News and Blog Report: Oilers are terrible when Taylor Hall is on the ice

[Mcnofool6110]

But we suck all the time...
I reckon the Oilers would be legendarily bad (no games won) without Hall...

 

[Ol' Jase]

I didn't think the was needed there, but apparently it was...

Lol...my bad.

 

[Moose Coleman]

But it is a missing variable that certainly exists! The fact that it can't be answered doesn't affect it's influence.

It doesn't prove its influence either. It seems weird you'd drop something that can't really be quantified into an rant about misinterpretation of stats.

Are you serious? You directly brought up Eberle's decrease in individual production as a parallel of what can happen to Hall.

Nope. My comment on shot quality was directed at its use as a catch-all for explaining high shooting percentages and not a particular comment on whether or not Hall will undergo the same regression.

I produced the individual production of both players over the last three years and the trend is obvious. Now it means nothing because you say so?

I'm not seeing your point. Are you saying that because Hall's counting numbers are in line with his career norms, there's no possibility of regression?

Again, it is a variable that is not addressed, and certainly not the only one.

But you tell me. How many variables are used in the assertion that the Oilers are terrible when Hall is on the ice as it pertains to this analysis?

Beats me, I didn't write it.

But hey, I'm sure you're more than welcome to do a more complete analysis if you care to. Show the corsi mob how its done.

Also to this:

Hell, it is more logical to claim that Hall's offensive output being consistent in the face of such "regression" show an improving player and not one who obviously has something wrong with him.

The point being made is that the regression hasn't happened yet.

 

[Ol' Jase]

It doesn't prove its influence either. It seems weird you'd drop something that can't really be quantified into an rant about misinterpretation of stats.

Obviously you are ignoring that it is, in fact, a variable in the sport of hockey, whether it is measureable or not. There are varying degrees of shots that have more likelihood of being successful. If you refute that as false, I can't help you.

Nope. My comment on shot quality was directed at its use as a catch-all for explaining high shooting percentages and not a particular comment on whether or not Hall will undergo the same regression.

As I clearly stated it was one of many unknown variables more than once, and again you are referring that I mean it in a stand alone context, I can only assume you are being purposely obtuse. Obvious, one of many variables isn't enough to convince you that I am simply referring to it as an unknown variable. Again, can't do anything if you're wilfully going to be spiteful.

I'm not seeing your point. Are you saying that because Hall's counting numbers are in line with his career norms, there's no possibility of regression?

The article clearly indicates that the data explicitly implies a regression has in fact taken place, and the prove is based on one unit based metric:

According to the general manager Edmonton Oilers, Taylor Hall is coming along nicely as a hockey player. He is making good decisions and playing better than ever this season. Maybe he's right. But the data emphatically points to something having gone sideways in the game of the Oilers’ best offensive player.

Beats me, I didn't write it.

But hey, I'm sure you're more than welcome to do a more complete analysis if you care to. Show the corsi mob how its done.

Or I can question an analysis that doesn't even use correct data to draw it's conclusions, point out the disparity of data, provide various individual derived statistics to at least call the hypothesis into question and since I have publicly retracted and claimed error for the Corsi mob comment, I can assume this is also a comment made out of spite. Any interest in refuting the numbers I am using?

Also to this:

The point being made is that the regression hasn't happened yet.

Except the point of the article it that regression has clearly happened, and the Oilers are terrible when Taylor Hall is on the ice.

Above

 

[dustrock]

I think Hall is a great player, but he would be an interesting player to play with.

He has developed into a pretty great passer, but he mostly wants the puck on his stick at all times and is so fast, he's usually ahead of the play and his teammates.

 

[nexttothemoon]

http://stats.hockeyanalysis.com/rat...00&teamid=12&type=goals&sort=F20&sortdir=DESC

5on5 zone start adjusted (most accurate way to look at 5 on 5 play)

Hall is #2 on the team (behind only Perron) in GF20 when he is on the ice. The team produces at a rate of .963 GF per 20 minutes when he's out there... the team produces at a rate of .984 GF per 20 minutes when Perron is out there.


Now looking at individual scoring stats... Hall is easily #1 on the team in offensive production... 2.89 pts per 60 minutes at 5 on 5 zone start adjusted. It's not even close... next is Perron at 1.97 pts per 60 minutes in that stat.

Hall faces the toughest competition among all Oiler forwards:

His opposition scores .847 and allows just .777 goals per 20 minutes. That's a ratio of .522... again that's the toughest faced by any forward on the team.


All the other stats... Fenwick, Corsi, shots... all superficial to what really matters.. goals for and against while you are on the ice.


Hall is clearly the #1 offensive bus driver on this team and the only player you could argue even remotely as good is Perron. Perron is better defensively but is quite a bit below offensively where Hall is far superior so the clear edge has to go to Hall.


It's fun to make up stuff just for the sake of arguments and debate, but facts are facts... Hall is the straw that stirs the drink on this team and without him this teams sucks even worse.

