Speculation: 2020-21 News/Rumors/Roster Thread Part III

[tbrown33]

No human being is watching every minute of every game happening in existence. This isn't a competition about who can watch the most hockey (and if you didn't, then tough shit, "that's on you" or whatever). Stats simply quantify events that happened in a game. Why is this even a discussion? How do you get upset at numbers. So weird.

 

[King'sPawn]

Its the same pointless waste of time as it was when they started this nonsense. Its simply collected numbers of already available information. If any of it tells you something that you didn't already know, that's on you.

Expected goals for and against is the single stupidest concept in sports. It tells you absolutely nothing about how the game actually works. It tells you who stood where when the inquantifiable acts occur. So an so is always in a good position to succeed. Yeah, I already know that. This one guy has a hard time getting the puck across the blueline while retaining possession. Yeah, I already know that too.

The most valuable piece of advice a viewer can receive is to stop following the puck, expand your view of the ice, and look to see who does what as plays develop. If you can do that, you will never need a number again and you will know what is about to happen and who is in position to profit or suffer from it. Its just that simple, and no data will ever match it.

You continue to dismiss analytics and look down on people who use it, putting them in the same group as people who mis-use it.

A very small sample of data by itself is useless. Yes.

But as more samples get gathered, it tells a story that either corresponds or disagrees with observation.

It's when there's a disagreement between observation and analytics that we need to examine "why?"

There are uses and misuses of analytics, just as there are uses and misuses of biased observers who focus solely on the "eye test" and fool themselves into thinking they are impartial.

 

[bland]

You continue to dismiss analytics and look down on people who use it, putting them in the same group as people who mis-use it.

A very small sample of data by itself is useless. Yes.

But as more samples get gathered, it tells a story that either corresponds or disagrees with observation.

It's when there's a disagreement between observation and analytics that we need to examine "why?"

There are uses and misuses of analytics, just as there are uses and misuses of biased observers who focus solely on the "eye test" and fool themselves into thinking they are impartial.

Here is what I ask you to consider then: If the numbers have to be interpreted to have value, you are simply transposing the opinion of the interpreter from the physical game to the numerical data. Either way, it is still opinion, and the nature of the data is simply more flawed due to its specificity - one must acknowledge that the data is telling only that which it is collecting, and in order to have value it is being compared and contrasted with other data and other numbers before reaching any discernable conclusions.

I do take a lot of issue with it because there are very few specificities in hockey, and using a single piece of data to form an opinion is inherently flawed. I believe it creates more misunderstandings than it reveals truths.

Does anybody here really need a number to inform them that Matt Roy is the least likely defenseman to be scored on while Kurtis MacDermid is the most likely?

Then what do you do with that data? Decide to give Roy more, harder shifts than MacDermid? That's been a part and parcel of the game for a hundred years without data to tell you.

Now I completely understand that it helps NFL offensive coordinators to know the success ratios of crossing patterns against specific coverages. But hockey is a much simpler sport with fewer personnel options and fewer moments of specialty play.

For instance, what real value is there in knowing that success a player has in.zone entries without cross-referencing the success/failure ratio of the opponent on the ice at the time? Even if you could get that data you would need to factor in the length of each players shift at the time, the time of the game, the momentum at the time, so many other factors just to undetstand it fully. There are no advanced stats available at this time that differentiate circumstances. Numbers cannot do that, they just combine all of them into one jumbled number that is really only marginally different from one player to the next. And from team to team? Forget it, its worthless.

 

[bland]

While I agree with the bolded portion of your post, I'd say your tone and dismissive language reveals your own bias and I may suggest that not everybody can be as astute or insightful as you fancy yourself. I frequently find myself agreeing with many of your posts but stats have their value to anyone that can't be the perfect viewer or evaluator of talent that you seem to fancy yourself.

Every stat should be viewed in context of other stats and compared against and with the "eye test" but humans have bias. The things you may be watching for may cause you to miss or dismiss the things you're NOT looking for and to suggest that those biases don't exist is naive.

PART of my huge issue with things like xG is that they are not simple "counting stats" and the idea that there may be a bias in the construction of the equation is never discussed. The notion that "equation stats" are objective is, to me, the real flaw.

Corsi is simple. It's just shot attempts and as we know from criticisms of SOG, you can heavily out shoot a team but if they're all from the perimeter and low danger areas it doesn't really mean you outplayed them.

Corsi is just a clearer expression of Shots On Goal and can reveal information that you may not have been aware of if all you're looking for is "how many goals did Player X score"

The perfect hockey watching machine may not need the added benefit of statistics but so very few of us are perfect.

