With regards to +/-, as a team the Pens are going to have to improve at 5 on 5. I was looking on NHL.com and noticed that the Pens are currently 23rd in the NHL in GF/GA ratio, at a pretty bad .79 ratio. That surprised me, because I didn't realize how poor the Pens have played at even strength.
Right now, the Pens are winning because of their specialty teams. That's great and all, but come playoff time (when the whistles are put away and PPs are scarce), the best teams at 5 on 5 are likely the teams who will go the deepest.
With regards to +/-, as a team the Pens are going to have to improve at 5 on 5. I was looking on NHL.com and noticed that the Pens are currently 23rd in the NHL in GF/GA ratio, at a pretty bad .79 ratio. That surprised me, because I didn't realize how poor the Pens have played at even strength.
Right now, the Pens are winning because of their specialty teams. That's great and all, but come playoff time (when the whistles are put away and PPs are scarce), the best teams at 5 on 5 are likely the teams who will go the deepest.
Yes and no. Last year we were excellent at 5 on 5, but special teams killed us.
I agree with the larger point though, and it's interesting to note how we've swapped positions over the last year. I think a bit of that is us icing fewer grinders and the injuries to Crosby, Malkin, and the blueline aren't helping.
Yes and no. Last year we were excellent at 5 on 5, but special teams killed us.
I agree with the larger point though, and it's interesting to note how we've swapped positions over the last year. I think a bit of that is us icing fewer grinders and the injuries to Crosby, Malkin, and the blueline aren't helping.
Yeah, I mean Crosby was so freaking dominant 5 on 5, if he was playing we'de be in the top 10 again probably. We managed to play very solid 5 on 5 without our stars at the end of last year and I'm sure the whole team will improve there once everybody settled into the system.
Martin will be fine. People shouldn't be worried about his regular season play.
IMO, the only legit concern is how he plays come playoffs. He kind of had a so-so track record for the Devils, as I recall. Not bad, but not a difference maker when NJ was making all those early playoff exits. He was pretty good for the Pens last year, but it was a short sample size.
I think it remains to be seen how PM and also ZM handle the rigors of deep Cup runs. They remain unproven in that regard.
Agreed... and both really lack the physicality that you like to see from d-men in the post-season. I think ZM has a strong game... but his physicality reminds me of a mix of Ryan Whitney and Scuderi. It's kinda surprising for how big he is.
Agreed... and both really lack the physicality that you like to see from d-men in the post-season. I think ZM has a strong game... but his physicality reminds me of a mix of Ryan Whitney and Scuderi. It's kinda surprising for how big he is.
When we won the cup the defense was hardly a group of big hitters.
Hitting is overrated in today's NHL, IMO. You'll be way more successful with a blueline full of skaters than you will with a blueline full of bruisers.
What Martin contributes to the Pens isn't going to be measured accurately with +/- ,although he'll probably finish the season with a + rating. If it's stats you want, you'd be better served to look at his Corsi. And what the Crosi data will show you is that when Martin is on the ice shots against go down, the other team has the puck less, in fact he's tied for second on the team on takeaways.
1) Corsi takes some noise out of +/- but the raw number is still an out-of-context stat that's heavily influenced by quality of competition. Here are the *WORST* d-men in relative Corsi so far this year (per Behind the Net):
Grossman (DAL)
Schenn (TOR)
Murray (SJ)
Ballard (VAN)
Schultz (MIN)
Regehr (BUF)
Polak (STL)
Scuderi (LA)
Clark (TB)
Oduya (WPG)
(Michalek is 26th, Engelland 29th, Martin 43rd)
It's a mix of very good shutdown d-men playing against tough competition, and weaker d-men. Not much to conclude here.
2) There is an almost perfect (negative) correlation between +/- and quality of competition for Pens d-men this year --
+/- from best to worst
Niskanen
Engelland, Orpik (tie)
Letang, Lovejoy (tie)
Michalek
Martin
Quality of competition from easiest to toughest
Lovejoy
Niskanen
Engelland
Orpik
Letang
Martin
Michalek
The thing with hockey and their stats, in my opinion, is that hockey, more than pretty much any other sport, is a team game. All 4 guys on the ice have an impact on a players stats.
That's one of the main reasons why I do not like hockey stats, especially a stat like +/-.
