Laughton-Giroux-Voracek
JvR-Hayes-Konecny
Raffl-Couturier-Pitlick Strictly to shutdown
Grant-Thompson-NAK

Easy to hate on Braun, but he has done some good things in this series. Broke up what was probably a sure goal last night. And he's been taking the punishment being 1st guy back on dump ins. With all due respect to the rest of the Flyers dmen, they're not as willing to eat glass as JB.
- Tomahawk

Niskanen and Provorov pretty adept in that department if the announcers are correct

Niskanen and Provorov pretty adept in that department if the announcers are correct
- easY

In my opinion Sean Couturier isn’t our most important player to have in the line up. He’s not even second. I think it’s Carter Hart and than Ivan Provorov.

This guy Demko is pretty good, isn't he?

In my opinion Sean Couturier isn’t our most important player to have in the line up. He’s not even second. I think it’s Carter Hart and than Ivan Provorov.
- SuperSchennBros

I agree. Coots is probably 3 although I think Hayes is giving him a real run for his money.

I agree. Coots is probably 3 although I think Hayes is giving him a real run for his money.
- hereticpride

I might have to agree. It’s clear Hayes isn’t on the same level as Coots defensively but Hayes seems to work well with what he has.

Of course but if what is available on the internet is so valuable, then why are they hiring stats guys? Why don't team just log onto the web? Again, it's not the science that is suspect. It is the data.
- MJL

They're hiring those guys so the subsequent iterations of their methods/models become proprietary. The public stuff is based on the work those guys shared before they got hired. So whatever the teams are using are rooted in the stuff we have available. Nothing is getting built from scratch.

They're hiring those guys so the subsequent iterations of their methods/models become proprietary. The public stuff is based on the work those guys shared before they got hired. So whatever the teams are using are rooted in the stuff we have available. Nothing is getting built from scratch.
- Tomahawk

Completely disagree. Anything publicly available online is easily replicated. They're hiring those guys to try and gain an advantage. To try and develop new methods of statistical evaluation. Reality is that relatively speaking, what is available publicly is archaic and not deeply used by NHL teams. What we've also seen is that NHL GM's who are rooted in analytics, don't have any real advantage. Chayka has been pretty bad and Dubas doesn't know how to build a team.

Completely disagree. Anything publicly available online is easily replicated. They're hiring those guys to try and gain an advantage. To try and develop new methods of statistical evaluation. Reality is that relatively speaking, what is available publicly is archaic and not deeply used by NHL teams. What we've also seen is that NHL GM's who are rooted in analytics, don't have any real advantage. Chayka has been pretty bad and Dubas doesn't know how to build a team.
- MJL

Taking 'the guy' off the market is the advantage. Anything he/she produces from that point forward becomes yours.

In many cases, these guys aren't creating new on ice events to count, they're developing models/algorithms using existing stats to create in-house aggregates like xG, win-shares, etc to evaluate performance.

(Archaic is kind of a weird term, considering analytics is relatively new to hockey still.)

Chayka quit, so who knows, and that team had more problems than just who the GM was. But TO is a young team on the rise. Like the Flyers. It's too early to tell what they are yet. And its not like either guy only signed analytics darlings, so like everybody else, the numbers are just part of what they use to make decisions.

Taking 'the guy' off the market is the advantage. Anything he/she produces from that point forward becomes yours.


- Tomahawk

Yes exactly but they aren't taking them off the market for corsi fenwick and expected goals. As I clearly said, they are taking them off to develop new propietary stats.

In many cases, these guys aren't creating new on ice events to count, they're developing models/algorithms using existing stats to create in-house aggregates like xG, win-shares, etc to evaluate performance.


- Tomahawk

This makes zero sense. Who said or thought that they would be creating new ice events? How can a stats guy create on ice events. You didn't think this comment through very well.

(Archaic is kind of a weird term, considering analytics is relatively new to hockey still.)


- Tomahawk

No they're not. They've been around for a significant period of time. Here is a quote from Bob McKenzie in 2014 on how NHL teams are already past using Corsi

"“The thing that makes me laugh a little bit about the whole analytics debate, in the media the old school media vs. the new school if that’s what you want to call it. People are still debating Corsi. I mean, the analytics in the National Hockey League team by team for the most part are so beyond Corsi that it’s not even a talking point. Corsi’s just a basic metric that some people use, some people don’t.

