News Article: Leafs AGM Dubas: Banking on hot shooting, goaltending is not a recipe for success

[Bomber0104]

What conclusions about what players?

Around the league, Corsi has a surprisingly good accuracy at gauging player ability to maintain possession in the offensive zone.

I will definitely agree with you that it is much more telling when measured as a team stat, but you're underselling it a lot when trying to measure individuals.

Scenario:

Team A is hemming Team B in their own end. Taking lots of shots attempts and CORSI'ing the hell out of Team B.

Team A's center cycles attempts to pass the puck back to his defenceman along the boards.

Team B's winger intercepts the puck, pokes it past Team A's defenceman, goes in for a breakaway, and scores.

------------------------------------------------------------------------------------------

CORSI implications for that Scenario:

1. Team A had a better shift, Team B had a worse shift, despite Team B scoring the goal. Team A simply had multiple shots and Team B only had one.
2. Three other players on the ice for Team A are no more or less responsible/at fault for the shot attempt (and goal) that Team B scored, despite the fact only the center and defenceman made the error
3. Four other players on the ice for Team B are no more or less responsible than the player who reclaimed possession, took the shot attempt, and scored for Team B.


------------------------------------------------------------------------------------------

CORSI conclusions:

In the eyes of CORSI, one player's success is four other players' success, and five other player's fault. One player's failure is four other players' failure, and five other player's success.

There is no measure of accountability or responsibility.

Therefore there is no way to measure or discern which individual players are drivers and which individual players are passengers.

 

[Pyromaniac3]

Scenario:

Team A is hemming Team B in their own end. Taking lots of shots attempts and CORSI'ing the hell out of Team B.

Team A's center cycles attempts to pass the puck back to his defenceman along the boards.

Team B's winger intercepts the puck, pokes it past Team A's defenceman, goes in for a breakaway, and scores.

------------------------------------------------------------------------------------------

CORSI implications for that Scenario:

1. Team A had a better shift, Team B had a worse shift, despite Team B scoring the goal. Team A simply had multiple shots and Team B only had one.
2. Three other players on the ice for Team A are no more or less responsible/at fault for the shot attempt (and goal) that Team B scored, despite the fact only the center and defenceman made the error
3. Four other players on the ice for Team B are no more or less responsible than the player who reclaimed possession, took the shot attempt, and scored for Team B.


------------------------------------------------------------------------------------------

CORSI conclusions:

In the eyes of CORSI, one player's success is four other players' success, and five other player's fault. One player's failure is four other players' failure, and five other player's success.

There is no measure of accountability or responsibility.

Therefore there is no way to measure or discern which individual players are drivers and which individual players are passengers.

This is a rare event. It has a very low chance of repeating and playing that way isn't going to win you games over a long run. Real world example: poor person wins a lottery, it doesn't mean that he is going to keep winning lotteries in the future.

Goals are incredibly rare event. And they are prone to fluctuations. Which is why it isn't a good idea to use standings or GF/GA to predict team record.


Here is why I switched over to analytics crowd: All teams in the past average around 7-9% in sh% over the course of a season regardless of team talent. This is a fact. Which means that to score more goals, they have to shoot more. To allow less goals against, they have to prevent shots against.

Having a high CF% also doesn't mean you will start winning immediately. It has been shown that luck is the biggest factor influencing your record up until game 72. After that point the cream rises to the top.

These stats are also just statistics. They have been shown to correlate highly with winning. It means there will be outliers and anomalies.

 

[Bomber0104]

This is a rare event. It has a very low chance of repeating and playing that way isn't going to win you games over a long run. Real world example: poor person wins a lottery, it doesn't mean that he is going to keep winning lotteries in the future.

Goals are incredibly rare event. And they are prone to fluctuations. Which is why it isn't a good idea to use standings or GF/GA to predict team record.


Here is why I switched over to analytics crowd: All teams in the past average around 7-9% in sh% over the course of a season regardless of team talent. This is a fact. Which means that to score more goals, they have to shoot more. To allow less goals against, they have to prevent shots against.

Having a high CF% also doesn't mean you will start winning immediately. It has been shown that luck is the biggest factor influencing your record up until game 72. After that point the cream rises to the top.

These stats are also just statistics. They have been shown to correlate highly with winning. It means there will be outliers and anomalies.

