View Single Post
Old
06-19-2011, 12:26 AM
  #121
Lafleurs Guy
Registered User
 
Lafleurs Guy's Avatar
 
Join Date: Jul 2007
Posts: 21,572
vCash: 500
Quote:
Originally Posted by MathMan View Post
Larry isn't under 35 either. Markov is just not going to fall off a cliff.
I don't think he'll fall off a cliff but if that knee gets hurt like it has again... he's done. Nobody knows how good he'll be or how durable that knee will be, he's not getting any younger and we aren't contenders this year anyway. Wiz makes more sense.

Quote:
Originally Posted by MathMan View Post
You're really underestimating Markov's defense. He's cerebral and positional, which isn't flashy, but it's still darned effective. You don't need to be big and hitty to be effective defensively, even if that gets you noticed more. In fact, I think hittiness can detract if you're not careful.
Lidstrom isn't big and hitty but he's fantastic defensively. Markov isn't. He's adequate defensively but he's one dimensional. He's just really good with the puck. He's a stellar offensive defenseman and that's about it. He's not Chris Pronger either, sorry.

Quote:
Originally Posted by MathMan View Post
Less than people think and, again, the key benefits of size are strength, reach and mass -- not necessarily physicality.
Right... we'll just have to disagree man. If you think that clearing the net or getting traffic in front of it has nothing to do with strength or toughness, I don't know what to say to you. Then again, you don't seem to value those traits anyway.

Quote:
Originally Posted by MathMan View Post
Spacek gets a bad rap because he tries to move the puck and occasionally turns it over. Which is what all puck-movers do to some extent; if you're Hal Gill and never pass the puck, you'll never get caught making a bad pass.

He is, however, capable of playing against top opposition and being effective in that role. People often forget he and Hamrlik held the fort together as the first pairing when Markov went down. The big problem he has is that when he's effective it's not showy and when he occasionally screws up, it's blatant, and because he always plays top opposition, every mistake he makes gets more . A common malady among defensemen who play top opposition. Atlanta fans think Bogosian is terrible for much the same reason.
Sorry, but whenever I've see Spacek, I have not been impressed. I'd be very happy to dump him. He's definitely a guy we could replace with somebody with sandpaper.

Quote:
Originally Posted by MathMan View Post
The point of Scott Gomez is having the puck on your guys' sticks in the opposing zone, rather than in your zone on the other guys' sticks. I don't think it's stretching ones' imagination to suggest this may result in more goals for than against.

Whether he personally gets points or not for it, I could care less. He could be the uncounted third assist on every one of Cammy's goals for all I care. His job is to get his team to generate scoring chances and prevent chances against.



If they shoot 9.5%, they may well end up being the highest-scoring team in the league. Just like 2007-2008. I insist on stressing that I see this as highly unlikely, however.



The point is, injuries can derail any team. Surely if Datsyuk and Zetterberg got injured, you wouldn't expect the Red Wings to be as good offensively, right?

That said, the Habs have had enough chances to be a top-third offensive club without Markov this year anyway, so they do have a fair amount of depth.



I don't think I ever did. I see this as the high end for him. It would require more PP scoring than he usually has.



It already includes goals. You want to give goals extra weight. But the point of Corsi is to measure puck possession, and eliminate the other factors that may or may not lead to a goal, many of which lie outside the player's control (goaltending foremost among them).

What you suggest would be like measuring puck possession in time, but giving a 10-second bonus for a goal for and a 10-second penalty for a goal against. It wouldn't be a puck possession measure anymore.

Or it's like measuring shots on goals, but giving a 10-shot bonus for goals. What would be the point of that?
I'm not going to address this anymore. In the other thread I said I'd leave this be and let you have the last word on things so I won't debate any further on this here.


Quote:
Originally Posted by MathMan View Post
GVT is probably more like what you're looking for here. As you like to point out, hockey is more complex than baseball, so a wins-over-replacement stat, while it would pretty much be the holy grail, is much harder to determine. GVT is an early attempt, but it doesn't factor in context (strength-of-opposition, strength-of-teammates) and it doesn't differentiate between sustainable performance elements and transient ones.

That's why a lot of the research focuses on things like puck possession -- that, and it's generally acknowledged that puck possession is a key sustainable skill for winning hockey teams.

Counting scoring chances is probably the most accurate measure we have -- but there's just not enough people available to watch all the games and painstakingly count them for all teams. And there's a high correlation between scoring chances counts and puck possession at the macro level.



I really suggest that you read Gabe Desjardins's blog at behindthenethockey.com if the subject interests you. He's one of the foremost guys and he routinely links to other sites that do research.

EDIT: Thanks for the research Roke.
I will check it out.
Quote:
Originally Posted by MathMan View Post
I also don't think it does a very good job with distinguishing transient versus sustainable contribution (aka luck vs. skill). I wonder if VUKOTA (the projection scheme based on GVT) takes that into account?
I will look at this as well.
Quote:
Originally Posted by Roke View Post
Lafleurs Guy, here's a link to the GVT (goals-versus-threshold) stat that MathMan was mentioning though it hasn't had an update since March 7th for some reason. There's also a primer about what it is and how it is calculated[/url] (clicking next column at the bottom of the article will bring you to parts II and then III). If you have some spare time and want to check it out that would be a decent start.

When I started reading about microstats it was the easiest stat for me to understand and it factors in offensive and defensive value to a degree making it somewhat holistic. It's main weak points are it doesn't account for quality of opposition where on the ice a player starts their shifts but it's a good first stat.

I wouldn't be comparing the value of goaltenders to the value of skaters in using it though - goaltenders tend to be rated highly across the board and it's the marginal difference between them that is more important.

It's a good stepping stone into the hockey microstat world though if you're inclined to want to read up more about it.
Okay thanks, I will check this out.

Lafleurs Guy is offline   Reply With Quote