[PGT] Coyotes 4 Flames 1

[taxbuster]

Originally Posted by Bingo

They likely are already doing so, but I'd get the analytics team to look at the Bruins, Lightning, etc and the Flames for a random ten game (or more segment) and isolate each high danger chance the teams get and give. Then compile this data and look at why they are different, perhaps assigning a measure 1 (really freaking dangerous), 2 (kind of dangerous) and 3 (in the correct zone but not very dangerous).

But this isn't as simple as "perimeter team winning the corsi battle" because they have great scoring chance totals and great scoring chance differentials.

There is a consideration with that - even with the "really freaking dangerous" shots: where other players are positioned at the time of the shot.

Somehow one needs to look at each shot in several lights: how close was it to the net? where were the G and D positioned for the other team? where were the F supporting the shot? There are a LOT of shades of grey there -- a shot from close in (what they like to call the "high danger" area) is a lot less dangerous if the goalie is square to the puck, a D-man is beside him and your only F nearby is the guy taking the shot.

OTOH, if the F is supported by 2 other Fs right beside him it's a MUCH higher chance shot. Even a drifting light shot from the point is more dangerous if there is an F or two close in front of the net - deflections or rebounds are much more likely.

The Flames get guys in close fairly regularly - but only one at a time. There's just no puck support so rebounds are easily cleared.

A "dangerous" shot with a low "likelihood" of recovery, rebound or deflection is really just a shot.

*If* they evaluate this way, then it may show something. The heat maps sure don't some days - they're all over the ice shooting, but clearly there is no one in position to put the puck away.

[zamler]

Originally Posted by Bingo

If it is bad luck then act accordingly.

Given the sample size under Gulutzan we are way past back luck territory.

[Bingo]

Originally Posted by zamler

Given the sample size under Gulutzan we are way past back luck territory.

There's that quick judgement I was looking to avoid.

Thanks for the example!

[Locke]

Originally Posted by Bingo

There's that quick judgement I was looking to avoid.

Thanks for the example!

I'm not sure I follow. You're insinuating that two full seasons isnt enough of a sample size?

[CSharp]

Originally Posted by Realtor 1

Foo looked good although I want to see it happen on a consistent basis as many call ups have early signs of promise with less opportunity. Does have the raw talent though!!
GG won as he outshot the other team.
Dave Cameron’s PP operating around 2% since the trade deadline. Odds of having twins is 3%

This is what I say about meaningless games - give the rookies a shot. Kind of BS right now trying to protect their minutes. If no one is scoring on the team, try the new bodies out and prepare for the new season.

[Bingo]

Originally Posted by Locke

I'm not sure I follow. You're insinuating that two full seasons isnt enough of a sample size?

My original long post was about finding the mix with analysis and avoid making assumptions either way.

So to say "it's not luck!" was pretty much exactly what I think needs to be avoided.

Of course luck plays a role. So does the coaching/structure/system, individual player performances, etc.

[zamler]

Originally Posted by Bingo

There's that quick judgement I was looking to avoid.

Thanks for the example!

How is me watching every single game Gulutzan has coached quick judgement?

[Bingo]

Originally Posted by zamler

How is me watching every single game Gulutzan has coached quick judgement?

I think the Flames have had a very unlucky season, but that's not to say the coach shouldn't go.

It's a mix of factors, simply ignoring the role of one of the factors is what I'd be looking to avoid if I was Treliving.

[VladtheImpaler]

Originally Posted by Bingo

I think the Flames have had a very unlucky season, but that's not to say the coach shouldn't go.

It's a mix of factors, simply ignoring the role of one of the factors is what I'd be looking to avoid if I was Treliving.

What's "unlucky", Bingo? What is the statistical probability that all (or virtually all) relevant players are "unlucky"? Is not the far more logical explanation systemic? Even in my mathematical genius heyday 30 years ago, I never took/was interested in statistics, so someone else can hopefully put actual numbers to this... I assume, as with everything, there would be a bell curve, with some being far too "lucky" and some far too "unlucky" and the majority around the fat middle. If the results are all at one end, that would suggest you have the wrong "normal", given the style of play...

[zamler]

Originally Posted by Bingo

I think the Flames have had a very unlucky season, but that's not to say the coach shouldn't go.

It's a mix of factors, simply ignoring the role of one of the factors is what I'd be looking to avoid if I was Treliving.

You're still not explaining why I made a quick judgement, that's what you accused me off. You might think I am wrong about the bad luck factor, but that's not what you said.

[Ryan Coke]

I absolutely agree that luck plays a part, the line between winning and losing can be very fine.

But any statistical analysis where ‘luck’ is an accepted conclusion, really isn’t worth much.

I think the idea you had Bingo of looking at scoring chances and rating what makes more or less dangerous ones is great. Really it means we are looking for a more accurate statistical representation that has a much stronger correlation with winning, because it s becoming more evident all the time that Corsi and HDSC haven’t achieved a useful level of predictability.

It is just a stat, and is not inherently bad or good. But it is bad as an overall measurement of a teams ability to win or lose consistently, much like face off %, hit counts, and +/- .

And as Zamler says, 160+ games is a pretty generous sample size.

[kukkudo]

Originally Posted by Bingo

I think the Flames have had a very unlucky season, but that's not to say the coach shouldn't go.

It's a mix of factors, simply ignoring the role of one of the factors is what I'd be looking to avoid if I was Treliving.

The team stinked from coaching to the players. Nothing was unlucky any team can fire 40 garbage shots at a goalie and make the stats look all pretty. Oiler fans have been doing this for years to make it look like they got a good team.

