Shots on goal/shot attempts statistics, overrated? (mainly in the playoffs)

[Artorius Horus T]

These days, its been for years now, teams take tons of shot attempts, shots on goal, its % hockey these days, advanced stats is the most popular thing.
= if you do this, % is higher to result on this and that

But is it all overrated?, especially in the playoffs


Teams that take tons of shots, have more shots on goal, etc. etc. gets beaten more than not in the playoffs these days,
because of, better overall team defending, goaltending and team system/tactics

I think teams trust too much on %' hockey, to advanced stats. Hard work, higher work ethic, team play, enough skill, speed and grit
beats % hockey, advanced stats darlings


Because, when playoffs comes, everything chances

I prefer quality over quantity

Click to read more...

 

[KevinRedkey]

These days, its been for years now, teams take tons of shot attempts, shots on goal, its % hockey these days, advanced stats is the most popular thing.
= if you do this, % is higher to result on this and that
But is it all overrated?, especially in the playoffs

Teams that take tons of shots, have more shots on goal, etc. etc. gets beaten more than not in the playoffs these days,
because of, better overall team defending, goaltending and team system/tactics

I think teams trust too much on %' hockey, to advanced stats. Hard work, higher work ethic, team play, enough skill, speed and grit
beats % hockey, advanced stats darlings


Because, when playoffs comes, everything chances

I prefer quality over quantity

Without any stats to back this up...

 

[SnowblindNYR]

Well there used to be a time when the analytics community pretended shot quality didn't exist. Thankfully they now take that into account. So yes quantity is less relevant. Also, it's 2024 not 2012 when the Kings spamming shots was a revolutionary and Stanley Cup winning tactic. You actually need to show some skill in the plays you make.

 

[Machinehead]

Teams that are winning the xG battle are losing like crazy in these playoffs.

The thing about the xG stats we use now is that you can still build them up by just shooting a lot.

We talk about the higher scoring league and the increased speed and skill. Perhaps the NHL is shifting more towards a quality meta.

The best ways of measuring quality vary from model to model and it's not something we've pinned down yet.

 

[ucanthanzalthetruth]

The Canes outshoot everyone which is incredibly misleading because yeah a lot of their shots are low probability, but even with that factored in, they have the puck so much they still usually play to a draw/marginally outplay their opponent. They have just gotten goalied/shitty luck vs Florida and NYR. Essentially losing 7 straight 50/50 games is just unfortunate. Ironically, having watched all 4 Florida games and all 4 NYR games, this was easily their worst game of those 8 and they won lol.

 

[Machinehead]

The Canes outshoot everyone which is incredibly misleading because yeah a lot of their shots are low probability, but even with that factored in, they have the puck so much they still usually play to a draw/marginally outplay their opponent. They have just gotten goalied/shitty luck vs Florida and NYR. Essentially losing 7 straight 50/50 games is just unfortunate. Ironically, having watched all 4 Florida games and all 4 NYR games, this was easily their worst game of those 8 and they won lol.

I was thinking the same thing.

I was like "man, that was our best game of the series."

The Rangers had a miserable 1st period and the Canes did their possession thing at the start of the 3rd. 30 of the last 40 minutes were just rush chance after rush chance by the Rangers.

Of course the Canes win this one.

That's the other thing, there's just a huge amount of variance in the playoffs in a sport with (relatively) not a ton of scoring compared to other sports, where the level of competition is high on both sides.

 

[SnowblindNYR]

The Canes outshoot everyone which is incredibly misleading because yeah a lot of their shots are low probability, but even with that factored in, they have the puck so much they still usually play to a draw/marginally outplay their opponent. They have just gotten goalied/shitty luck vs Florida and NYR. Essentially losing 7 straight 50/50 games is just unfortunate. Ironically, having watched all 4 Florida games and all 4 NYR games, this was easily their worst game of those 8 and they won lol.

I'd say it was the worst defensively of the 4. I still think game 1 was the worst overall for them because until it was 6 on 5 they didn't do shit offensively, save for one nice rush and a fluke goal. But yeah, the Rangers had a high level scoring chance it seemed every time they entered the zone. But the Canes produced more quality offensively than probably the other three games and Igor was off his game. Coupled with the Rangers not being able to finish and you get a win.

 

[kabidjan18]

I can't help but think your post is in response to my post celebrating the Stars winning while losing SOG, xG, Corsi, and HDCF.

There are countless variables and no stat does a good job capturing all of them. I've consistently felt that the biggest variable that those stats don't accurately capture is the battle in front of the net. Past a certain distance or from a certain angle, most goalies will stop most shots they see all the way. So the art of defending is keeping shots a combination of A. far, B. wide, C. visible. That's why the guys wrestling in front of the net are so important. They block shots from good angles. They try to box opposing forwards out of the line of sight of their goalie. And they try to push puck possession wide. Conversely, if you're a forward, you want to place a hug butt right in the face of the goalie. He won't see anything. And if the puck squirts out to you for a rebound shot it will be close and from a good angle.

