Jump to content
The Official Site of the Montréal Canadiens
Canadiens de Montreal

#27 Alex Galchenyuk 2016-17


habs_93
 Share

Recommended Posts

I see the flashes of talent, but honestly ... I don't like his effort at times or his decisions.    One moment he hangs on to the puck too long.    Next shift, he's making a creative pass his wingers can't anticipate.   Maybe some of that is confidence ... destroyed by MT for sure ... but since his injury I honestly don't believe he's played himself back onto the 1st or 2nd line ... if I were a coach I know I wouldn't be satisfied with his performances.

 

 

Link to comment
Share on other sites

I think a lot of that is confidence.  But the way you build a guys confidence isn't by throwing him on the 4th line with your 'character' players.  While ott and martinsen are big bodies they definitely can't anticipate it when Chucky makes a creative pass.  Perhaps the reason he's holding onto the puck for so long is he sees what happens when he gets it to his line mates?  

No matter what you think of his effort he did manage the 2nd most shot attempts last night on the team with 7.  Only pacioretty and weber had more; and they both had much more ice time.

Link to comment
Share on other sites

3 hours ago, HabsAlways said:

I see the flashes of talent, but honestly ... I don't like his effort at times or his decisions.    One moment he hangs on to the puck too long.    Next shift, he's making a creative pass his wingers can't anticipate.   Maybe some of that is confidence ... destroyed by MT for sure ... but since his injury I honestly don't believe he's played himself back onto the 1st or 2nd line ... if I were a coach I know I wouldn't be satisfied with his performances.

If we keep playing him with pluggers the bolded part will continue happening.  We have the luxury right now of having two lines that are playing pretty well so we can actually give Galchenyuk 2 decent linemates (Lekhs & Shaw) with no ill effect.   If we want a little more skill on the fourth line sit two of Ott, Martensen and King and bring back Mitchell and Carr. 

Link to comment
Share on other sites

Todd in one of the posted inside out videos was joking about difficult fathers causing problems for players and I wonder if Galchenyuk has been dealing with a lot of meddling through his young career and this is impacting his play. At this stage one would expect him to have become a more consistent player but he continues to give the puck away and his body language after these mistakes is getting worse. He looked lost out their for significant periods of time, and took a penalty when given the chance to take the draw on the power play. For a player with his skill level it is tough to watch him struggle mentally with his game.

Link to comment
Share on other sites

Just like Dave posted in the "Habs Lines" thread, they could go back to the lines that were working at the beginning of the season when they were scoring lots of goals:

 

     Pacioretty-Galchenyuk-Gallagher

     Radulov-Plekanec-Lehkonen

 

Galchenyuk probably does need to play with the highest skill players we've got.  I would be willing to bet that he and Lehkonen on the same line would be dynamite.  

Link to comment
Share on other sites

10 hours ago, Habberwacky said:

Todd in one of the posted inside out videos was joking about difficult fathers causing problems for players and I wonder if Galchenyuk has been dealing with a lot of meddling through his young career and this is impacting his play.

I imagine having the first five years of his career wasted playing for one of if not the worst coach in the NHL might have something to do with it, too.

Link to comment
Share on other sites

19 hours ago, Habberwacky said:

Todd in one of the posted inside out videos was joking about difficult fathers causing problems for players and I wonder if Galchenyuk has been dealing with a lot of meddling through his young career and this is impacting his play. At this stage one would expect him to have become a more consistent player but he continues to give the puck away and his body language after these mistakes is getting worse. He looked lost out their for significant periods of time, and took a penalty when given the chance to take the draw on the power play. For a player with his skill level it is tough to watch him struggle mentally with his game.

I thought that what Todd said in that video was very revealing, and may explain some of the reasons why Galchenyuk has been struggling. Nothing worse than an interfering hockey Dad.

Link to comment
Share on other sites

3 hours ago, Habs=stanleycup said:

I thought that what Todd said in that video was very revealing, and may explain some of the reasons why Galchenyuk has been struggling. Nothing worse than an interfering hockey Dad.