Again, I have no issue with using shot differentials to analyze a team. Maybe even certain lines. At a player level there is no stat that can accurately tell me this player is a better possession player than that player. You can most certainly tell me this team is a better possession team than that one, because the shot differential can somewhat gauge it.

 

[leafstilldeath*]

Dallas Eakins:


How many times do people have to be told this?

Your Bozak fanboy bias is showing. Nowhere in my post do I mention corsi. My argument is based on GF and GA.

May be you need to read the posts completely before moving forward with the keyboard battle

 

[leafstilldeath*]

2014/2015 (This season) based on shots against Leafs are ranked 26th in the entire league with 33.6 S/A

http://www.nhl.com/ice/teamstats.ht...&sort=avgShotsAgainstPerGame&viewName=summary

2013-2014 Leafs became the worst team in history in terms of shots against ranking 30th with approx. 36 shots against/game

http://www.nhl.com/ice/teamstats.ht...&sort=avgShotsAgainstPerGame&viewName=summary

not to say that coaching staff aren't aware of the problem this time around



As can be seen from this vid; Coaching staff is being very cautious about how not playing the right way.

That said, Carlyle's team has been consistently outshot and the first line is WORST in regards to shots forced vs shots against.

 

[leafstilldeath*]

Scenario:

Team A is hemming Team B in their own end. Taking lots of shots attempts and CORSI'ing the hell out of Team B.

Team A's center cycles attempts to pass the puck back to his defenceman along the boards.

Team B's winger intercepts the puck, pokes it past Team A's defenceman, goes in for a breakaway, and scores.

In the history of hockey what % of total team goals or for that matter winning goals are scored in the scenario presented above?

Come on people please ffs use some "logic" and "reasoning" COME ON!

 

[Daisy Jane]

As can be seen from this vid; Coaching staff is being very cautious about how not playing the right way.

That said, Carlyle's team has been consistently outshot and the first line is WORST in regards to shots forced vs shots against.

to be fair to Randy, which I think he never gets his fair shake a lot of the time. from day 1 last year, he kept saying if we don't play and win the right way it will catch up (even in between the whole 'don't critique a win' rhetoric, there was always a BUT" there).

his problem (which we've seen between two teams now) is that he doesn't know how to counteract the heavily outshot/poor possession aspect).

 

[Pyromaniac3]

Again, I have no issue with using shot differentials to analyze a team. Maybe even certain lines. At a player level there is no stat that can accurately tell me this player is a better possession player than that player. You can most certainly tell me this team is a better possession team than that one, because the shot differential can somewhat gauge it.

If a player consistently has good shot differentials, despite playing with different linemates and defense pairing, isn't he a good possession player? We eliminated almost all the factors.

 

[ULF_55]

Your Bozak fanboy bias is showing. Nowhere in my post do I mention corsi. My argument is based on GF and GA.

May be you need to read the posts completely before moving forward with the keyboard battle

You didn't?

Year Goals Assists Points Shots Corsi TOI GF20 GA20 GF% CF20 CA20 CF%
07-08 9 15 24 148 273 890:07:00 0.831 0.921 47.4 19.66 16.72 54
08-09 23 16 39 165 285 813:46:00 1.18 0.59 66.7 21.95 19.32 53.2

 

[leafstilldeath*]

to be fair to Randy, which I think he never gets his fair shake a lot of the time. from day 1 last year, he kept saying if we don't play and win the right way it will catch up (even in between the whole 'don't critique a win' rhetoric, there was always a BUT" there).

his problem (which we've seen between two teams now) is that he doesn't know how to counteract the heavily outshot/poor possession aspect).

if it is the players then some players need to go. That said, that 1st line of

JVR-Bozak-Kessel

has remained intact no matter how badly they have played. Yes they put up points but they also get scored on alot, they are hemmed in the zone, and are a rush team.

Isn't it Carlyle's fault for keeping that line together for the last 20 games of previous season when we weren't getting anything done, and even this season when evenstrength that line has been a disaster.

How is it that Carlyle expects different results while running the same experiments?

For the same period of time when Carlyle broke up that first line and was rolling

Winnik-Kadri-Kessel
JVR-Bozak-Clarkson
Komarov-Holland-Santo


things seemed to have settled down a bit although Bozak line was still one of the worst lines in terms of getting hemmed in BUT they were not as bad as before.