Yes and no. Last year we were excellent at 5 on 5, but special teams killed us.
I agree with the larger point though, and it's interesting to note how we've swapped positions over the last year. I think a bit of that is us icing fewer grinders and the injuries to Crosby, Malkin, and the blueline aren't helping.
I was referring to this season in terms of my commentary on their 5 on 5 play. In fact, your point is exactly what struck me when I was looking at the stats last night -- we've flip flopped from last season, where even strength we were producing well, but on the PP we were horrible (especially after Sid/Geno went down).
If we could combine our 5 on 5 play from last year with our special teams play from this year, we'd be almost unbeatable.
Hopefully when (whenever that is) Crosby returns, along with Kennedy and Michalek, the special teams (especially the improved PP) continues to play well, and we'll see the even strength production go up as well.
The thing with hockey and their stats, in my opinion, is that hockey, more than pretty much any other sport, is a team game. All 4 guys on the ice have an impact on a players stats.
That's one of the main reasons why I do not like hockey stats, especially a stat like +/-.
Right, but if we have a large enough sample size and we're looking at the right stats we can usually get a good grasp of how much impact a player is making on those numbers.
People need to keep in mind though that we're dealing with incomplete information, which means two things. One, that the data probably doesn't tell the whole story. Two, and most importantly, that the data gives us a better understanding of individual player impact than not having it at all provided we interpret the data as intended. That last part eludes many posters however and as a result sours many on the concept of using statistics in their analysis. Quite a shame, really.
Right, but if we have a large enough sample size and we're looking at the right stats we can usually get a good grasp of how much impact a player is making on those numbers.
People need to keep in mind though that we're dealing with incomplete information, which means two things. One, that the data probably doesn't tell the whole story. Two, and most importantly, that the data gives us a better understanding of individual player impact than not having it at all provided we interpret the data as intended. That last part eludes many posters however and as a result sours many on the concept of using statistics in their analysis. Quite a shame, really.
I don't need a statistic to tell me something I can gain by watching. IMO, stats are used in place of watching a player far too much.
Statistics are not accompanied by perspective, and that perspective can only be gained by watching a player.
I don't need a statistic to tell me something I can gain by watching. IMO, stats are used in place of watching a player far too much.
Statistics are not accompanied by perspective, and that perspective can only be gained by watching a player.
It's not even understanding a player, but understanding the game.
Martin is not an offensive dynamo but is struggling nonetheless with I believe 3 points, while going up against the best players in the game, night in night out. His defense partner has been struggling at unfortunately the same time he is.
He rarely gets premiere PP time which is where this team is getting its production from this year.
How in God's name would he be anything but a minus player in that scenario?
It's not even understanding a player, but understanding the game.
Martin is not an offensive dynamo but is struggling nonetheless with I believe 3 points, while going up against the best players in the game, night in night out. His defense partner has been struggling at unfortunately the same time he is.
He rarely gets premiere PP time which is where this team is getting its production from this year.
How in God's name would he be anything but a minus player in that scenario?
other than the fact that you arent awarded a "plus" for a pp goal scored, i agree.
paul martin is fine. his plus minus will be fine. hes a beast - start paying attention to what he does.
other than the fact that you arent awarded a "plus" for a pp goal scored, i agree.
paul martin is fine. his plus minus will be fine. hes a beast - start paying attention to what he does.
Our ES production hasn't been what it was last year. That was my point.
Neal would be +million if PP goals counted towards plus/minus.
I don't need a statistic to tell me something I can gain by watching.
You realize that this is the same logic that baseball traditionalists used, right? In fact, just about all of the cliches that hockey fans use to dismiss statistics can be applied to baseball before sabremetrics proved otherwise. With hockey it's just about finding the right statistics, and I'm not positive we've found them yet. That doesn't mean we should stop looking or dismiss the progress we've made thus far. And yes, significant progress has been made in the use of statistical analysis over the years. Anybody who is ignoring them is doing themselves a major disservice.
Quote:
IMO, stats are used in place of watching a player far too much.
Why does it have to be one or the other? It doesn't take superhuman capabilities to use both observational and statistical analysis to form an opinion.
Quote:
Statistics are not accompanied by perspective, and that perspective can only be gained by watching a player.