Geez......interfering hockey dads.....the need for the head coach to speak French.....doesn't this franchise care about winning any more? 

Link to comment
Share on other sites

On 4/10/2017 at 11:08 PM, BigTed3 said:

He has only reached 30 once, but that's not a very fair way of assessing him as a player. His first season was a lockout-shortened season and he was only given 12 minutes of ice a game. His production increased in each of the next 3 seasons, and his production per ice time was actually better this year than it has been in any of the three preceding seasons. So what's different? His ice time has been kept low, and he was injured this season and missed time. Galchenyuk has averaged about 16 minutes of ice the past three years... compare that to Pacioretty at 19 minutes or Radulov at over 18 or so on. Do you know how many Canadiens had more points per ice time than Galchenyuk? None. Not a single one (unless you count Hudon and his 2 points in 3 games). Despite missing over 20 games, he still managed to put up more points than players like Byron and Danault and Lehkonen and Plekanec and Gallagher, all of whom were given just as many top 6-9 opportunity as he was.

So extrapolate his production over a full 82 games, and he would have had 23 goals and 59 points, despite getting tossed around the line-up and having to spend some of that time recovering from injury. Now give him 19 minutes a game instead of 16 and that would extrapolate to 27.3 goals. Now keep him healthy and play him as a 1C instead of giving him games playing with Shaw or Ott or so on... he'd be at 30 goals or more. If you keep him as a 3rd line winger, no, he'll never get to 30, but my statement was that if you get to play him as a 1C, then he'd score 30+. He's a very capable player who's effect is being minimized by poor usage.

As for +/-, I won't put any stock in that. I don't believe Bobby Farnham and his 0 rating are better than Galchenyuk or that Zack Redmond and his +6 have more value than Petry and his +3. Or are Sergachev (+1), Beaulieu (+8), and Pateryn (+1) all superior defensively to Jordie Benn and his -1?

My assesment is not fair? FACT: he has only scored 30 goals once in 5 seasons.  Why would anyone bet he would score  30 to 40 goals in each of the next 5 seasons? I guess if you were given really high odds and a large payout.

Extrapolate? FACT: he has scored 30 goals once in 5 seasons. you have a lot of ifs and's or but's in this paragraph. if the leafs beat us all this season we would be in the wild card and they would be in first, but they didn't.

No stock in plus minus? I guess you only read what you wanted, I was comparing players over 20 games. And in no way was i discussing value of a player how you came up with value is beyond me!  Does having a higher corisca indicate you have more value to a team? No. All these stats are flawed, but they do indicate a trend. Having the worst plus minus on a team and a negative plus minus on a team that was a plus team, is certainly indicating that he is on the wrong end of the puck more then he in on the good end of the puck.

Do i think he has tremedous offensive skill and upside I certainly do. Do i want to blame Therrien for his game? Yep. Do i think he should have been at center from the get go and live with the mistakes? Yep. Do I believe he would be better if theses scenarios were different? Yep. Do i know for a fact he would be? Nope. Can  change them? Nope. Fact is he has been nothing short of brutal from the red line in on the defensive side of the puck. And the puck has ended up in our net more then in the opposing net this season when he is on the ice.  

Oh and by the way I do think he has value the type of value that could net us a really good two way center :) 

Link to comment
Share on other sites

1 hour ago, Ravadak said:

All these stats are flawed,

Okay. How?

1 hour ago, Ravadak said:

Having the worst plus minus on a team and a negative plus minus on a team that was a plus team, is certainly indicating that he is on the wrong end of the puck more then he in on the good end of the puck.

Measures with no validity are equally as invalid whether moderate or extreme in nature, and +/- objectively has no value in determining individual ability.