The stuff that teams are doing, I don’t want to say it’s more sophisticated because it is shrouded in mystery. But there’s no question that it’s more sophisticated because they link things more to video and break things down more analytically that way. I kind of chuckle because this whole argument over Corsi, Fenwick and PDO and all these things. Most of the teams that I’m aware of are years past the whole Corsi debate.”

Chayka quit, so who knows, and that team had more problems than just who the GM was. But TO is a young team on the rise. Like the Flyers. It's too early to tell what they are yet. And its not like either guy only signed analytics darlings, so like everybody else, the numbers are just part of what they use to make decisions.
- Tomahawk

Chayka quitting has nothing to do with his performance. It was poor. It's not too early to tell what the Leafs are. They're a top heavy team, maxed out to the cap that is overall poorly constructed. These are the analytic darlings. The guys who were going to give these teams a big edge. Hasn't materialized. You're right, numbers are PART of the decision making. A small part.

Yes exactly but they aren't taking them off the market for corsi fenwick and expected goals. As I clearly said, they are taking them off to develop new propietary stats.
- MJL

I don't think you really read what I wrote very carefully.

AFAIK, shot-attempts are still the backbone of everything. Each team will weigh it against their own measure of shot quality, and that will be their secret sauce. Just like expected goals is one model, teams have their own. These aren't 'new proprietary stats, they're just proprietary models. One team may value low-high point shots more, or some teams may take into account pucks crossing the golden road, etc, but it's still the same basic thing.

This makes zero sense. Who said or thought that they would be creating new ice events? How can a stats guy create on ice events. You didn't think this comment through very well.
- MJL

It's been said in the past that teams will track things like board pins or whatever. Eventually when puck-tracking and player-tracking becomes a reality, they'll be able to track all sorts of new things like puck-control quality, pass-stability, stick-speed, etc. Until then, they're all still counting shot attempts and where they took place and how.

No they're not. They've been around for a significant period of time. Here is a quote from Bob McKenzie in 2014 on how NHL teams are already past using Corsi

"“The thing that makes me laugh a little bit about the whole analytics debate, in the media the old school media vs. the new school if that’s what you want to call it. People are still debating Corsi. I mean, the analytics in the National Hockey League team by team for the most part are so beyond Corsi that it’s not even a talking point. Corsi’s just a basic metric that some people use, some people don’t.

The stuff that teams are doing, I don’t want to say it’s more sophisticated because it is shrouded in mystery. But there’s no question that it’s more sophisticated because they link things more to video and break things down more analytically that way. I kind of chuckle because this whole argument over Corsi, Fenwick and PDO and all these things. Most of the teams that I’m aware of are years past the whole Corsi debate.”
- MJL

Yes, NHL teams have been tracking random things forever as in-house stats, but predictive analytics is a new phenomenon. The modeling is what's new. Teams weren't hard into the math of it. Now they are.

Chayka quitting has nothing to do with his performance. It was poor. It's not too early to tell what the Leafs are. They're a top heavy team, maxed out to the cap that is overall poorly constructed. These are the analytic darlings. The guys who were going to give these teams a big edge. Hasn't materialized. You're right, numbers are PART of the decision making. A small part.
- MJL

It's funny how Dubas and Chayka are held up as poster boys when they're basically doing the same thing as the other 29-GMs.

Leafs have multiple franchise players. Tavares, Matthews and Marner are all earning every penny of their contracts. They are getting their moneys worth out of Nylander too. Rielly is on a Gostisbere type contract giving them legit 1D quality play. A lot of old/bad money is about to come off the books and they have a lot of youth filtering in. They're not unlike where the Flyers at the tail end of Hexy's tenure. Carter Hart changed everything for us. If the Leafs can find some stability in net, they'll be fine too. The Pens have been running with a top-heavy salary structure for years and were able to make it work.

I don't think you really read what I wrote very carefully.

AFAIK, shot-attempts are still the backbone of everything. Each team will weigh it against their own measure of shot quality, and that will be their secret sauce. Just like expected goals is one model, teams have their own. These aren't 'new proprietary stats, they're just proprietary models. One team may value low-high point shots more, or some teams may take into account pucks crossing the golden road, etc, but it's still the same basic thing.