You say this while ignoring all of the advantages that statistics bring to the table. For starters, they have no biases, and they are able to reliably measure whatever it is they're recording. We can use that data to find trends that lead to success, which gives us a better understanding of what drives winning. You'd have to be crazy to think that teams aren't looking to take advantage of this kind of analysis. As long as you understand the advanced statistics you're looking at, that information can only enhance your knowledge. So just because you're looking at incomplete information does not mean that it doesn't have value. It simply means you need to use that data within the correct context to come to a meaningful conclusion.
Stats also allow us to debate in a controlled medium where all of the information is readily available to use. When you get two people who are arguing based on what they saw it usually doesn't go anywhere, because realistically where can it go? I saw the sky as blue and you saw it as green. Not much of a discussion to be had except to agree to disagree.
What I find disingenuous about the people who brush off advanced statistical analysis is that they do use stats to make arguments, they just use the most basic among them, which also happen to be the most flawed. Players who accumulate goals and assists aren't particularly valuable if they give up as many on their own net, yet these are often the most highly rated players in the league because of their point totals and on ice flashiness. These players are also among the highest paid in the league. But without looking at the right statistics most people would never come to this conclusion (and many still haven't despite the overwhelming evidence of shot analysis). Yet you say that watching games tells you everything that a statistic would? Unless you are superhuman, this is just false.
A good example of observational analysis being misleading would be if you asked somebody who the better player is at even strength between Ilya Kovalchuk and Jordan Staal. Watching them play you would probably say that it's Kovalchuk by a mile. His speed, his shot, his hands, all of the physical attributes that wow you when he's on the ice are seemingly miles ahead of Staal, a player who many feel isn't particularly fast, doesn't stickhandle very well, and has a very mediocre shot. Anybody who simply watches these two play would come to the conclusion that Kovalchuk would be a much more effective player 5 on 5, however I can assure you that this is not the case in reality. Staal is far more dominant territorially at even strength and Kovalchuk would be a significant dog to have a higher plus minus than Staal over any period of time. That despite Kovalchuk likely facing easier competition and having better linemates.
In general players like Jordan Staal are criminally underrated when going by the eye test, while players like Kovalchuk get overrated due to their highly skilled nature. You see this kind of thing all the time in varying degrees and that is why observational analysis isn't infallible either. The reality is that no one form of analysis is going to fully encompass the impact of that individual, whether it be advanced statistics, basic statistics, or observing their play on the ice.
Denying statistical value in hockey at this point is like denying science. The evidence is out there. Educate yourself.
You realize that this is the same logic that baseball traditionalists used, right? In fact, just about all of the cliches that hockey fans use to dismiss statistics can be applied to baseball before sabremetrics proved otherwise. With hockey it's just about finding the right statistics, and I'm not positive we've found them yet. That doesn't mean we should stop looking or dismiss the progress we've made thus far. And yes, significant progress has been made in the use of statistical analysis over the years. Anybody who is ignoring them is doing themselves a major disservice.
Why does it have to be one or the other? It doesn't take superhuman capabilities to use both observational and statistical analysis to form an opinion.
You say this while ignoring all of the advantages that statistics bring to the table. For starters, they have no biases, and they are able to reliably measure whatever it is they're recording. We can use that data to find trends that lead to success, which gives us a better understanding of what drives winning. You'd have to be crazy to think that teams aren't looking to take advantage of this kind of analysis. As long as you understand the advanced statistics you're looking at, that information can only enhance your knowledge. So just because you're looking at incomplete information does not mean that it doesn't have value. It simply means you need to use that data within the correct context to come to a meaningful conclusion.
Stats also allow us to debate in a controlled medium where all of the information is readily available to use. When you get two people who are arguing based on what they saw it usually doesn't go anywhere, because realistically where can it go? I saw the sky as blue and you saw it as green. Not much of a discussion to be had except to agree to disagree.
What I find disingenuous about the people who brush off advanced statistical analysis is that they do use stats to make arguments, they just use the most basic among them, which also happen to be the most flawed. Players who accumulate goals and assists aren't particularly valuable if they give up as many on their own net, yet these are often the most highly rated players in the league because of their point totals and on ice flashiness. These players are also among the highest paid in the league. But without looking at the right statistics most people would never come to this conclusion (and many still haven't despite the overwhelming evidence of shot analysis). Yet you say that watching games tells you everything that a statistic would? Unless you are superhuman, this is just false.