Link to comment
Share on other sites

We saw a line of Gallagher-Galchenyuk-Lehkonen for a while in Game 2.  I think that could be a dynamic scoring line.  One could make the argument that having Galchenyuk on the 4th line with pluggers for the entire 60 minutes cost us Game 1.  That was essentially a one-goal game until we pulled Price.  Oh, well, what's done is done.  At least Coach Julien made the adjustment in Game 2.  

Link to comment
Share on other sites

On 4/15/2017 at 9:50 AM, habs_93 said:

Okay. How?

Measures with no validity are equally as invalid whether moderate or extreme in nature, and +/- objectively has no value in determining individual ability.

Okay. How?

Do stats predict a future outcome 100% of the time?

Do stats predict how good a player will be 5 years from now?

Do stats predict what team will win the cup this season?

No no and no, because there are two many variables. So stats have flaws. That is how :)

I have a question for you. In its simplest form. If a player is minus 5 at the end of the night, Does that mean he was on the ice when his team was scored on 5 more times then when his team scored? If you answer yes to that, then it is a measurement none the less in its simplest form. Now if you choose to ignore it because of all the other variables or noise that is your choice. But it is a measurement and it is accurate fixed number that he/she was on the ice more times 5 more times when they were scored against. Fluke or bad luck maybe 1 game, 5 games, 10 games, 20 games, but after 30 games, 40 games 50 games it becomes a trend.

Link to comment
Share on other sites

8 hours ago, Ravadak said:

Okay. How?

Do stats predict a future outcome 100% of the time?

Do stats predict how good a player will be 5 years from now?

Do stats predict what team will win the cup this season?

No no and no, because there are two many variables. So stats have flaws. That is how :)

I have a question for you. In its simplest form. If a player is minus 5 at the end of the night, Does that mean he was on the ice when his team was scored on 5 more times then when his team scored? If you answer yes to that, then it is a measurement none the less in its simplest form. Now if you choose to ignore it because of all the other variables or noise that is your choice. But it is a measurement and it is accurate fixed number that he/she was on the ice more times 5 more times when they were scored against. Fluke or bad luck maybe 1 game, 5 games, 10 games, 20 games, but after 30 games, 40 games 50 games it becomes a trend.

Here's the thing though... stats don't have to predict 100% of the outcomes to be valid. Stats predict trends. If I were rolling a dice and I told you this dice was weighted so that "6" was likely to come up 50% of the time and the other five numbers were each likely to come up 10% of the time, which number would you put money on coming up? You'd probably put your money on 6, even though you'd be wrong half the time. The fact is that 6 is still the most likely event to occur.

Same thing in hockey. Corsi and advanced metrics HAVE been shown to correlate with future success. It's not 100% correlation, but that doesn't invalidate the fact that these numbers have value in predicting outcomes. If we got to the finals, would you want the Habs to be the team with the 55% chance of winning the Cup or the one with a 45% chance of winning? Even though the difference might seem small to you, why would you not take the extra 10% odds on your side?

As far as +/- goes, it's a stat that shows you what has happened, you're right, but it has very little predictive value for what will happen next. A lot of that stat comes down to things a player can't control, so it's not something that's reproducible and it's generally not an accurate reflection of a player's added value to the team. The fact that a coin flip was heads the last eighteen times does not mean it's more likely to be heads the next time you flip the coin.

Link to comment
Share on other sites

19 minutes ago, BigTed3 said:

Here's the thing though... stats don't have to predict 100% of the outcomes to be valid. Stats predict trends. If I were rolling a dice and I told you this dice was weighted so that "6" was likely to come up 50% of the time and the other five numbers were each likely to come up 10% of the time, which number would you put money on coming up? You'd probably put your money on 6, even though you'd be wrong half the time. The fact is that 6 is still the most likely event to occur.

Same thing in hockey. Corsi and advanced metrics HAVE been shown to correlate with future success. It's not 100% correlation, but that doesn't invalidate the fact that these numbers have value in predicting outcomes. If we got to the finals, would you want the Habs to be the team with the 55% chance of winning the Cup or the one with a 45% chance of winning? Even though the difference might seem small to you, why would you not take the extra 10% odds on your side?