- Tomahawk

No, it's not the same basic thing. You're simply wrong. If it was, they wouldn't need to hire an analytics guy.

It's been said in the past that teams will track things like board pins or whatever. Eventually when puck-tracking and player-tracking becomes a reality, they'll be able to track all sorts of new things like puck-control quality, pass-stability, stick-speed, etc. Until then, they're all still counting shot attempts and where they took place and how.


- Tomahawk

That's not creating new ice events. That is impossible! Those things are already a reality. That is what you don't understand. This is the same conversation we've had many times. Teams are using video and software to collect better data. Raw shot data is archaic.

Yes, NHL teams have been tracking random things forever as in-house stats, but predictive analytics is a new phenomenon. The modeling is what's new. Teams weren't hard into the math of it. Now they are.


- Tomahawk

You're wrong here. You're way behind the times. Random things. You're just making it up. Predictive analytics or the attempt to is nothing new. Analytic guys have been writing articles about attempting to do so for years, since the beginning. Teams aren't hard into the math of it, they're hard into the stats. The math is left up to the analytics guy. Do you think any coach wants to look at the math?

It's funny how Dubas and Chayka are held up as poster boys when they're basically doing the same thing as the other 29-GMs.

Leafs have multiple franchise players. Tavares, Matthews and Marner are all earning every penny of their contracts. They are getting their moneys worth out of Nylander too. Rielly is on a Gostisbere type contract giving them legit 1D quality play. A lot of old/bad money is about to come off the books and they have a lot of youth filtering in. They're not unlike where the Flyers at the tail end of Hexy's tenure. Carter Hart changed everything for us. If the Leafs can find some stability in net, they'll be fine too. The Pens have been running with a top-heavy salary structure for years and were able to make it work.
- Tomahawk

What don't you understand? As analytic darlings, they're supposed to have an advantage over other teams. It hasn't happened. Chayka was pretty bad and quit before he was fired. Dubas has got his team in a jam. Why can't he uncover hidden cheap gems with his analytic advantage? His team is an a cap jam and they lack depth. The defense is poor. It's a poorly constructed team.
Yet the old dog, out to pasture GM Lou Lamoriello has his team one game away from the conference finals.

That's not creating new ice events. That is impossible! Those things are already a reality. That is what you don't understand. This is the same conversation we've had many times. Teams are using video and software to collect better data. Raw shot data is archaic.
- MJL

Current video + AI event tracking is limited in scope and capability. It's basically a more efficient/precise method of tracking events than having humans do it manually, and still an approximation. It's not giving you the physics and precision that chips in pucks and on players would.

What don't you understand? As analytic darlings, they're supposed to have an advantage over other teams. It hasn't happened. Chayka was pretty bad and quit before he was fired. Dubas has got his team in a jam. Why can't he uncover hidden cheap gems with his analytic advantage?
- MJL

Analytics have already played a part in winning Stanley Cups for various teams. Why ignore that fact and only talk about two teams?

Current video + AI event tracking is limited in scope and capability. It's basically a more efficient/precise method of tracking events than having humans do it manually, and still an approximation. It's not giving you the physics and precision that chips in pucks and on players would.


- Tomahawk

In theory. The point is that analytics with NHL teams are far more advanced than the analytics we have available to us as fans. It all points to one thing. Analysis made by fans and pundits who are paid to do so using analytics is very suspect.

Analytics have already played a part in winning Stanley Cups for various teams. Why ignore that fact and only talk about two teams?
- Tomahawk

I don't recall at any point in time stating that analytics aren't used as part of teams player development and selection. I'll repeat it again. Those two teams claimed that they were going to heavily use analytics to gain an advantage over the rest of the league. That's why they hired each guy. It hasn't panned out. It's just more evidence that analytics are only a small part of running an NHL team and that human intelligence, is still more valuable.

In theory. The point is that analytics with NHL teams are far more advanced than the analytics we have available to us as fans. It all points to one thing. Analysis made by fans and pundits who are paid to do so using analytics is very suspect.




