[News] USA Olympic Roster announced

[Tinordi]

I feel like I started a tempest. All I'm saying is that it's just a bit concerning what Burke seems to prioritize and his methods for doing so. I'm still giving him the benefit of the doubt and we wont know how he really will approach this rebuild until he's had time to hire a GM and make some trades up to the deadline. But, articles like this do lead me to be a bit suspicious of his approach.

[19Yzerman19]

Originally Posted by SuperMatt18

I agree 100% but there is no proof that those teams used advanced stats to be good at advanced stats. My point is more that Chicago may use advanced stats but I don't think they drafted Toews and Kane because they had a strong Corsi in junior.

That's not the point. The point is that the results are clear: being good at puck possession, building your team so that the roster you ice has a high Corsi%, is something that GMs should strive for, because teams that have a high Corsi% are more successful than others. Good possession teams make the playoffs, bad ones tend not to, so if you're running a team, you should probably try as hard as possible to fall into the former category.

The point is advanced stats are helpful but shouldn't be the be all end all for determining if a player is good.

I agree with this. For example, Justin Williams is consitently one of the best possession players in the league, but I'd still rather have Phil Kessel. However, it certainly is a strong attribute to have, and since there are only so many Phil Kessel level talents in the league, it's absolutely an important factor to consider when making personnel decisions.

[Oling_Roachinen]

Is the Burke vs Stats thing coming from the one joke Burke made?

They all were joshing with each other, Burke wasn't the only one to make fun of his 'thoroughness':

http://espn.go.com/olympics/hockey/s...pics-was-named

Lombardi jokingly asks if he wants any written reports.

"No, orally. I don't want you to write anything else down," Poile quips as the call comes to an end.

[Oling_Roachinen]

When it comes to the Yandle decision there's very little for me to worry about. First it's important to realize that they are discussing some of the top defensemen in the world. I'm sure Burke could find a spot for Yandle on the Flames.

Secondly, there's no reason to believe Burke discounted anything. Yandle's merit to the team was his puck-moving and offensive capabilities. It was far from just Burke who had issues with his defense (you can tell the coaches didn't want Yandle on the team either). And even his biggest supporter in Lombardi didn't have much faith in his defensive abilities

"If you ask our coaches," Lombardi says, "Yandle scares us more than Shattenkirk."

"He's not great defensively but he's not soft," Lombardi continues.

So then you have to weigh if his great offense is worth his defensive liabilities. Burke wasn't the only one to question it, obviously, especially considering that they intended to have Pavelski play the point on the PP along with Suter and Shattenkirk, while Fowler and well their entire defense is capable of playing on the PP. So they didn't seem to need a PP specialist.

And then there's just the fact that Burke was far from the only one who suggested it.

"What if we just lose our first couple of games? We're just no [darn] good," Bowman says.

You know guys such as Suter, McDonagh and Martin will bounce back, he says. But the coaches need to be able to move guys around as they see fit if the team doesn't play well early in the tournament.

If that's the case, then it's not a bad thing to have guys ranked four through eight who can be mixed around.

Stan Bowman wants to be able to mix and match if needed. Yandle, a player they all agreed they don't quite have faith in his defense, would not be able to play in the top 4. If there was an injury they wouldn't feel comfortable having him move up the depth chart. It wasn't just Burke saying this stuff, it seemed closer to everyone but Lombardi.

Poile reports that the coaches feel that based on current level of play Fowler is the best for them. He's a better all-around player and can play in more situations in their estimation.

If there is an injury in the top group, they would feel more comfortable moving Fowler up the depth chart than Jack Johnson or Yandle.

Although the Yandle issue was discussed in depth in California, the coaches don't believe there is that big a drop-off in offensive talent if Yandle isn't on the roster, plus they are planning to use Joe Pavelski on the point, which will minimize the need for a purely offensive defenseman in man-advantage situations.

Johnson vs Ryan is one thing, but I think anything to do with Burke and Yandle is being blown out of proportion.

[Red John]

Originally Posted by 19Yzerman19

This is not only wrong, it doesn't even make any sense. Of course stats describe what happened on the ice. Regardless of the stat, that's the case. The "goals" stat describes who scored how many goals. Plus / Minus describes who was on the ice for goals for their team and against their team.