EDIT : I just realized I should have read this back to myself before posting cus I used "fancy yourself" twice in a row and jesus it sounds obnoxious.

But Doc, I DO fancy myself!

None of this would matter at all if it wasn't for the fact that fans, reporters and now even the broadcasts are misusing data to imply that things are rosier or just plain different from reality. It leads to people choosing to believe something false because a number on a page says so without context, and I could argue for days how that is a plague on our society, from sports to entertainment to sociology to politics.

 

[All The Kings Men]

But Doc, I DO fancy myself!

None of this would matter at all if it wasn't for the fact that fans, reporters and now even the broadcasts are misusing data to imply that things are rosier or just plain different from reality. It leads to people choosing to believe something false because a number on a page says so without context, and I could argue for days how that is a plague on our society, from sports to entertainment to sociology to politics.

Well on THAT we agree... but it seems to me the flaw there is with the user not with the tool and when you figure out a solution to that particular problem please don't bother posting it on a hockey message board. Spread it to people that actually matter!

BUT... to your point about misusing stats to paint an inaccurate picture... during the last game against the Blues I heard plenty of talk about the SOG being representative of the ice being tilted towards the Blues. That's not a "fancy" stat... it's just a stat that, in my opinion, has a subjective standard applied to its definition... and wait...

Now I'm lost as to which side of the issue I'm on so I'll bow out again...

 

[Raccoon Jesus]

But Doc, I DO fancy myself!

None of this would matter at all if it wasn't for the fact that fans, reporters and now even the broadcasts are misusing data to imply that things are rosier or just plain different from reality. It leads to people choosing to believe something false because a number on a page says so without context, and I could argue for days how that is a plague on our society, from sports to entertainment to sociology to politics.

No, that's actually the exact opposite of what's been explained repeatedly. Context is the most valuable part of that. Yes, there are some that blow it and rely entirely on the numbers, and there are unfortunately several national 'pundits' that get hung up there, but it's pretty insulting to people who watch 82-or-whatever-the-season-du-jour-games are of the Kings to suggest that's what we're hanging our hats on.

But bluntly, this also explains why you're so stubborn about your player views. Nothing wrong with that, I appreciate that you're so steady and steadfast on what you believe--but you cling to the idea that we're just looking at spreadsheets and ignoring the game. Frankly, all I do is watch the game and see after what the stats say to see if it supports or doesn't at all support what I'm seeing then ask 'why.' It doesn't mean I need stats to develop a viewpoint, but it sure helps when I can look at, as you say, Matt Roy, and explain via gameplay footage and stats why he was effective and the outcome of that effectiveness. Similarly, if someone is getting caved in game after game but not getting scored on, CF% would suggest it's probably coming. IE--watch tonight. If our 2nd line can't pull their heads out, they'll be the liability that lets in goals.

 

[All The Kings Men]

Similarly, if someone is getting caved in game after game but not getting scored on, CF% would suggest it's probably coming. IE--watch tonight. If our 2nd line can't pull their heads out, they'll be the liability that lets in goals.

Thank god I'm not the only one...

 

[Raccoon Jesus]

Thank god I'm not the only one...

Gonna quote you because I was going to edit it but you beat me to it

I was going to add I'm sure what @bland will say--and I agree with--you don't need a stat to see that the 2nd line was having issues. But it helps to understand to what degree! I mean, if the Kopitar line is nearly 80% CF and the Carter line is 20%--that's an awful disparity that I didn't catch the degree with with my eye--especially because (I think you pointed this out on twitter) the shots on goal weren't matching the game flow because the Kings kept firing attempts into every player and everywhere but on net.

I don't see what's so problematic about looking at that for more descriptive depth.

 

[All The Kings Men]

Gonna quote you because I was going to edit it but you beat me to it

I was going to add I'm sure what @bland will say--and I agree with--you don't need a stat to see that the 2nd line was having issues. But it helps to understand to what degree! I mean, if the Kopitar line is nearly 80% CF and the Carter line is 20%--that's an awful disparity that I didn't catch the degree with with my eye--especially because (I think you pointed this out on twitter) the shots on goal weren't matching the game flow because the Kings kept firing attempts into every player and everywhere but on net.

I don't see what's so problematic about looking at that for more descriptive depth.

Not to mention hitting the post a few times... which STILL doesn't make sense to me. Why isn't that a shot on goal but a puck that deflects off of like five bodies before winding up in the net is? Or is it? I've never found definitive answers about that.