A good example of observational analysis being misleading would be if you asked somebody who the better player is at even strength between Ilya Kovalchuk and Jordan Staal. Watching them play you would probably say that it's Kovalchuk by a mile. His speed, his shot, his hands, all of the physical attributes that wow you when he's on the ice are seemingly miles ahead of Staal, a player who many feel isn't particularly fast, doesn't stickhandle very well, and has a very mediocre shot. Anybody who simply watches these two play would come to the conclusion that Kovalchuk would be a much more effective player 5 on 5, however I can assure you that this is not the case in reality. Staal is far more dominant territorially at even strength and Kovalchuk would be a significant dog to have a higher plus minus than Staal over any period of time. That despite Kovalchuk likely facing easier competition and having better linemates.
In general players like Jordan Staal are criminally underrated when going by the eye test, while players like Kovalchuk get overrated due to their highly skilled nature. You see this kind of thing all the time in varying degrees and that is why observational analysis isn't infallible either. The reality is that no one form of analysis is going to fully encompass the impact of that individual, whether it be advanced statistics, basic statistics, or observing their play on the ice.
Denying statistical value in hockey at this point is like denying science. The evidence is out there. Educate yourself.
Stats don't always tell the right story though. You can look at that sina character who posts here to see that. Some of those stats rate Martin and Michalek poorly last season, but that was the furthest thing from the truth.
I think the advanced statistics aren't as valuable as you want them to be. You're a stats guy. We have established this. I, and a lot of others are not. I'll base my observation off of what I see...you can keep your stat sheets.
Statistics measure what they measure, not what we wish they measured. Plus/minus does not measure defensive acumen. It takes one mistake or one exceptional play by the opposition to result in a goal against. Plus/minus does not track which one of the 12 guys made that play.
Who's the better defender 2009-2010 Jeff Schultz or 2010-2011 Nicklas Lidstrom? +50 vs -2. Obviously Schultz AINEC, right?
Statistics measure what they measure, not what we wish they measured. Plus/minus does not measure defensive acumen. It takes one mistake or one exceptional play by the opposition to result in a goal against. Plus/minus does not track which one of the 12 guys made that play.
Who's the better defender 2009-2010 Jeff Schultz or 2010-2011 Nicklas Lidstrom? +50 vs -2. Obviously Schultz AINEC, right?
And a guy can have poor advanced stats. It doesn't measure a players capability though, and stats never add perspective. They are black and white. They are what they are.
If looking at a team structure, you don't go by stats...you go by a players skill set and what he can bring. For instance, if you're in need of a goal scorer on your team, are you going to look to see how territorially dominant he was? No. You're going to look for the guy who has the shot and the skill.
Stats don't always tell the right story though. You can look at that sina character who posts here to see that. Some of those stats rate Martin and Michalek poorly last season, but that was the furthest thing from the truth.
I already addressed this. No one piece of analysis will ever tell the whole story. Why should statistics be dismissed under that logic when the same thing is true of observation?
Which stats are you referring to that rated Martin and Michalek poorly last year? I'm especially curious about Martin, who is an advanced statistics dynamo.
Quote:
I think the advanced statistics aren't as valuable as you want them to be. You're a stats guy. We have established this. I, and a lot of others are not. I'll base my observation off of what I see...you can keep your stat sheets.
Several teams have already adopted this kind of analysis. In fact, the Penguins are supposedly one of the greatest advocates of advanced statistics in the NHL currently. You can continue to put on the blinders if you'd like, I prefer to be open minded. Stats, observation, all of it is essential to understanding the bigger picture.
Quote:
Originally Posted by Sideline
Who's the better defender 2009-2010 Jeff Schultz or 2010-2011 Nicklas Lidstrom? +50 vs -2. Obviously Schultz AINEC, right?
Who's the better player, 05-06 Cheechoo or 09-10 Crosby? 56 goals vs 51 goals. I guess that settles it - goals are worthless.
On a serious note, using Lidstrom to demonstate the flaws of plus minus doesn't work when he's +436 lifetime. All you've done is shown that small sample sizes have violatile results, which we already knew.