Instead of keeping this line up Carlyle went back to the 1st line and the results are for everyone to see.

Personally I would prefer:

Komarov-Kadri-Kessel
Lupul-Holland-Santo
JVR-Bozak-Panik
Winnik-Smith-Clarkson/Kozun/Booth

after PK (Komarav, Winnik, Holland JVR)

Carlyle could send in a loaded line of Lupul-Kadri/Bozak-Kessel for a shift as opposition's top D-men were probably on PP and try to take advantage of the situation; and then falling back to rolling four lines.

Carlyle keeps on saying last season that the way we are winning is going to catch up to us but never does anything to change it. I see a paradox here. You don't?

 

[leafstilldeath*]

You didn't?

http://hfboards.mandatory.com/showpost.php?p=93973749&postcount=165

Players just don't become worse all of a sudden defensively. Kessel shows tremendous improvement with respect to goals against stat in 2008/2009. If Kessel was that terrible his goals against stat would be higher. And by the way Kessel had a tremendous season not only by being a significantly better goal differential player in 2008/2009 but also scored 60 points in 70 games.


Is about time people stop blaming Kessel for bozak's defensive ineptitude

where in the blue hell do you see the word corsi in my argument? I give goals for and against; goals differential argument.

I posted every stat that was available people can take whatever they want from that.

and FYI, CorsiRel is an individual stat; have you looked at CorsiRel of your boy bozak?

 

[Snow Dog]

http://hfboards.mandatory.com/showpost.php?p=93973749&postcount=165



where in the blue hell do you see the word corsi in my argument? I give goals for and against; goals differential argument.

I posted every stat that was available people can take whatever they want from that.

and FYI, CorsiRel is an individual stat; have you looked at CorsiRel of your boy bozak?

Recently I've heard more than one hockey analyst praising Bozak for his play this season.I'll take a 'pro's'opinion over math anyday.By the way the same guys were praising Kadri as well.But Bozak is crap and Kadri is great in some people's opinion.They are BOTH playing well this season and we should be happy.

 

[leafstilldeath*]

Recently I've heard more than one hockey analyst praising Bozak for his play this season.I'll take a 'pro's'opinion over math anyday.By the way the same guys were praising Kadri as well.But Bozak is crap and Kadri is great in some people's opinion.They are BOTH playing well this season and we should be happy.

Bro opinions are like A-holes; everybody has got one and nothing but poop falls from it.

Without evidence and facts everything is nothing but an opinion.

 

[Snow Dog]

Bro opinions are like A-holes; everybody has got one and nothing but poop falls from it.

Without evidence and facts everything is nothing but an opinion.

For all of history athletes have been judged by experts in their sport and what that expert sees in that athlete at the time and hopefully into the future.Just because recently tracked stats have come into play these experts opinions should not be discounted.

 

[leafstilldeath*]

For all of history athletes have been judged by experts in their sport and what that expert sees in that athlete at the time and hopefully into the future.Just because recently tracked stats have come into play these experts opinions should not be discounted.

which experts are calling bozak a great player? what is their track record? How many players have they predicted that they will be greatand these players went on to have good to great careers contributing to team's success.

You name these experts, you show evidence from thepast and then you can claim that yes Mr. X or Mr.Y have the track record of good prediction so based on that whatever they are saying about Bozak may be true.

Then you "may" have a case, without that its nothing but just an opinion

 

[Menzinger]

Scenario:

Team A is hemming Team B in their own end. Taking lots of shots attempts and CORSI'ing the hell out of Team B.

Team A's center cycles attempts to pass the puck back to his defenceman along the boards.

Team B's winger intercepts the puck, pokes it past Team A's defenceman, goes in for a breakaway, and scores.

------------------------------------------------------------------------------------------

CORSI implications for that Scenario:

1. Team A had a better shift, Team B had a worse shift, despite Team B scoring the goal. Team A simply had multiple shots and Team B only had one.
2. Three other players on the ice for Team A are no more or less responsible/at fault for the shot attempt (and goal) that Team B scored, despite the fact only the center and defenceman made the error
3. Four other players on the ice for Team B are no more or less responsible than the player who reclaimed possession, took the shot attempt, and scored for Team B.