I don't recall at any point in time stating that analytics aren't used as part of teams player development and selection. I'll repeat it again. Those two teams claimed that they were going to heavily use analytics to gain an advantage over the rest of the league. That's why they hired each guy. It hasn't panned out. It's just more evidence that analytics are only a small part of running an NHL team and that human intelligence, is still more valuable.
- MJL

Not any more suspect than anything else, really. Do I take a take a Tanner blog w/ a huge barrel of salt? You bet I do. Do I also still believe that "insiders" always know best? Not after watching Chia & Co.'s debate over Seguin, lol.

Intelligence seems to be at a premium in hockey reporting, forum discussion AND NHL front offices.

You also can't single out instances like Chayka to discredit the analytics movement, just like analytics people shouldn't hold up the way AMac and Lecavalier's contracts worked out for the Flyers as irrefutable proof that the numbers are ironclad.

Not any more suspect than anything else, really. Do I take a take a Tanner blog w/ a huge barrel of salt? You bet I do. Do I also still believe that "insiders" always know best? Not after watching Chia & Co.'s debate over Seguin, lol.

Intelligence seems to be at a premium in hockey reporting, forum discussion AND NHL front offices.


- Tomahawk

What's anything else? A good analyst like Bill Meltzer can analyze a game and players without using analytics same as a good coach and scout can. Then you have analysts who think they can analyze any player or team simply using nothing but analytics. I know which is best. Every GM makes mistakes. Some for good reasons and some for bad reasons.

You also can't single out instances like Chayka to discredit the analytics movement, just like analytics people shouldn't hold up the way AMac and Lecavalier's contracts worked out for the Flyers as irrefutable proof that the numbers are ironclad.
- Tomahawk

It's not about discrediting analytics. Why don't you get that? Analytics are a useful tool if used properly. Otherwise NHL teams wouldn't devote resources to it. What I'm discrediting is analytics over human intelligence. What I'm discrediting is discussions where analytic numbers always trump human insight. I'll repeat again, the numbers we have access to as fans, are mediocre at best.

What I'm discrediting is discussions where analytic numbers always trump human insight. I'll repeat again, the numbers we have access to as fans, are mediocre at best.
- MJL

Nobody with any sense ever argued that numbers were anything more than a check on anecdotal decision making. More information, better decisions.

Anyway, there's no point in getting into circular arguments about this.

Nobody with any sense ever argued that numbers were anything more than a check on anecdotal decision making. More information, better decisions.

Anyway, there's no point in getting into circular arguments about this.
- Tomahawk

Again, we have a fundamental disagreement. In making your first statement here, you're implying that analytics are better than human intelligence in decision making. They're not. They're a supplemental tool to aid decision making. Nothing more. It's not a circular argument to me.

In making your first statement here, you're implying that analytics are better than human intelligence in decision making..
- MJL

Nope.

AV sucks. Nada mas.

Nope.
- Tomahawk

Yes, you did. It can't be a check on that human intelligence if it is not superior. Common sense.

Nope.
- Tomahawk

Yes, you did. It can't be a check on that human intelligence if it is not superior. Common sense.

Yes, you did. It can't be a check on that human intelligence if it is not superior. Common sense.
- MJL

Are you being serious?

For example, US congress and the judiciary are checks on the executive... are they superior to the power of the president? No, they're designed to be coequal branches of gov't.

I dunno man, seems like you're trying really hard to make this more contentious than it has to be. We actually agree that stats and intuition should work in balance.

Are you being serious?

For example, US congress and the judiciary are checks on the executive... are they superior to the power of the president? No, they're designed to be coequal branches of gov't.

I dunno man, seems like you're trying really hard to make this more contentious than it has to be. We actually agree that stats and intuition should work in balance.
- Tomahawk

So now you're comparing government over sight to hockey analytics? Clearly you are the one who is being contentious. Analytics are not a check, they're a tool and one part of player and team evaluations. That's what you don't get. You've put the cart before the horse. Human intelligence is still far superior and likely always will be.

So now you're comparing government over sight to hockey analytics? Clearly you are the one who is being contentious. Analytics are not a check, they're a tool and one part of player and team evaluations. That's what you don't get. You've put the cart before the horse. Human intelligence is still far superior and likely always will be.
- MJL

smh