Jump to content
The Official Site of the Vancouver Canucks
Canucks Community

[Waivers] Frk


Provost

Recommended Posts

15 hours ago, Provost said:

I will slow it down for you so you understand the math.

 

Excluding special teams, excluding shift starts on the fly, excluding shift starts in the neutral zone (your favourite stat excludes all of those)... it leaves around 10 oZone or dZone starts per game for a forward... give or take depending on their line, but is a close median and makes the math easy for you to follow.

 

10 starts measured and if you start 4 of the 10 in the oZone that is 40% oZone starts.  If you start 6 of the 10 in the oZone, thst is 60% oZone starts.  The difference is literally two starts (just like I said) and they could both be on one single shift.

 

As as you said, there can be multiple starts per shift, and using 20-25 total shifts per game as an average for a forward. times that multiplier of number of starts per shift, the percent of those starts that is measured in that single oZone starts stat is a fraction of their total deployment.  Just using a conservative 2.5 starts per shift gives you 50-62.5 total starts, of which your favourite particular stat only measures 15-20%, and excludes the rest.  No sane person relies on a small fraction of the deployment to rate a player.  Especially the way you use it to call out other people posting as being stupid.  It is a very limited measurement and you use it like it is the bible to thump over people.

 

The link I posted literally talked about why it is important to consider SHIFT starts rather than just ones that happen to start with a face off in the oZone or dZone.

 

It is a poor metric that excludes far too much.  

If you take a sample of one game, of course, it is a poor, useless metric.  But over the course of the year, this metric shows tendancy, average and becomes more relevant.

 

It is the same metric as batting average in baseball ....a guy that hit 0.600 or 3/5 on any given night, only has two more hits than a guy that bats 0.200 or 1/5.  Doesn't mean that guy is a better player than the other guy.  But as the season goes on, with an increase population of data, the average is a pretty good metric.

Link to comment
Share on other sites

19 hours ago, Provost said:

I didn’t say we should pick him up.  I said that twice already including the post you quoted.  I also said I don’t know where he would fit on our roster.

 

I said “some” team might.

 

Going back to my OP, to fit on our team, we would have to have another long term injury.  With Sutter and Baertschi out, we already have Mcewan and Gaudette up.  Say if Virtanen or Erickson went out, then that is a position where we would want to look at guys like Frk as a temporary alternative to putting Schaller back in the lineup or pulling another guy from the farm who is farther down the depth chart.  The other alternative would be calling Gagner back up... and he might be better off in the minors as it is remotely possible someone takes him off our hands st the deadline as playoff insurance... and him already waived and in the minors is more valuable to a team close to the cap than him in the NHL.

Maybe not directly....but you start a post on Frk being on waivers and you comment that our 2 PP unit need skills....if you are not insinuating that we should pick him up for our 2nd PP unit, what is the relevance of that comment in a Martin Frk thread?

Link to comment
Share on other sites

1 minute ago, timberz21 said:

If you take a sample of one game, of course, it is a poor, useless metric.  But over the course of the year, this metric shows tendancy, average and becomes more relevant.

 

It is the same metric as batting average in baseball ....a guy that hit 0.600 or 3/5 on any given night, only has two more hits than a guy that bats 0.200 or 1/5.  Doesn't mean that guy is a better player than the other guy.  But as the season goes on, with an increase population of data, the average is a pretty good metric.

It is a not a performance indicator at all and doesn’t pretend to be.

It is a relatively weak indicator of how a coach deploys players.

It excludes too much to have an accuracy and has too much noise to make even that limited info accurate.

It isn’t entirely meaningless as it shows a general tendency towards how a player is deployed.  It just isn’t strong enough to use in isolation to tell anything of real substance about a player.

You just can’t exclude 80-90% of a player’s deployment and then say the stat measures their deployment.

 

Starts are pretty random.  If you end up starting in the offensive zone because the offensive players just finished a PP shift (not counted in their stats)... then on that that one shift, there are two quick shots on goal that the goalie covers and you face off a couple of times, then a shot goes into the stands.  You suddenly have four oZone starts within 20 seconds and your ratio shoots up past the actual offensive players because there are so few actual oZone or dZone starts in a game.

Link to comment
Share on other sites

9 minutes ago, timberz21 said:

Maybe not directly....but you start a post on Frk being on waivers and you comment that our 2 PP unit need skills....if you are not insinuating that we should pick him up for our 2nd PP unit, what is the relevance of that comment in a Martin Frk thread?

Don’t try to read in things between the lines you think are insinuated when there are literally words there that say differently.