 

[Raccoon Jesus]

Not to mention hitting the post a few times... which STILL doesn't make sense to me. Why isn't that a shot on goal but a puck that deflects off of like five bodies before winding up in the net is? Or is it? I've never found definitive answers about that.

My rudimentary understanding is it's 'puck that will end up in net if goalie doesn't save it', shots used more to calculate saves than anything? spitballing

 

[Statto]

While I agree with the bolded portion of your post, I'd say your tone and dismissive language reveals your own bias and I may suggest that not everybody can be as astute or insightful as you fancy yourself. I frequently find myself agreeing with many of your posts but stats have their value to anyone that can't be the perfect viewer or evaluator of talent that you seem to fancy yourself.

Every stat should be viewed in context of other stats and compared against and with the "eye test" but humans have bias. The things you may be watching for may cause you to miss or dismiss the things you're NOT looking for and to suggest that those biases don't exist is naive.

PART of my huge issue with things like xG is that they are not simple "counting stats" and the idea that there may be a bias in the construction of the equation is never discussed. The notion that "equation stats" are objective is, to me, the real flaw.

Corsi is simple. It's just shot attempts and as we know from criticisms of SOG, you can heavily out shoot a team but if they're all from the perimeter and low danger areas it doesn't really mean you outplayed them.

Corsi is just a clearer expression of Shots On Goal and can reveal information that you may not have been aware of if all you're looking for is "how many goals did Player X score"

The perfect hockey watching machine may not need the added benefit of statistics but so very few of us are perfect.

EDIT : I just realized I should have read this back to myself before posting cus I used "fancy yourself" twice in a row and jesus it sounds obnoxious.

It’s also, as usual, about using the stats properly. xG stuff is built to be used over large sample sizes and the people use it to exam a single game, or even half a dozen games... they then wonder why it frequently doesn’t reflect what they just watched... shocker! It’s the same as plus/minus it’s a great stat IF you use it properly but I can count on one hand the number of times I’ve seen it used proper when discussing on an Internet Forum. Btw, I’m completely with RJ you use everything available to build up the big picture of performance, there is no single approach that can do it on its own.

 

[Statto]

My rudimentary understanding is it's 'puck that will end up in net if goalie doesn't save it', shots used more to calculate saves than anything? spitballing

Yep, for s%... nothing more.

 

[All The Kings Men]

It’s also, as usual, about using the stats properly. xG stuff is built to be used over large sample sizes and the people use it to exam a single game, or even half a dozen games... they then wonder why it frequently doesn’t reflect what they just watched... shocker! It’s the same as plus/minus it’s a great stat IF you use it properly but I can count on one hand the number of times I’ve seen it used proper when discussing on an Internet Forum. Btw, I’m completely with RJ you use everything available to build up the big picture of performance, there is no single approach that can do it on its own.

I get that... but if I use it to compare specific games and they ALL don't add up to much in my opinion than why would I pay attention to the aggregate? Corsi I get... if you outshoot your opponent chances are more times than not you'll be the btter team and from what I can recall Corsi is like a 55%-60% accurate in predicting future success so I take it with the grain of salt it requires.

xG.... I just don't know how I feel about it yet. It's based on equations and not every site or analyst used to same equation so.... it's clearly subjective.

 

[All The Kings Men]

My rudimentary understanding is it's 'puck that will end up in net if goalie doesn't save it', shots used more to calculate saves than anything? spitballing

That's the definition I've just always had a problem with it. To me a shot that hits the post should count just as much as a shot that may have sailed an inch wide or hit a post but we'll never know because the goalie made a huge windmill save and the person counting the stat is sitting somewhere that totally skews their vision and judgement.

Shots on goal vs shot attempts to me is a fairly silly distinction in a sport where pucks hit players, sticks and divots on the ice to become goals or missed shots all the time.

 

[Statto]

Not to mention hitting the post a few times... which STILL doesn't make sense to me. Why isn't that a shot on goal but a puck that deflects off of like five bodies before winding up in the net is? Or is it? I've never found definitive answers about that.

In terms of assessing play, shots on goal is not the best. So when I was working with a pro team in the early to 2000’s the coach wanted to know the number of scoring chances for each team at any given moment rather than the number of shots. That way if it helped him understand how well the system is being executed.