It makes tons of sense actually. If you watch a game and your observation is "Joe Shmuck has a great scoring touch" then at the end of the night you look down and see he had 2 goals, that's great it corroborates your observation. But if you skip the game and see the boxscore, you are missing the context. What if one goal was an empty netter? What if the defenseman fell on a routine play allowing an easy tap in? What if he scored both goals being matched against the 4th line? By looking at stats only you only see the result - not what actually happened on the ice.

So-called advanced stats are more predictive simply because goals themselves are a rare enough event that a certain amount of randomness plays into whether a goal is created, whereas shot totals create a larger sample size and as a result, you can discern a clearer pattern that has less randomness and noise. Which is why they're predictive of who's going to be successful at the end of the year - see quote from Kent below.

There's certainly a correlation between being GOOD at "advanced stats" and being successful. From Kent Wilson of Flames Nation:

Again, oftentimes over a large sample size, stats will often corroborate what happened on the ice. Looking at Joe Shmuck again, if he has a good scoring touch regardless of situation over a long period of time his goal stat line will likely reflect that fact. "Advanced stats" takes shots instead of goals into effect because it gives a larger sample size as you said.

But it still doesn't actually describe what happened on the ice - only the result.

[Inglewood Jack]

sigh...nobody is ever going to be coerced to the other side in a stats vs. intuition argument. that is about as likely as it was in the owners vs. players sides during the lockout.

I think it's time for us stat nerds to just start our own "Advanced Fricken' Statistics" or "Fenwick Close Appreciation" threads so we can analyze possession metrics in peace.

[PeteMoss]

Originally Posted by 19Yzerman19

That's not the point. The point is that the results are clear: being good at puck possession, building your team so that the roster you ice has a high Corsi%, is something that GMs should strive for, because teams that have a high Corsi% are more successful than others. Good possession teams make the playoffs, bad ones tend not to, so if you're running a team, you should probably try as hard as possible to fall into the former category.

I agree with this. For example, Justin Williams is consitently one of the best possession players in the league, but I'd still rather have Phil Kessel. However, it certainly is a strong attribute to have, and since there are only so many Phil Kessel level talents in the league, it's absolutely an important factor to consider when making personnel decisions.

Here's the thing... how do you build a team with high corsi? In theory would you not build the Oilers? A bunch of puck possession guys who can dangle and hold the puck?

Baseball... which kicked off the advanced stat brigade (moneyball, etc) is a series of one on one events so you can easily pull data and make decisions that way. It was also created by a bunch of true mathematical/statisical minds. Now you've got every two bit internet guy screaming... you need possession! That's great... but is it actually a skill a player can possess?

You look at Calgary's CORSI here - http://www.behindthenet.ca/nhl_stati...34+45+46+63+67 - TJ Galiardi is one of the best guys, Jiri Hudler is in the bottom tier. That is a small example... but its why using the current advanced stats out there on an individual player basis is stupid. I'm sure Iginla and Bouwmeester's CORSI/FENwick or whatever else went from bad to good as soon as they left the Flames and were on better teams last year. Its way too dependent on your teammates.

I will close with this... Corsi and these advanced stats are clearly good for telling what happened and can predict on a team level what may happen. But until they can tell you how to fix it... they are useless for a GM beyond saying 'well we are being really lucky or unlucky right now and we might want to make a move to based on that'

[19Yzerman19]

Here are two really good articles about the two most talked about "snubs".

Travis Yost debunks the notion that Bobby Ryan is bad defensively and doesn't produce on the power play: http://www.hockeybuzz.com/blog.php?post_id=56874

Again, I'm always wary about using goal-based analysis, but a lot of these guys now have thousands of even-strength minutes. At some point, there should be an overlapping as to where shots translate into goals, particularly with years of data. As you might expect, most US guys are plus-percentage players at EV. But, even against the team average, Ryan's mere decimal points from the top-three here. If Ryan was the defensive liability Team USA's pegged him out to be here, why does it not show?

