (@ PNC Arena) - 27.67 shots/game
OPP Average Shots Allowed (Away) - 27.28 shots/game

OPP Home/Away Shot Differential - 0.39 shots

It's been a long while since I've taken a stats course and calculated statistical significance, but I have a feeling a difference of 0.66 shots between their home and away performance isn't significant.

Let's just clarify this portion. The original comment was that it didn't FEEL like they had that many shots. Many agreed, a small subset made these types of comments and before the 2nd period even started multiple people had reviewed the tape and confirmed the numbers were accurate.

We're not some sort of one big unit of goo that is easy to hate. Many of us have different reactions and opinions. And in this case, it's a case of fans looking for the voice they want to hear (there is always a group of people saying SOMETHING) which is something that is easy to latch onto and make fun of.

 

[Machinehead]

Aren't there models which account for the player's individual SH% based on their last ~400 shots and apply that to the equation too?

Not that it makes the stat perfect, but I think there is more to existing models than shot type and location. And there should be.

Manny Elk does shooting metrics and according to his numbers, the Hurricanes are historically bad at shooting, so that makes some sense.

 

[Ahoy there]

They have an awful group of forwards, like maybe the worst center depth in the league, a pretty decent young defense, a new goalie semingly twice a week, a rookie coach that just thinks they need to grind grit and grapple their way out into scoring goals, an owner that just took over and decides to change everything because he could surely be better and know more about running a hockey team than the last guy, and a GM that traded two recent #5OV picks for Dougie Hamilton, Ferland, and a college prospect who I'm sure can't wait to jump into that mess and then also traded Jeff Skinner in the offseason going into his contact year for the equivalent of the change you might find in your car.

But ya with those SA differentials and poor shooting percentages, must just be bad luck

as much as I'd like to, I find nothing with which to argue against this analysis.

 

[Maukkis]

Manny Elk does shooting metrics and according to his numbers, the Hurricanes are historically bad at shooting, so that makes some sense.

That would make sense, seeing as the Canes have next to no shooting talent.

Since it is basically established that the best scoring talent can consistently outperform their xG, I checked which Canes have achieved that this year. It seems like only three of their regulars are doing that: Ferland, TVR, and Pesce. Only one forward as well... most other teams have at least a handful of forwards who are scoring way more than their xG value suggests.

Then again, I don't think that works both ways (as in that bad shooters would consistently underperform their xG). Anyway, I'm starting to think finishing ability is the issue here, and it might even be magnified due to a system that heavily promotes shot volume.

 

[Volica]

Peters team has similar shots inflation. Flames outshooting other team by large margin many times but too much perimeter non-quality shots.

It's why Calgary's also top 10 in scoring chances for, high danger scoring chances for and, oh yeah, 6th in GF/Game.
But whatever helps you sleep at night.

 

[Filthy Dangles]

Are they less "stupid" then making up random explanations based on made up scenarios you heard some talking head mention about some other team some time in the distant past? Because that’s what the alternative you are offering.

And yet you claim sample size and yet every year the Canes shoot well below their xPected numbers say. Sure shooting talent and finishing ability play a role but there's more to it. they're chances themselves are overstated. That heatmap can't discern how much time and space a shooter had, if a seam or goal line pass was made prior to the shot (which greatly increases Sh%) etc.

 

[Ahoy there]

An in-game commercial shown often during local coverage is a clip of Forsland announcing where he says "throw it at the goal; you never know!"

I find that to be a perfect summation of the Hurricanes shot selection.

 

[Revelation]

Wow it's almost like spamming muffins from the point on NHL butterfly goalies isn't as effective as being conservative with your shots and making them move side to side before ripping one on a gaping net.

 

[lomiller1]

And yet you claim sample size and yet every year the Canes shoot well below their xPected numbers say.

You seem to be trying to move the goalpost on "well below" because their sh% this year is nowhere near as bad as where they have been historically. Remember, I already presented actual numbers the last few years they have been shooting 7.1% when the league median was 7.5%. 0.4 percentage points below league median I can buy as poor finishing talent. This year they are shooting 3 full % points below the league median and 2 points below their own norms even though sh% are way up league wide.

 

[lomiller1]

Wow it's almost like spamming muffins from the point on NHL butterfly goalies isn't as effective as being conservative with your shots and making them move side to side before ripping one on a gaping net.

No team in the NHL tries to hang onto the puck for a perfect shot, because it doesn't work. It's bad stratigy on NHL ice. No one coaches it no one does it and generally speaking players that like to play that way get sent back to the KHL. The way teams generate offense in the NHL is to get the puck to the net and look for deflections and rebounds. Aslo, as I've pointed out multiple times Caroline doesn't "shoot muffins from the point" they are amoung the best teams in the NHL at generating shots from the high danger places on the ice.

 

[Revelation]

No team in the NHL tries to hang onto the puck for a perfect shot, because it doesn't work. It's bad stratigy on NHL ice. No one coaches it no one does it and generally speaking players that like to play that way get sent back to the KHL. The way teams generate offense in the NHL is to get the puck to the net and look for deflections and rebounds. Aslo, as I've pointed out multiple times Caroline doesn't "shoot muffins from the point" they are amoung the best teams in the NHL at generating shots from the high danger places on the ice.

I didn't realize Kuznetsov put up 32 NHL playoff points from the KHL last year, makes that even more impressive.

The Caps won the cup last year by giving up tons of stoppable shots to keep Holtby in rhythm while scoring on breakaways, one timers and plays that took the goalie out of the picture. Full throated puck possession works if you're the prime Red Wings or Blackhawks with half your roster consisting of top end players, otherwise it's largely reduced to the puck constantly being intercepted by heady players and ending in the back of your net since most NHL goals now are from counterattacks on breakdowns.

 

[pabst blue ribbon]

	Top 10 5v5 Goal Differential teams with >50% xgf	Top 10 5v5 Goal Differential teams with <50% xgf	Bottom 10 5v5Goal Differential teams >50% xgf	Bottom 10 5v5 Goal Differential teams with <50% xgf
17/18	7	3	3	7
16/17	9	1	2	8
15/16	7	3	2	8
14/15	8	2	2	8
13/14	9	1	2	8
12/13	8	2	3	7
11/12	7	3	0	10
10/11	8	2	3	7
09/10	9	1	3	7
08/09	8	2	2	8
07/08	7	3	1	9

[TBODY] [/TBODY]

For all the crap that expected goals is getting in this thread, it seems to positively corelate with actual results (goal differential)

Every season at least 7 of the Top 10 teams who finish top 10 in Goal differential also have an xg above 50%
At most 3 teams that are in the bottom of Goal differential manage to have an xg above 50%