 

[Tarus]

Goals don't have any worth when discussing the performance of a player, only shots on net matter.

 

[CornKicker]

But we suck all the time...
I reckon the Oilers would be legendarily bad (no games won) without Hall...

i came here to say the oilers are equally as bad when he isnt on the ice so i dont get it

 

[CornKicker]

Goals don't have any worth when discussing the performance of a player, only shots on net matter.

thats why radek dvorak was a superstar

 

[Addison Rae]

And on the contrary Hall was a fantastic possession player last year, especially in relation to his team. He is a fantastic player, I put him on par with Tavares the fact people discrete such a tremendous young star baffles my mind.

 

[40oz]

Injury or coaching?
I think it's pretty clear that entering every new season with a new head coach wreaks havoc on many aspects of a players game. This isn't an indictment of Eakins or Hall (that can be discussed from many other aspects) but it certainly is interesting and is something worth discussing.

I enjoy the reactions to articles like this because people either think it's something definitive (positive or negative) or they're open to read and interpret it and and wonder what it may mean. Writing off stats because you don't agree with some of it's conclusions is ridiculous. If anything it's more of a reason to take the chance learn about them and to try and to identify why the author's interpretation is different than yours.

 

[Ol' Jase]

Injury or coaching?
I think it's pretty clear that entering every new season with a new head coach wreaks havoc on many aspects of a players game. This isn't an indictment of Eakins or Hall (that can be discussed from many other aspects) but it certainly is interesting and is something worth discussing.

I enjoy the reactions to articles like this because people either think it's something definitive (positive or negative) or they're open to read and interpret it and and wonder what it may mean. Writing off stats because you don't agree with some of it's conclusions is ridiculous. If anything it's more of a reason to take the chance learn about them and to try and to identify why the author's interpretation is different than yours.

Who's writing them off?

It's the major leaps in logic and individual player analysis based SOLELY on a unit derived metric that has people questioning the end conclusion.

I still haven't heard anything as to the accuracy of the data in the first place, and why different sources have different data, or why individual metrics that would contradict the hypothesis are not taken into account.

I find that extremely curious.

 

[Oi'll say!]

http://oilersnation.com/2014/1/16/there-is-something-wrong-with-taylor-hall

So many jumps in logic and confirmation bias I just had to post it.

As always, the usual questions arise, most importantly why do people continue to use unit derived metrics to evaluate individual performance? But to outright claim "the Oilers are terrible" when Hall is on the ice in one breath and in the next show the team scoring at EV at an elite level when he is on the ice, and then claim definite proof of Hall "having something wrong"...it's really something else to witness.

Advanced stats had, and still have, so much promise in terms of analysis tools. It's a damn shame that the spotlight is being shone on them by people who simply can not interpret them properly.

Willis interpreting advanced stats is like a complete novice moving chess pieces around the board. They may know the basic moves of all of the pieces but have no clue how to move them with any semblance of purpose.

 

[Ol' Jase]

Willis interpreting advanced stats is like a complete novice moving chess pieces around the board. They may know the basic moves of all of the pieces but have no clue how to move them with any semblance of purpose.

The article is one thing, the amount of followers that attribute this as positive proof of his hypothesis full stop is what really makes me wonder.

I will say it again. Advanced stats have an important role in overall analysis, but the amount of misuse and "misusers" is growing daily.

 

[Replacement*]

Goals don't have any worth when discussing the performance of a player, only shots on net matter.

This would seem to be the most basic statement in the thread but its one that needs to be made apparently.

Willis is looking at stats that should only ever be supplemental information and he's using it as primary information. He's also selectively doing that using any team metric shot stats that make Hall look bad.

The basic thing is one should NEVER make conclusions derived on supplemental information while completely ignoring primary information.

Primary information for offensive players involve goal, assists, pts. Something Hall is extremely good at and only getting better. Strangely, in the want to make it look like Hall hampers this club that is entirely left out by Willis.

Next, due to the use of selective advanced stats tunnel vision false attribution is arrived at. With the conclusion that Hall is hurting the club. I think other views could be had that the respective overall games of linemates Eberle, and RNH have regressed. They have regressed SIGNIFICANTLY. For both players their production has decreased relative to their best years. In the case of RNH especially his all round game has completely fallen apart and he possesses the worst GA of any forward in the NHL. A more alarming stat considering Nuge is not regularly on pk. More trouble, and I think this comes closest to explaining some poor results, is RNH GA is over TWICE what it was last season.
The suggestion can be made that SA, GA, have become more pronounced due to Halls linemates stuggling more with all round play this year. I think that would be the simplest conclusion. Somehow it wasn't arrived at.

 

[Shizuka]

While stats and the light they are casting may be misinterpreted and used to forward all sorts of agendas.. there is no way in hell you can deny that Hall has not been good the last stretch of games. Come on now, watch the games.