Tyler Dellow discusses why Jack Johnson isn't very good at hockey and shouldn't have been considered in the first place: http://www.mc79hockey.com/?p=6538

Jack Johnson’s a pretty solid example of the latter type. Hockey people seem to really like him. He was drafted third overall. He was traded for a decent return to Los Angeles. Dean Lombardi gave him a lot of money. Columbus traded Jeff Carter for him. Still, the stats on the guy are amazing. Basically everyone who he plays with, he makes them worse. His teams invariably do better at 5v5 when he’s not on the ice. If Corsi% is too complex for you, this little stat is pretty amazing: of the 231 defencemen who played at least 2000 5v5 minutes between 2007-13, Johnson is 223rd in terms of his team’s share of the goals when he’s on the ice: 42.8%. There is nothing good that you can say about him. Not a thing.

[strombad]

Originally Posted by 19Yzerman19

Here are two really good articles about the two most talked about "snubs".

Travis Yost debunks the notion that Bobby Ryan is bad defensively and doesn't produce on the power play: http://www.hockeybuzz.com/blog.php?post_id=56874



Tyler Dellow discusses why Jack Johnson isn't very good at hockey and shouldn't have been considered in the first place: http://www.mc79hockey.com/?p=6538

Clearly two brilliant hockey minds that we should listen to over a team of hockey executives that include Lombardi, Poile, and Burke.

The personal blog market is just brimming with NHL executive level talent.

[Tinordi]

ad hominem attack, the last refuge of the ignorant

[19Yzerman19]

Originally Posted by Red John

It makes tons of sense actually. If you watch a game and your observation is "Joe Shmuck has a great scoring touch" then at the end of the night you look down and see he had 2 goals, that's great it corroborates your observation. But if you skip the game and see the boxscore, you are missing the context. What if one goal was an empty netter? What if the defenseman fell on a routine play allowing an easy tap in? What if he scored both goals being matched against the 4th line? By looking at stats only you only see the result - not what actually happened on the ice.

What you're describing is exactly the point of these so-called "advanced" stats. You're describing internal, random events within a game. What possession numbers are designed to do is to track much larger sample sizes (an entire season) of much larger sets of events (shots instead of goals). And yes, many of these stats do take into account things like playing against the fourth line - quality of competition is a very frequently used stat, and will often be combined with possession to adjust a player's possession statistics to account for how good the players he plays against are.

These guys aren't stupid, if you've thought of a flaw in the methodology that seems obvious to you, they've probably also thought of it and either tested it and found that it doesn't actually impact the results, or they've tried to find a way to account for it.

Originally Posted by PeteMoss

Here's the thing... how do you build a team with high corsi? In theory would you not build the Oilers? A bunch of puck possession guys who can dangle and hold the puck?

No. The Oilers are a terrible possession team. They're 27th in the league. Why would it lead you to build that team? If I'm taking over as GM and want to improve the team I've just inherited from the perspective of puck possession, I'd probably sign guys who are a) good at it, and b) see their teams generate more shots when they're on the ice as opposed to when they're on the bench. There are lots of ways to look at this; pure fenwick, WOWY (measuring whether a player does better or worse with or without a particular teammate), Corsi Rel, etc. In other words, I'd probably sign guys like Justin Williams and Logan Couture.

[strombad]

Originally Posted by Tinordi

ad hominem attack, the last refuge of the ignorant

I've never seen someone so aptly describe their own post before, brilliant!

Don't be so hard on yourself, I don't think you're ignorant.

[Oling_Roachinen]

Originally Posted by 19Yzerman19

Tyler Dellow discusses why Jack Johnson isn't very good at hockey and shouldn't have been considered in the first place: http://www.mc79hockey.com/?p=6538

Going by the same criteria Leopold is ahead of Karlsson over the last 3 years. And O'Brien is ahead of Giordano.

I think that's all we need to know about that arbitrary stat.

[Red John]

Originally Posted by 19Yzerman19

What you're describing is exactly the point of these so-called "advanced" stats. You're describing internal, random events within a game. What possession numbers are designed to do is to track much larger sample sizes (an entire season) of much larger sets of events (shots instead of goals). And yes, many of these stats do take into account things like playing against the fourth line - quality of competition is a very frequently used stat, and will often be combined with possession to adjust a player's possession statistics to account for how good the players he plays against are.

Thing is these events aren't actually random. A hockey game is full of decisions every single second - decisions about what angle to take to the puck when skating, decisions about whether to make a head fake, tons of tiny little things that go on. Each of these tiny decisions ultimately makes an impact on the final outcome. People who have played the game at a high level understand this.