Quote:
Originally Posted by JTG
And a guy can have poor advanced stats. It doesn't measure a players capability though, and stats never add perspective. They are black and white. They are what they are.
Stats never add perspective? You do realize that there are several advanced stats which are used with the sole intention of adding perspective, right? Quality of competition, quality of linemates, offensive zone usage, relative corsi, PDO, etc. - all of these things are there to add context.
Quote:
If looking at a team structure, you don't go by stats...you go by a players skill set and what he can bring. For instance, if you're in need of a goal scorer on your team, are you going to look to see how territorially dominant he was? No. You're going to look for the guy who has the shot and the skill.
These stats are far from "science."
If you're looking for a goalscorer you would obviously look for a guy who registers a lot of shots and scores a lot of goals. I've yet to see somebody who accumulates a lot of those without having 'the shot and the skill' that you speak of.
Last edited by Malkin4Top6Wingerz: 11-14-2011 at 09:38 PM.
The plus/minus stat was never designed to be a stand-alone stat. It paints a pretty good picture when used with other numbers, though.
Martin's -11 looks ugly. He probably should be a minus player regardless though, because we haven't been great five on five so far this season and Martin routinely sees the most or second-most ice for us. So logic says he'll be a minus player. That said, using the eyeball test, I think it's fair to say Martin's played great some nights, decent some nights and fairly bad other nights. And I think we ALL can agree that he's been pretty brutal since paired with 58.
Craig Adams sees the least amount of ice among our forwards not named MacIntyre, Jeffrey, Tangradi and sometimes Vitale. So HIS -7 tells a much different tale than Martin. Especially since Adams gets only PK time, not PP time. So he'll never get a minus on special teams and can only get a plus while short-handed. In his case, I believe the +/- stat is fairer. Adams really hasn't played well this season, aside from his usual fearless PKing (something several guys could do as his replacement on this team).
Staal and Dupuis have the best +/- on the team right now, and it's hard to find two more consistent players so far this year.
Meanwhile, the Sullivan-Malkin-Neal line is a minus-6 collectively. Not a shocker. None are particularly stellar defenders, and the line is designed for all-out offense. They've scored 16 goals, so the production is there. If the minuses continue, though, and if our five-on-five play doesn't improve when Crosby returns, then I think you may see the line broken up.
That said, if that line becomes Line 2, they'll get more favorable match-ups. That'll help their +/- rating.
SO, to summarize, the plus/minus stat is a starting point for analysis, not a conclusion.
I already addressed this. No one piece of analysis will ever tell the whole story. Why should statistics be dismissed under that logic when the same thing is true of observation?
Which stats are you referring to that rated Martin and Michalek poorly last year? I'm especially curious about Martin, who is an advanced statistics dynamo.
Several teams have already adopted this kind of analysis. In fact, the Penguins are supposedly one of the greatest advocates of advanced statistics in the NHL currently. You can continue to put on the blinders if you'd like, I prefer to be open minded. Stats, observation, all of it is essential to understanding the bigger picture.
Who's the better player, 05-06 Cheechoo or 09-10 Crosby? 56 goals vs 51 goals. I guess that settles it - goals are worthless.
On a serious note, using Lidstrom to demonstate the flaws of plus minus doesn't work when he's +436 lifetime. All you've done is shown that small sample sizes have violatile results, which we already knew.
Stats never add perspective? You do realize that there are several advanced stats which are used with the sole intention of adding perspective, right? Quality of competition, quality of linemates, offensive zone usage, relative corsi, PDO, etc. - all of these things are there to add context.
If you're looking for a goalscorer you would obviously look for a guy who registers a lot of shots and scores a lot of goals. I've yet to see somebody who accumulates a lot of those without having 'the shot and the skill' that you speak of.
Hockey is a team game, and a player should never be helped or punished by something someone else does on the ice. Individual stats shouldn't be given on a team basis. In baseball and football, it's much, MUCH easier to give out individual stats.
Those stats at behindthenet.ca take +/- into consideration, along with shots on goal, shots missed, and blocked shots while a player is on the ice. So a player can have nothing to do with the play at all, and his stats are affected. I think the site uses a flawed foundation to then make statistical observations
These are not stats I, myself, can put a lot of stock in. Like I said...you can use your stats. I don't believe in them.