Dallas did a good job on all of them. Oettinger was able to see most of the shots coming his way. The Stars themselves blocked 24 shots, almost as much as Oettinger saved. And they pushed the Avs puckhandlers wide. They also trapped well in the neutral zone in the 3rd period. So even though possession usually correlates to scoring, here it did not. Effective defensive play negates possession-heavy teams.

 

[ijuka]

These days, its been for years now, teams take tons of shot attempts, shots on goal, its % hockey these days, advanced stats is the most popular thing.
= if you do this, % is higher to result on this and that

But is it all overrated?, especially in the playoffs


Teams that take tons of shots, have more shots on goal, etc. etc. gets beaten more than not in the playoffs these days,
because of, better overall team defending, goaltending and team system/tactics

I think teams trust too much on %' hockey, to advanced stats. Hard work, higher work ethic, team play, enough skill, speed and grit
beats % hockey, advanced stats darlings


Because, when playoffs comes, everything chances

I prefer quality over quantity

I mean, it's less predictive than goals or expected goals, but it's still quite predictive of wins.

But does anyone think that shots or shot attempts are more predictive than goals or expected goals? I don't think so.

Hence, no, I don't think that it's overrated.

 

[Bounces R Way]

Personally think the publicly available xG models are trash in trash out. A 4th line plug shoveling a puck into a set goalie's pads at the side of the net from two feet away is worth 3x what a Ovechkin or Pastrnak bomb one timer from the circle.

If I walked into a hedge fund and claimed to have a model to predict futures on pork prices and they asked me what my average forecasting error was and I answered 38% they would likely laugh me out of the building. The data is simply not good enough.

 

[Grifter3511]

These days, its been for years now.

You lost me right there.

 

[HarrySPlinkett]

Analytics and playing the percentages are fine for the regular season and player evaluation.

Playoffs is about going through the men across from you.

 

[HairyKneel]

Analytics and playing the percentages are fine for the regular season and player evaluation.

Playoffs is about going through the men across from you.

Exactly. Stats, especially a lot of the fancy ones are for neckbeards. Watch the game, hell try and understand the game and if you aren’t a fat pig,,go for a skate and join a beer league. There is so much action and noise out there that even the sweaty neckbeards might understand more about playing at a rudimentary level. Advanced stets have a small place in hockey. It isn’t baseball.

 

[HolyHagelin]

I was thinking the same thing.

I was like "man, that was our best game of the series."

The Rangers had a miserable 1st period and the Canes did their possession thing at the start of the 3rd. 30 of the last 40 minutes were just rush chance after rush chance by the Rangers.

Of course the Canes win this one.

That's the other thing, there's just a huge amount of variance in the playoffs in a sport with (relatively) not a ton of scoring compared to other sports, where the level of competition is high on both sides.

A 4-3 game in hockey is basically the same as 28-21 in football. There isn’t less scoring it just counts differently.

 

[HolyHagelin]

I can't help but think your post is in response to my post celebrating the Stars winning while losing SOG, xG, Corsi, and HDCF.

There are countless variables and no stat does a good job capturing all of them. I've consistently felt that the biggest variable that those stats don't accurately capture is the battle in front of the net. Past a certain distance or from a certain angle, most goalies will stop most shots they see all the way. So the art of defending is keeping shots a combination of A. far, B. wide, C. visible. That's why the guys wrestling in front of the net are so important. They block shots from good angles. They try to box opposing forwards out of the line of sight of their goalie. And they try to push puck possession wide. Conversely, if you're a forward, you want to place a hug butt right in the face of the goalie. He won't see anything. And if the puck squirts out to you for a rebound shot it will be close and from a good angle.

Dallas did a good job on all of them. Oettinger was able to see most of the shots coming his way. The Stars themselves blocked 24 shots, almost as much as Oettinger saved. And they pushed the Avs puckhandlers wide. They also trapped well in the neutral zone in the 3rd period. So even though possession usually correlates to scoring, here it did not. Effective defensive play negates possession-heavy teams.

Ot only that, but baiting teams into less effective shots can frequently generate rush chances for your offense. You can see something like a few great shifts of possession in a row, get off 5-6 shots, then the d cleans up one of the rebounds and 1LW is streaking in alone on the goalie, bang. Happens all the time.

 

[Machinehead]

A 4-3 game in hockey is basically the same as 28-21 in football. There isn’t less scoring it just counts differently.

Football is also a high-variance sport, especially considering how few games there are.

 

[SoupNazi]

Without any stats to back this up...

Yeah, I want to see years worth of stats for this, not just one series or a couple games.