As far as +/- goes, it's a stat that shows you what has happened, you're right, but it has very little predictive value for what will happen next. A lot of that stat comes down to things a player can't control, so it's not something that's reproducible and it's generally not an accurate reflection of a player's added value to the team. The fact that a coin flip was heads the last eighteen times does not mean it's more likely to be heads the next time you flip the coin.

All stats have variables but +/- does also show predictive trends if that player continues to stay a - player or if it decreases or increases in value. Corsi +/- there is no absolute , because there are variables in hockey and values just as when Gallager isn't scoring but he is a pest if he is taking another player especially the goalie off their game than yes that has a value which can be big. That alone is why years ago the Canadian Olympic team went from compiling a 'team" of all scorers and all the most skilled players to a team with balance and players that would be role players during the games. So even variables which sometimes can't be qualified are things that are looked at and for good reasons. So all stats and variables have to be taken into account even face offs blocked shots experience ect. If corsi or fenwick were 100% definitive than all teams scouts coaches GM's ect. would only use those. So we may disagree with what stats or even variables mean more , but we are fans and that makes it fun , we can debate about players buy just watching the games also without having to even quote any stats.

Link to comment
Share on other sites

11 hours ago, Ravadak said:

Do stats predict a future outcome 100% of the time?

No, no one says they do, and they don't have to.

11 hours ago, Ravadak said:

Do stats predict how good a player will be 5 years from now?

There are individual statistics with limited demonstrated projective validity out over multiple seasons. They don't provide much information, but some is better than none.

11 hours ago, Ravadak said:

Do stats predict what team will win the cup this season?

See above.

11 hours ago, Ravadak said:

because there are two many variables. So stats have flaws. That is how :)

There's nothing wrong about knowing very little about statistics, but glib truisms are intellectually dishonest.

As for the rest, and this:

2 hours ago, CaptWelly said:

All stats have variables but +/- does also show predictive trends

I'm so over talking about +/-. No, it doesn't. It's demonstratively useless. I understand the desire for a simple, "straightforward" measure of defensive ability, but clinging to +/- is pointless. And yet another appeal to authority fallacy isn't going to convince anyone who's done any kind of independent investigation of it for themselves.

2 hours ago, CaptWelly said:

If corsi or fenwick were 100% definitive

You repeat this over and over. And I'll remind you again that no one here—no one—has ever claimed that any analytical tool was 100% accurate. Repeating something you know to be false which misrepresents what people have said is extremely impolite.

2 hours ago, CaptWelly said:

we can debate about players buy just watching the games also without having to even quote any stats.

You certainly can. But when you misrepresent what people say, or contend that there's somehow some kind of alternative mathematics, that's something else entirely.

Link to comment
Share on other sites

11 hours ago, BigTed3 said:

Here's the thing though... stats don't have to predict 100% of the outcomes to be valid. Stats predict trends. If I were rolling a dice and I told you this dice was weighted so that "6" was likely to come up 50% of the time and the other five numbers were each likely to come up 10% of the time, which number would you put money on coming up? You'd probably put your money on 6, even though you'd be wrong half the time. The fact is that 6 is still the most likely event to occur.

Same thing in hockey. Corsi and advanced metrics HAVE been shown to correlate with future success. It's not 100% correlation, but that doesn't invalidate the fact that these numbers have value in predicting outcomes. If we got to the finals, would you want the Habs to be the team with the 55% chance of winning the Cup or the one with a 45% chance of winning? Even though the difference might seem small to you, why would you not take the extra 10% odds on your side?

As far as +/- goes, it's a stat that shows you what has happened, you're right, but it has very little predictive value for what will happen next. A lot of that stat comes down to things a player can't control, so it's not something that's reproducible and it's generally not an accurate reflection of a player's added value to the team. The fact that a coin flip was heads the last eighteen times does not mean it's more likely to be heads the next time you flip the coin.