[Bingo]

Originally Posted by VladtheImpaler

What's "unlucky", Bingo? What is the statistical probability that all (or virtually all) relevant players are "unlucky"? Is not the far more logical explanation systemic? Even in my mathematical genius heyday 30 years ago, I never took/was interested in statistics, so someone else can hopefully put actual numbers to this... I assume, as with everything, there would be a bell curve, with some being far too "lucky" and some far too "unlucky" and the majority around the fat middle. If the results are all at one end, that would suggest you have the wrong "normal", given the style of play...

Hartley's 14/15 team had a lot go it's way, there's little doubt.

This season felt a bit like the antitheses to that. I'm certainly not calling the whole thing luck, but it's not just systems too. Bad bounces, odd calls, and the one that really sticks out to me is leading the league in missing the net.

Missed shots now up to 1207, when the next closest team is 1120 is just odd as hell. Missing the net could be execution too, don't get me wrong, but it's certainly not structure.

Before the torches get lit, I'm not suggesting they retain Gulutzan, but I'd want the GM to look at some of the unlikeliness of some of the stats to repeat if everything was left as is as part of the analysis.

Too many players on this team are below their position's average in shooting percentage. Is that all on coaching?

I just wouldn't discount anything.

[Locke]

Originally Posted by Bingo

My original long post was about finding the mix with analysis and avoid making assumptions either way.

So to say "it's not luck!" was pretty much exactly what I think needs to be avoided.

Of course luck plays a role. So does the coaching/structure/system, individual player performances, etc.

Right, okay, but after 160+ games one would think that luck would eventually meander towards the mean and that sample size would indicate that conclusions can be drawn and they wouldnt be 'knee-jerk.'

After 2 full seasons and 160 games can one not conclude that it isnt 'luck?'

That would have to be immensely bad luck, almost systemically bad luck.

[zamler]

Missing the net more than any other team is not bad luck that's being bad at shooting the puck.

[Itse]

Originally Posted by taxbuster

There is a consideration with that - even with the "really freaking dangerous" shots: where other players are positioned at the time of the shot.

Somehow one needs to look at each shot in several lights: how close was it to the net? where were the G and D positioned for the other team? where were the F supporting the shot? There are a LOT of shades of grey there -- a shot from close in (what they like to call the "high danger" area) is a lot less dangerous if the goalie is square to the puck, a D-man is beside him and your only F nearby is the guy taking the shot.

OTOH, if the F is supported by 2 other Fs right beside him it's a MUCH higher chance shot. Even a drifting light shot from the point is more dangerous if there is an F or two close in front of the net - deflections or rebounds are much more likely.

The Flames get guys in close fairly regularly - but only one at a time. There's just no puck support so rebounds are easily cleared.

A "dangerous" shot with a low "likelihood" of recovery, rebound or deflection is really just a shot.

*If* they evaluate this way, then it may show something. The heat maps sure don't some days - they're all over the ice shooting, but clearly there is no one in position to put the puck away.

If I was to make an attempt at creating a "dangerous shot" chart, I would probably track something like "how many players from your team are directly involved", and by being involved I mean
a) being open for a pass in an alternate scoring position
b) screening the goalie
c) passed the puck within the last 1 second (2 if there's 2 passes)

This kinf of the stuff that makes things hard for a goalie. If there's another guy open, they often can't fully commit to the shot that's coming without risking a pass and a tap-in, plus it forces them to control the rebounds which is more difficult than just stopping the puck. If there's screen they can't see the puck. If the puck just moved from one player to another they had to quickly move to get in position again. Combine these three and you have a great scoring chance.

I'm not saying this is some perfect formula, but ultimately this kind of stuff is a lot more important than simply where the shot comes from.

Or, alternatively you might do a simple count: are there more your guys closer to the goalie than opposing players when the shot goes?

[Bingo]

Originally Posted by Locke

Right, okay, but after 160+ games one would think that luck would eventually meander towards the mean and that sample size would indicate that conclusions can be drawn and they wouldnt be 'knee-jerk.'

After 2 full seasons and 160 games can one not conclude that it isnt 'luck?'

That would have to be immensely bad luck, almost systemically bad luck.

Well that's an assumption that the whole thing has been constant for 160 games. It hasn't.

The Flames had different problems last year. Their powerplay was top 10, they won on home ice, their shooting percentage was mid pack.

Analytically they were middle of the pack in corsi and shots, and near the bottom in high danger splits.

The Flames were 22nd in terms of missed shots last year, now they're first.

This isn't a running plot for two years, the two seasons are completely different and unique.

[Bingo]

Originally Posted by zamler

Missing the net more than any other team is not bad luck that's being bad at shooting the puck.

Well then it's a one year sickness of being bad at shooting the puck because they didn't have that problem last year.

[Itse]

Originally Posted by Bingo

Well that's an assumption that the whole thing has been constant for 160 games. It hasn't.

The Flames had different problems last year. Their powerplay was top 10, they won on home ice, their shooting percentage was mid pack.

Analytically they were middle of the pack in corsi and shots, and near the bottom in high danger splits.

The Flames were 22nd in terms of missed shots last year, now they're first.

This isn't a running plot for two years, the two seasons are completely different and unique.

They also changed their powerplay tactics and made lots of other little changes. From an analytics perspective this should offer a good comparison point for what works and what doesn't, since so many parts of the team are the same.

That's not what I've seen anyone do with stats in hockey. If Gulutzan has been doing that, he clearly can't draw any useful conclusions out of the stats, which means he's not good at using stats or the stats don't really help.

[redforever]

Originally Posted by Locke

Right, okay, but after 160+ games one would think that luck would eventually meander towards the mean and that sample size would indicate that conclusions can be drawn and they wouldnt be 'knee-jerk.'

After 2 full seasons and 160 games can one not conclude that it isnt 'luck?'

That would have to be immensely bad luck, almost systemically bad luck.

This this this. Stats should now be pointing to the results being the norm with this coach and his system.