 

[Soli]

Has anyone made the joke/truth that the title should read: 'News and Blog Report: Oilers are terrible when Oilers are on the ice' yet?

 

[Replacement*]

The article is one thing, the amount of followers that attribute this as positive proof of his hypothesis full stop is what really makes me wonder.

I will say it again. Advanced stats have an important role in overall analysis, but the amount of misuse and "misusers" is growing daily.

I would say they are being too guilty of trying to be intelligent by half. Good job on outing this as I think its one of those hallmark moments where the advanced stats crowd seem to have it all wrong.

Its beyond me how so much impetus can be put on SA/SF instead of actual primary counting numbers like GF/GA. Or the persons individuals pts which reflects their relative involvement in production.

It seems that in the want to go full on corsi the stats crowd are cutting off their arms to save their fingers.

I'll say this too, because apparently it needs stating. Shots are a proxy for puck possession, it does not equate puck possession. Next, possesion is a proxy for who is being the better team in a game. It doesn't in all instances indicate who is the better team or was in control of the game.
Next, as often mentioned not all shots are created equal. Shots stats largely ignore that thinking it all equals in the wash. Yet in context we know teams took more shots on goal this year from all angles and distances because of our goaltending. We know this because teams were stating that intent often. So obviously SA are going to increase somewhat.

It should never need stating but Goals are a much more homogenous stat than shots. Shots can be many different things ranging from completely harmless to dangerous. Goals, being a scoring result, are obviously more similarly dangerous instances (dubnyk notwithstanding) because they are goals.

Lastly again captain obvious here, the Goal is to score a goal, the goal is to score more goals than the opponent. The goal of hockey is not to have more shots than the opponent. Yet we have people placing more priority on the latter.

 

[Replacement*]

Has anyone made the joke/truth that the title should read: 'News and Blog Report: Oilers are terrible when Oilers are on the ice' yet?

Thats a double secret advanced stat that requires like lots of math.

 

[KeithIsActuallyBad]

Really starting to think Hall isn't as good as we think he is.

 

[Ol' Jase]

I would say they are being too guilty of trying to be intelligent by half. Good job on outing this as I think its one of those hallmark moments where the advanced stats crowd seem to have it all wrong.

Its beyond me how so much impetus can be put on SA/SF instead of actual primary counting numbers like GF/GA. Or the persons individuals pts which reflects their relative involvement in production.

It seems that in the want to go full on corsi the stats crowd are cutting off their arms to save their fingers.

I'll say this too, because apparently it needs stating. Shots are a proxy for puck possession, it does not equate puck possession. Next, possesion is a proxy for who is being the better team in a game. It doesn't in all instances indicate who is the better team or was in control of the game.
Next, as often mentioned not all shots are created equal. Shots stats largely ignore that thinking it all equals in the wash. Yet in context we know teams took more shots on goal this year from all angles and distances because of our goaltending. We know this because teams were stating that intent often. So obviously SA are going to increase somewhat.

It should never need stating but Goals are a much more homogenous stat than shots. Shots can be many different things ranging from completely harmless to dangerous. Goals, being a scoring result, are obviously more similarly dangerous instances (dubnyk notwithstanding) because they are goals.

Lastly again captain obvious here, the Goal is to score a goal, the goal is to score more goals than the opponent. The goal of hockey is not to have more shots than the opponent. Yet we have people placing more priority on the latter.

This is an excellent post.

I'll add one more thing to it as well. These numbers that are used completely discount two facts.

1. Blocked and missed shots have 0% probability of going in the net, which destroys the ridiculous notion that shot quality means nothing when it comes to Corsi.

2. When a player is generating offense at the elite level that Hall is right now (See the P/60 link), wouldn't logic dictate that attempted shots may suffer due to the fact that more are going in the net???
I'm not the smartest guy, but I do know that shot attempts on an offensive possession end when a goal is scored.

 

[Ol' Jase]

Really starting to think Hall isn't as good as we think he is.

Yeah, 2nd in the league in P/60. Bum.

He generates offense at an elite level unseen on this team in decades. I'd say that's pretty good.

 

[KeithIsActuallyBad]

Yeah, 2nd in the league in P/60. Bum.

He generates offense at an elite level unseen on this team in decades. I'd say that's pretty good.

Still can't play a lick of defense.

And he crumbles a lot under physical pressure.

 

[Ol' Jase]

Still can't play a lick of defense.

And he crumbles a lot under physical pressure.

So it's his lack of defensive ability that is responsible for our GA being as high as it is. Cool.

And his offensive numbers kind of say he doesn't exactly crumble a lot in any situation.

Kid is an offensive force in the NHL at 21, and if he not a high level two way forward right now, that means he's no damn good. Wow.

 

[Draiskull]

I dont need to read stats to know that Hall is one of the worst in the league in D zone... Give me grebeshkov, blindfold him and he would be as effective.. offensive zone is a different story.. Hall is our MVP blueline in.