I agree with you 100%! on most of your post.

However, I do think + - is a measurement and can be predictive, is it as valuable as some other metrics u have mentioned? No, I.E., Galchenyuk 2013, bottom 2+-. 2015 bottom 5, 2016, dead last + -, is that just bad luck? 3 out of 4 years. I would be willing to bet based on the previous 3 out of 4 years +- he  will likely be in the bottom 1/3 more often then not. Just like i would bet you he does not score 30 to 40 goals in the next five season, because he has only reached 30 goals once in 5 season. 

Anyway the reason for my post regarding flaws in metrics was a response to habs_93, who asked how stats are flawed, and my answer was quite simple unless they can predict 100% of an outcome they have to have flaws.

Can they provide insight into what may happen? u bet they can :), are they always right? nope. flawed :)

Link to comment
Share on other sites

4 minutes ago, Ravadak said:

However, I do think + - is a measurement and can be predictive

Demonstrate it using standard statistical methods and document exactly how you do it so other people can try the same process themselves and confirm your work.

4 minutes ago, Ravadak said:

and my answer was quite simple unless they can predict 100% of an outcome they have to have flaws.

That isn't how this works.

Link to comment
Share on other sites

1 minute ago, habs_93 said:

Demonstrate it using standard statistical methods and document exactly how you do it so other people can try the same process themselves and confirm your work.

I will show you how it works. I will put 100 dollars up in each on the next 5 seasons. That galchenyk ends up in teh bottom half of +- on the habs, if you hmatch the bet and put 100 each season that he ends up in the top half of +-

That isn't how this works.

Says Who?

 

Link to comment
Share on other sites

13 minutes ago, Ravadak said:

Says Who?

People who've done even a rudimentary study of statistics.

Again, there's nothing wrong with not being an expert (or even a moderately interested layman) with regards to statistics. Just don't try to pretend that argumentation based on fallacies is equal in weight to empirical evidence.

Link to comment
Share on other sites

21 minutes ago, habs_93 said:

People who've done even a rudimentary study of statistics.

Please give me some names?

People created +- also :) 

Again, there's nothing wrong with not being an expert (or even a moderately interested layman) with regards to statistics. Just don't try to pretend that argumentation based on fallacies is equal in weight to empirical evidence. 

Are you trying to insult me :)  (informal fallacy - Appeal to ridicule: a specific type of appeal to emotion where an argument is made by presenting the opponent's argument in a way that makes it appear ridiculous) - Just because u can not comprehend that  + - even in the simplest form is still a statistic that provides a measurement. 

 

21 minutes ago, habs_93 said:

 

 

Link to comment
Share on other sites

^^ If we come back to Galchenyuk, let's all agree that Galchenyuk is a skilled player who is an offensive threat, but that his +/- numbers have been poor. The real question is what does that mean for this player going forward? So I'll ask the question this way instead: would you rather have a player like Galchenyuk on the team who puts up points but has a weak +/- statistic or would you rather have a player like Byron or Plekanec, who had good +/- numbers this season? Byron had a good season, but I'd still prefer Galchenyuk on my team. And there's not even a question for me when it comes to Galchenyuk having had a better season than Plekanec. Coming back to Byron, he was a +21 this season but a -9 last year and a -2 the year before. So which number more accurately represents what he'll give us next year?

+/- doesn't take into account special teams play. It doesn't take into account the times when players just hopped on or off the ice and get a number attached to their name for a play they weren't involved in. It doesn't account for when a goal is scored while they are on the ice but where the play was not related to their ability either. For example, if a goalie lets in a shot from center ice because they flub the puck. Or an empty net goal for/against... does it make a difference to know the Habs allowed 7 empty net goals against this year and given that AG is on the ice for a lot of the offensive situations at the end of the game, he tacks up minuses for many of these... if we only accounted for changes attributed to these negatives (which are not negatives related to poor defensive play), then Galchenyuk would conceivably go from being a -5 to a plus player at +2. Does that suddenly make him a good defensive player instead of a poor one?