 

I literally said outright 

 

“Not sure if we have the room for him depending on length of injuries, but a low risk pick up for some team.”

 

That also included the scenario where we WOULD need a guy like him... if we had a bunch of long term injuries.

Link to comment
Share on other sites

6 minutes ago, Provost said:

It is a not a performance indicator at all and doesn’t pretend to be.

It is a relatively weak indicator of how a coach deploys players.

It excludes too much to have an accuracy and has too much noise to make even that limited info accurate.

It isn’t entirely meaningless as it shows a general tendency towards how a player is deployed.  It just isn’t strong enough to use in isolation to tell anything of real substance about a player.

You just can’t exclude 80-90% of a player’s deployment and then say the stat measures their deployment.

 

Starts are pretty random.  If you end up starting in the offensive zone because the offensive players just finished a PP shift (not counted in their stats)... then on that that one shift, there are two quick shots on goal that the goalie covers and you face off a couple of times, then a shot goes into the stands.  You suddenly have four oZone starts within 20 seconds and your ratio shoots up past the actual offensive players because there are so few actual oZone or dZone starts in a game.

Aren't all hockey metric useless in isolation.  That's why they all have to be interpreted in relation to another, so the the metric itself is fine, but obviously GM aren't handing out contracts based on Ozone start %.

 

It might excluses 80-90% of the player's deployment, however, it does the same for every player.  So when you compare player to player their ratio is based on the same data, so why can't you use it to compare player.  If Ozone/DZone only count for 10-20% of deployment....then why does home teams have last change and why are line matching so crucial for many coaches if it's not that important?

Link to comment
Share on other sites

12 minutes ago, Provost said:

Don’t try to read in things between the lines you think are insinuated when there are literally words there that say differently.

 

I literally said outright 

 

“Not sure if we have the room for him depending on length of injuries, but a low risk pick up for some team.”

 

That also included the scenario where we WOULD need a guy like him... if we had a bunch of long term injuries.

Then why in the hell you say our 2nd PP unit could use some skills?  It has zero relevance in this thread.

 

And you never say you don't want to pick him....you say you're not sure if we have room.   IMO that says you want to pick him, your just not sure who you're letting go to fit him.

Link to comment
Share on other sites

1 minute ago, timberz21 said:

Aren't all hockey metric useless in isolation.  That's why they all have to be interpreted in relation to another, so the the metric itself is fine, but obviously GM aren't handing out contracts based on Ozone start %.

 

It might excluses 80-90% of the player's deployment, however, it does the same for every player.  So when you compare player to player their ratio is based on the same data, so why can't you use it to compare player.  If Ozone/DZone only count for 10-20% of deployment....then why does home teams have last change and why are line matching so crucial for many coaches if it's not that important?

 You can’t use it to compare players because the 80-90% of their other deployment is not captured at all.  What would you be using it to tell you about the players?

Oldnews uses it in isolation to say a player has the hard or easy minutes or and whether they are an offensive or shutdown player.

 

Horvat could kill every penalty, play on the top unit PP, jump over the boards on the fly to be matched up against the opposition’s top line... and the stat captures zero of that.  It just captures the few times a game he takes a face off in two of the three zones.

He could be deployed exactly the same way for 5 games and the oZone ratio for each game could have zero relationship with each other... he could be 80% oZone starts one game and 20% the next with no difference in how he is deployed.  Go look how volatile the numbers are game to game, and it shows how much noise and randomness there is to it.  Those players aren’t deployed vastly different from game to game.

 

With enough games it will indicate a weak tendency towards ho a player is deployed in general.

Link to comment
Share on other sites

39 minutes ago, timberz21 said:

Aren't all hockey metric useless in isolation.  That's why they all have to be interpreted in relation to another, so the the metric itself is fine, but obviously GM aren't handing out contracts based on Ozone start %.

 

It might excluses 80-90% of the player's deployment, however, it does the same for every player.  So when you compare player to player their ratio is based on the same data, so why can't you use it to compare player.  If Ozone/DZone only count for 10-20% of deployment....then why does home teams have last change and why are line matching so crucial for many coaches if it's not that important?

This isn't true. If you are comparing 10-20% of the deployment of players with each other, what exactly is being compared at that point? Let's compare my leg with your arm and see the relation. Based on these results, your hand can't walk as well as my foot and my foot can't grab anything to save my life. Yet, we've compared 20% of my body with 20% of your body. Even if we were to compare legs with legs, it doesn't really show much about us.

 

If you want a solid comparison between 2 relations, a 20% sample size is not good enough.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...