 

[Statto]

I get that... but if I use it to compare specific games and they ALL don't add up to much in my opinion than why would I pay attention to the aggregate? Corsi I get... if you outshoot your opponent chances are more times than not you'll be the btter team and from what I can recall Corsi is like a 55%-60% accurate in predicting future success so I take it with the grain of salt it requires.

xG.... I just don't know how I feel about it yet. It's based on equations and not every site or analyst used to same equation so.... it's clearly subjective.

It’s a guide, a tool to assist predicting outcomes, nothing more. By the way, I’m not yet convinced it’s a good model for hockey.

Edit: My apologies if I sounded patronising

 

[All The Kings Men]

It’s a guide, a tool to assist predicting outcomes, nothing more. By the way, I’m not yet convinced it’s a good model for hockey.

Edit: My apologies if I sounded patronising

Not at all patronizing. I just meant I'm not convinced of it's value same as you.

 

[All The Kings Men]

In terms of assessing play, shots on goal is not the best. So when I was working with a pro team in the early to 2000’s the coach wanted to know the number of scoring chances for each team at any given moment rather than the number of shots. That way if it helped him understand how well the system is being executed.

That jives with what the article I linked to said about xG vs Corsi vs SC

 

[King'sPawn]

Here is what I ask you to consider then: If the numbers have to be interpreted to have value, you are simply transposing the opinion of the interpreter from the physical game to the numerical data. Either way, it is still opinion, and the nature of the data is simply more flawed due to its specificity - one must acknowledge that the data is telling only that which it is collecting, and in order to have value it is being compared and contrasted with other data and other numbers before reaching any discernable conclusions.

I do take a lot of issue with it because there are very few specificities in hockey, and using a single piece of data to form an opinion is inherently flawed. I believe it creates more misunderstandings than it reveals truths.

Does anybody here really need a number to inform them that Matt Roy is the least likely defenseman to be scored on while Kurtis MacDermid is the most likely?

Then what do you do with that data? Decide to give Roy more, harder shifts than MacDermid? That's been a part and parcel of the game for a hundred years without data to tell you.

Now I completely understand that it helps NFL offensive coordinators to know the success ratios of crossing patterns against specific coverages. But hockey is a much simpler sport with fewer personnel options and fewer moments of specialty play.

For instance, what real value is there in knowing that success a player has in.zone entries without cross-referencing the success/failure ratio of the opponent on the ice at the time? Even if you could get that data you would need to factor in the length of each players shift at the time, the time of the game, the momentum at the time, so many other factors just to undetstand it fully. There are no advanced stats available at this time that differentiate circumstances. Numbers cannot do that, they just combine all of them into one jumbled number that is really only marginally different from one player to the next. And from team to team? Forget it, its worthless.

You continue to oversimplify these anylytics and misrepresent it. It is still very young. We are still looking at how the numbers are being used, which will also lead to the growth of analytics.

If analytics say, at face value, that MacDermid was LEAST likely to have goals scored against him, instead of the most, then you have to ask yourself:
- are we misusing the data? Or
- are we misusing MacDermid?

Numbers alone are faulty as they can be absent of context. Eye test alone are also faulty as they can be misled by bias and human error of perception.

Anyone who looks at a shift and says "MacDermid had 2 expected goals for and 0 expected goals against that shift. He should be used more" is an idiot. Even in the span of a game.

But over the course of a season, we find Kurtis MacDermid had averaged better possession metrics than Sean Walker, even though Walker was expected to show more offense and was given more offensive opportunities, then the team has something to work with. Did they *REALLY* put Walker in a better position? Or are they weaker in 3rd pairing right side than expected? Or are they straight misusing their numbers?

There's a reason teams are actually using analytics and why some have actual departments specializing in it.

 

[Nickster]

Canucks looking like they might be really hurting for cash real soon:

 

[YP44]

Imama should not be getting NHL minutes, no matter who the Kings are playing

Keeps him from taking AHL minutes and fulfills the Kings need to have a certain number of taxi guys available.

 

[YP44]

Dallas sure is bad too haha

 

[Mats26]

So 20 goals scored so far leads the west, without Toffoli and Carts\Kopi with only 2. Finally getting some goals from up and down the lineup. I think our D only scored one, so our forward group has been producing, only going to get better when the more talented prospects make it in the lineup.

 

[Raccoon Jesus]

Re: advanced stats, as we were talking about yesterday, interesting note about blending the hockey side and the numbers side:

 

[bland]

Just so I don't have to quote everybody, could somebody please provide some instances where they believe advanced analytics have been properly applied and changed anything, big or small?

I see fans, reporters and broadcasters consistently misuse numbers to justify opinions without context, cannot recall a single instance where they have been applied in a positive manner.