------------------------------------------------------------------------------------------

CORSI conclusions:

In the eyes of CORSI, one player's success is four other players' success, and five other player's fault. One player's failure is four other players' failure, and five other player's success.

There is no measure of accountability or responsibility.

Therefore there is no way to measure or discern which individual players are drivers and which individual players are passengers.

Your scenario seems quite specific - not to mention goals in general happen far less than shots. As you know, Corsi even on an individual player level is looking at long term trends, which would likely even out these obscure individual game occurrences.

As for the passenger vs driver issue - that's exact what WOWY is for - you can learn quite quickly which players are central to owning possession.

 

[MastuhNinks]

Also you can't just walk up to a player and tell them they need to have a better corsi. You don't want guys skating over the blueline and letting a soft floater go on net because any shot attempt will improve your corsi. The players have to be playing the game, that's the funny thing about the stat. I guarantee you there's not a single coach in the NHL that tells his team to get and prevent shot attempts at all costs, regardless of quality.

 

[Bomber0104]

In the history of hockey what % of total team goals or for that matter winning goals are scored in the scenario presented above?

Come on people please ffs use some "logic" and "reasoning" COME ON!

...

Yeah

The point of my post is not how common or not the scenario is of repeating itself...

It's how a very basic CORSI "stat" does not apportion anything in the way of statistical meaningfulness.

Kessel and JVR take a shot a piece on net. Great.

The puck comes the other way and the puck goes in on Bernier/Reimer.

Kessel and JVR had a better shift.

Yet they're still -1.

I suppose 2 shots are better than 1 shot and a goal against, right?

Are you actually going to tell me that?

 

[Bomber0104]

Your scenario seems quite specific - not to mention goals in general happen far less than shots. As you know, Corsi even on an individual player level is looking at long term trends, which would likely even out these obscure individual game occurrences.

"Likely even out" ?

Is that how you defend a mock statistic?

It is "likely" going to be true, because "I say so"?

I've already proven how CORSI does nothing to show who is responsible for a good offensive or defensive play.

Prove to me how these insignificant events demonstrate anything about a players' offensive or defensive ability.

If a player has a "positive CORSI", tell me if that means he is a better player on the offensive or defensive side of the puck.

Tell me.

 

[Delicious Dangles*]

There's an obvious lack of statistical understanding as well as hockey understanding when people try and quantify an individual player's possession ability.

Except the exact same thing is true on a team level, but people have no problem misusing that.

 

[Tyler Biggs*]

Leafs are showing depth as they have had pretty consistent scoring on all four lines. Goal tending has been solid and, with the exception of the game on Wednesday, the D has been tighter. The fact that we are playing so well without Leo or Polak shows that we are deeper this season. I was a little surprised by Bernier's start as I didn't hear them say anything about Reimer experiencing a problem. Must have missed it. Hopefully he will be good to go tonight if needed.

 

[TML1]

Hottest team in the League right now. Officially.

 

[Tyler Biggs*]

It's not that Carlyle has to "buy into this whole idea of playing better," rather the players have to buy into what Randy has been preaching for the past two years and that is to tighten up defensively. Hopefully the players realize this after playing a complete game last night!!!

 

[Budsfan]

Eric Prime ‏@PrimeSN590
Leafs' entourage (includes president Brendan Shanahan & Dubas) watching Canadian juniors.

 

[Menzinger]

If a player has a "positive CORSI", tell me if that means he is a better player on the offensive or defensive side of the puck.

Tell me.

Your arguments seem to be shifting now from the individual v team Corsi issue to challenging the very idea of it.

You know the answer to your question already.

A player with a strong Corsi raiting on 5v5 play will have higher numbers of scoring opportunities via shots (offensive impact) while clearly impacting the defensive side of the play by preventing the other side from having the same opportunity by controlling the puck. And on the individual level this is clearly measurable- last season Jay Mcclement was lauded time and time again for his supposed defensive prowess, yet his Corsi numbers clearly demonstrated he was a black whole in terms of possession.

As for the qualitative issue of the value of an individual play (as in the scenario your previous post described), well, that's why we still watch games. Not a single person has ever claimed all you need is Corsi to evaluate an individual player -it's a strawman to suggest this is the case.