Basic stats people would say the only thing that matters is goals. Advanced stats people have taken it one step further and trying to track shots since shots produces goals. Problem is, what produces shots? Skating, stickhandling, passing, and tons of little decisions that happen with every player on the ice at every moment of a game. And there is no stat out there that can track those decisions.

That's the problem, and that's why most of the "stats crowd" are basement bloggers, or at the very least not the ones who played the game and are working in the game - because they are trying to reverse engineer the game and use the final result to try and quantify the process. Instead of watching the process unfold with their eyes, drawing conclusions based on those observations, and letting the results fall where they may. If their observations are accurate, most of the time the results will provide confirmation.

[trackercowe]

Burke sure seems to be taking the brunt of this from the media, as no matter where you turned to today his quotes and opinion were being discussed. I think he could have handled things differently, but really don't have much issue with it. It's not like he was the only person making the decision, he's just the most prominent member making his voice heard. I wouldn't be at all surprised if other people on the selection panel said worse things about Ryan or Yandle, but Burnside paid special attention to Burke because of the reputation he has.

One thing I heard on McCown from that idiot Sid Seixeiro is that players in Calgary should be worried about what Burke thinks of them behind closed doors after what he said about Ryan. Well seeing how honest Burke is I think most players know full well what Burke thinks of them, and what they need to improve. He's not going to sugar coat things, like say Oilers management. It's probably for the best, as if there's any team out there that needs to be realistic about where they're at it's the Calgary Flames.

[Dagger]

Originally Posted by SuperMatt18

I agree 100% but there is no proof that those teams used advanced stats to make the decisions that allow them to be good at advanced stats.

My point is more that Chicago may use advanced stats but I don't think they drafted Toews and Kane because they had a strong Corsi in junior. Or that they use advanced stats to make personel decisions because that would mean they should trade Patrick Kane for Backlund since Backlund has a much better relative corsi then Patrick Kane.

The point is advanced stats are helpful but shouldn't be the be all end all for determining if a player is good.

It's how they supplant their young talent with veterans. That's largely where LA, Chicago and Pittsburgh have done well where as Burke's impatient approach has proven to be ineffective.

[Dagger]

Originally Posted by trackercowe

Burke sure seems to be taking the brunt of this from the media, as no matter where you turned to today his quotes and opinion were being discussed. I think he could have handled things differently, but really don't have much issue with it. It's not like he was the only person making the decision, he's just the most prominent member making his voice heard. I wouldn't be at all surprised if other people on the selection panel said worse things about Ryan or Yandle, but Burnside paid special attention to Burke because of the reputation he has.

One thing I heard on McCown from that idiot Sid Seixeiro is that players in Calgary should be worried about what Burke thinks of them behind closed doors after what he said about Ryan. Well seeing how honest Burke is I think most players know full well what Burke thinks of them, and what they need to improve. He's not going to sugar coat things, like say Oilers management. It's probably for the best, as if there's any team out there that needs to be realistic about where they're at it's the Calgary Flames.

I think it would have been more headline grabbing if someone like Poile, Shero or Lombardi came out and said things as strongly as Burke. But in the piece they were at least portrayed as more interested in contributing their opinions in a professional manner, whilst Burke was more interested in soundbites.

[d_phaneuf]

Originally Posted by strombad

Clearly two brilliant hockey minds that we should listen to over a team of hockey executives that include Lombardi, Poile, and Burke.

The personal blog market is just brimming with NHL executive level talent.

they are not giving their opinion

they are showing numbers, numbers don't lie, people's bias/eye tests do

[CliffFletcher]

Anyone who thinks Burke isn't cognizant of every word he utters when there's a reporter in the room needs to give his head a shake.

[Oling_Roachinen]

Originally Posted by diane_phaneuf

they are not giving their opinion

they are showing numbers, numbers don't lie, people's bias/eye tests do

For the Johnson one, here's the list he was using:
http://stats.hockeyanalysis.com/rati...T&sortdir=DESC

Starts off nice, Lidstrom top of the charts is a good indication. And then it's Huskin. Lepisto is right there at the top...I think we all remember when he got snubbed of his Norris.

Follow down the list and you have Weber and Doughty sandwiched right in between McKee and Meszaros. Exactly what you would expect.

The fact he used that list of an indication that Johnson is so terrible is bias and is lying, just under the guise of numbers.