 

[HolyHagelin]

Football is also a high-variance sport, especially considering how few games there are.

In some ways yes in some ways no; most football teams have their best player making most of their offensive plays which reduces variance.

For context, i used to help create new proprietary predictive stats for a football gambling advice site, so i am pretty versed in football stats. I worked with scott kacsmar before he got super famous.

 

[Machinehead]

In some ways yes in some ways no; most football teams have their best player making most of their offensive plays which reduces variance.

For context, i used to help create new proprietary predictive stats for a football gambling advice site, so i am pretty versed in football stats. I worked with scott kacsmar before he got super famous.

This is true, the QB position does reduce variance a lot.

People will bring up goaltending, but there's just no comparison.

The best goaltender isn't that far ahead of like, the 15th best goaltender.

 

[Essenege]

For the most part good analytics teams fare well in the playoffs. At the end of the day I take the team who controls scoring chances 5v5 over the ones that don’t.

That being said, in a 7 games series, having a 55% xGF vs a team with a 52% xGF won’t give you a huge advantage. All the imponderables becomes much more important.

 

[BLONG7]

The Canes outshoot everyone which is incredibly misleading because yeah a lot of their shots are low probability, but even with that factored in, they have the puck so much they still usually play to a draw/marginally outplay their opponent. They have just gotten goalied/shitty luck vs Florida and NYR. Essentially losing 7 straight 50/50 games is just unfortunate. Ironically, having watched all 4 Florida games and all 4 NYR games, this was easily their worst game of those 8 and they won lol.

Canes have the puck all the time, hence the shots..................maybe their goalie gets out played the other teams goalie??
Food for thought...

Analytics and playing the percentages are fine for the regular season and player evaluation.

Playoffs is about going through the men across from you.

The old..............there are players who get you to the playoffs, and then there are players who get your through...........the playoffs.

Hard for stats to then, show everything going on out there, on the ice.
Hockey has a puck, that can produce lucky bounces, and OT wins etc

 

[BLONG7]

Analytics and playing the percentages are fine for the regular season and player evaluation.

Playoffs is about going through the men across from you.

The old..............there are players who get you to the playoffs, and then there are players who get your through...........the playoffs.

Hard for stats to then, show everything going on out there, on the ice.
Hockey has a puck, that can produce lucky bounces, and OT wins etc

 

[HolyHagelin]

This is true, the QB position does reduce variance a lot.

People will bring up goaltending, but there's just no comparison.

The best goaltender isn't that far ahead of like, the 15th best goaltender.

The offense also has control over how many plays the qb is in control of. Uptempo/grind it out, run heavy v pass heavy, etc.

In hockey, the other team decides how many shots your goalie faces.

 

[JaegerDice]

They're obviously less predictive over the course of a 7 game series than an 82 game season..

That said, even with that weakness taken into consideration, they're still better than the #willtowin, #heart, #grit, #snot nonsense narratives that the meatballs like to pedal.

Well there used to be a time when the analytics community pretended shot quality didn't exist. Thankfully they now take that into account. So yes quantity is less relevant. Also, it's 2024 not 2012 when the Kings spamming shots was a revolutionary and Stanley Cup winning tactic. You actually need to show some skill in the plays you make.

This wasn't a Kings invention, nor were they the only ones that did so. The 2008 and 2009 Red Wings, the 2010 and 2013 Blackhawks, the 2016 Penguins, and on down the list to the 2022 Avalanche all were top 5 in the league in shot attempts per game, shots on goal per game, etc.

More over, while their shot volume was telling, it was shot differential that spoke to their dominance.

They were able to fire-house the other team in shots because they were great at getting the puck back and putting shots on over and over, and then getting the puck away from the other guys quickly to repeat the process. Within the output of 'shots' is all the inputs like plays, battles, and decisions that allow you to keep the pressure on and diffuse pressure against quickly.

That will always be true, so long as hockey only has one puck on the ice.

 

[Mickey Marner]

Head-to-head match-ups produce much different results than the regular season and you need to account for score effects. At 5v5 against the Rangers, Washington had a 52.03 CF% and a 54.99 xGF%, which on the surface makes it seems like they out-played the Rangers. But New York actually had a better CF% and xGF% when the game was tied, when they were leading or when they were trailing. The Capitals were just trailing so much more often that their total shot analytics were positive.

	NYR CF%​	NYR XGF%​	NYR Minutes​	WSH CF%​	WSH XGF%​	WSH Minutes​
Leading​	42.92%​	35.86%​	119:30​	33.33%​	0%​	3:21​
Trailing​	66.67%​	100%​	3:21​	57.08%​	64.14%​	119:30​
Tied​	58.33%​	62.54%​	53:49​	41.67%​	37.46%​	53:49​
All Scores​	47.97%​	45.01%​	176:40​	52.03%​	54.99%​	176:40​