Now I don't think Galchenyuk is a good defensive player by any means. But I do think his positive offensive contribution outweighs his defensive woes and I do think that his Corsi/point production/etc are better markers of his worth than a tool like +/-.

 

Link to comment
Share on other sites

 

17 minutes ago, Ravadak said:

Just because u can not comprehend that  + - even in the simplest form is still a statistic that provides a measurement. 

Okay, I'll talk about +/- again :). I've never once said +/- did not meet the dictionary definition of statistic. If you're interested in dictionary definitions, then I can understand the argument. I'm not, particularly; I'm interested in hockey, and quantifying as much of it as possible, as predictively as possible. What I've said, and said repeatedly for a while now, is that +/- does not describe what people think it describes. It is demonstrably not predictive or a representation of individual skill. On an individual level it is, at best, a gross measure of a mixture of luck and quality of teammates/quality of competition. If we didn't have better ways to measure those, there'd be a reason to look at +/-, albeit not very seriously. But we have better tools now; PDO and team relative/QoC Corsi, both of which do a monumentally better job of describing their individual bits that +/- aggregates and makes less precise. There is objectively no reason to pay attention to +/-.

I'm asserting an empirical observation which I have an extremely high level of confidence in. Because I've personally done the tests with data from NHL's Real Time Scoring System—assembled for analysis with a parser I wrote myself—that showed it. And I'm not some privileged high priest of this stuff, either. There's nothing exotic or complicated needed to comprehensively shred the myth of +/-. Anyone with a statistics textbook printed after World War II has the formulas required. These days, reasonably fast world-class statistical analysis software is available free of charge for anyone with a personal computer. Online instruction material for statistics is also available, free of charge, to anyone with the time and interest. The majority of journal articles dealing with NHL hockey analysis are published without paywalls, as well. I did it the hard way for fun, but multiple sources for NHL game data exist.

I encourage you wholeheartedly to not believe my assertions. President Reagan had a phrase for the ages: trust but verify. Don't use Excel, its statistical functions have been suspect for 20 years. Download RStudio and a season-by-season dataset of all NHL players since 2007-08 from Corsica.hockey. Filter it to eliminate players without more than two consecutive half seasons, and then subset the data to give you more longitudinal options (3-year players, 4-year players, etc etc). Run a Pearson correlation coefficient test to see how well one season's +/- correlated to the next season's. Do this longitudinally, as well, and see if the r-squared increases, decreases, or doesn't move very much. Run ANOVA. Then, do it game-by-game, and time series. See how well one game's +/- correlates to the next few games, or how the first half season's +/- correlates to the second, or how 5/10/15/20 game blocks correlate to their next 5/10/15/20 game block. Try an autocorrelation test. Split-halves, split-periods, and so on. Compare +/- to on-ice shooting and save percentages, point percentages, scoring rates, etc.

If you give it a look and still believe +/- is a reasonably accurate representation of individual skill and holds predictive value, I would be extremely interested in hearing your conclusions. If you don't have the time or inclination, I could point you to scholarly journal articles produced by people who've done all of this and more.

Link to comment
Share on other sites

1 hour ago, BigTed3 said:

^^ If we come back to Galchenyuk, let's all agree that Galchenyuk is a skilled player who is an offensive threat, but that his +/- numbers have been poor. The real question is what does that mean for this player going forward? So I'll ask the question this way instead: would you rather have a player like Galchenyuk on the team who puts up points but has a weak +/- statistic or would you rather have a player like Byron or Plekanec

The answer for me would be quite simple to your question. I would rather have Galchenyk. However I am not sure the coaching staff feels that way. If they did I am sure he would have been starting on one of the top two lines not on the 4th line in the first round of the playoffs.

i think your question should not be about who we already have but who we could get by moving Galchenyk.

Link to comment
Share on other sites

Guest
This topic is now closed to further replies.
 Share

  • Recently Browsing   0 members

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