HFBoards

HFBoards (http://hfboards.hockeysfuture.com/index.php)
-   Pittsburgh Penguins (http://hfboards.hockeysfuture.com/forumdisplay.php?f=23)
-   -   Corsi Numbers for Pens (http://hfboards.hockeysfuture.com/showthread.php?t=1342265)

vodeni 02-05-2013 12:50 PM

Corsi Numbers for Pens
 
Despres has the highest Corsi score, followed by Nisky, Malking, Sid and Letang
it is alarming that our third line has a very low number, basically saying that they do not posses the puck and have a very little offensive production

http://www.behindthenet.ca/nhl_stati...4+45+46+63+67#

What do you make of it?

IcedCapp 02-05-2013 12:51 PM

small sample, for now.

Frederick Stanley 02-05-2013 12:55 PM

I never head about the Corsi Numbers until yesterday, when someone from another forum posted it.

Am I the only one? How long does this statistic exist?

td_ice 02-05-2013 12:59 PM

Quote:

Originally Posted by Frederick Stanley (Post 59093387)
I never head about the Corsi Numbers until yesterday, when someone from another forum posted it.

Am I the only one? How long does this statistic exist?

I have heard of it for awhile now, but I know little about it. Haven't really studied it or researched it.
I have no idea if it holds any validity or not. But I guess that it's validity would be very subjective. Meaningful for some, meaningless for others.

JordanStaal#1Fan 02-05-2013 01:01 PM

Quote:

Originally Posted by Frederick Stanley (Post 59093387)
I never head about the Corsi Numbers until yesterday, when someone from another forum posted it.

Am I the only one? How long does this statistic exist?

I have no clue what "corsi" is whatsoever. Hockey isn't baseball...

KIRK 02-05-2013 01:08 PM

http://thehockeywriters.com/corsi-nu...hl-statistics/

Quote:

Based on his Corsi Numbers, right-winger Teddy Purcell was a good signing by the Tampa Bay Lightning.

Haven’t heard of “Corsi Numbers”? You’re not alone.

The NHL isn’t known as a statistically progressive organization. It often lags behind other major sports leagues when it comes to analytical number-crunching.

But new categories of statistics are gaining popularity and these state-of-the-art numbers are shining light into the dim corners of one of the world’s fastest and most complex games.

Corsi Numbers have gained popularity as a statistical measure of a player's contribution - the results are often surprising.

Darth Vitale 02-05-2013 01:13 PM

Because it does not distinguish between good scoring chances and bad, or shots that get through to the net in some cases, it's a pretty meaningless thing AFAICT but I don't know that much about it.

JordanStaal#1Fan 02-05-2013 01:17 PM

So MA Bergeron is a good player acccording to corsi? The guy only plays on the PP of course he is gonna be on the ice for more shots for than shots against...

I really don't get the relevance of it all.

Shady Machine 02-05-2013 01:20 PM

Quote:

Originally Posted by JordanStaal#1Fan (Post 59094441)
So MA Bergeron is a good player acccording to corsi? The guy only plays on the PP of course he is gonna be on the ice for more shots for than shots against...

I really don't get the relevance of it all.

It's measured only for 5 on 5 play from what I can tell, so MAB's PP time does not impact his corsi number. However, it doesn't and can't take into account the amount of offensive zone starts he would get over other dmen on their team. So yeah, as the article says, it's not a silver bullet and needs to be taken in context. I'd say he is an outlier.

vodeni 02-05-2013 01:47 PM

Quote:

Originally Posted by Shady Machine (Post 59094639)
It's measured only for 5 on 5 play from what I can tell, so MAB's PP time does not impact his corsi number. However, it doesn't and can't take into account the amount of offensive zone starts he would get over other dmen on their team. So yeah, as the article says, it's not a silver bullet and needs to be taken in context. I'd say he is an outlier.

it is adjusted by the number of offensive or defensive starts, they count that in, also there is a relative number that takes into consideration a quality of the opponents that one player is up against while on ice, in other words, Brandon Sutter and mat Cooke although with the worst offensive Corsi number, get bounce back because the opposing five they face have very high Corsi number, quite complicated but IMO it can show some real trends in development of a player...
and yes its only 5 on 5

td_ice 02-05-2013 01:53 PM

Is there any evidence that any NHL coaching staff/Gm's use this data?? Like in baseball, Sabremetrics is used by many teams.

Sidney the Kidney 02-05-2013 01:58 PM

I'm not sure how accurate these ratings are when our shutdown pairing of Orpik/Martin (who have played well for the most part) are near the bottom of the roster. :help:

wgknestrick 02-05-2013 02:06 PM

Quote:

Originally Posted by vodeni (Post 59093147)
Despres has the highest Corsi score, followed by Nisky, Malking, Sid and Letang
it is alarming that our third line has a very low number, basically saying that they do not posses the puck and have a very little offensive production

http://www.behindthenet.ca/nhl_stati...4+45+46+63+67#

What do you make of it?

Welcome to the complex world of starting to understand much more about hockey. Corsi is a great stat, but with every stat, there is "context" that is critical to understand. Corsi is heavy context related and this refers to how a player is used by a coach.

Pens 3rd line is a great example. Not only do they play against top competition (ie positive +- players, or 1st/2nd liners), they also start a high percentage of their starts in the defensive zone. This zone start metric is probably the biggest reason they are last in corsi on the pens. Byslma prefers to have them on the ice for defensive zone faceoffs (along with the 4th line for some reason).

http://www.behindthenet.ca/nhl_stati...61+62+64+65+66

Bylsma seems to like to heavily load his lines to optimize his chances of scoring. Top 2 lines (players who "can" score and have higher shooting %s) tend to start more shifts in offensive zone, while the 3rd/4th lines are tasked with tough assignments of getting the puck out of their defensive zone and pushing play into the other end.

The corsi numbers ARE NOT zone start adjusted which is very misleading. They are adjusted relative to team only.
Zone starts is the biggest factor to adjust for.


Depres is doing fairly well despite having a relatively easy assignment by Bylsma. Dan is sheltering Depres from other team's top lines, and a lot of defensive zone starts. He is still doing well though, indicating he should be able to handle more of a defensive load.

wgknestrick 02-05-2013 02:10 PM

Quote:

Originally Posted by Sidney the Kidney (Post 59096613)
I'm not sure how accurate these ratings are when our shutdown pairing of Orpik/Martin (who have played well for the most part) are near the bottom of the roster. :help:

They are not ratings. Corsi is only 1 stat to use. Orpik and Martin have been doing fantastically by all statistical measures.

GAA / 60min of ice time is my preferred defensive stat. It accounts for both corsi, and on ice shooting % while a defender is out on the ice. Just remember the context also applies to this as well.

http://www.behindthenet.ca/nhl_stati...4+25+26+27+28#

vodeni 02-05-2013 02:15 PM

Quote:

Originally Posted by wgknestrick (Post 59097217)
They are not ratings. Corsi is only 1 stat to use. Orpik and Martin have been doing fantastically by all statistical measures.

GAA / 60min of ice time is my preferred defensive stat. It accounts for both corsi, and on ice shooting % while a defender is out on the ice. Just remember the context also applies to this as well.

http://www.behindthenet.ca/nhl_stati...4+25+26+27+28#

I guess that's why Lovejoy is in the line up again tonight, so he can improve his over 4 GAA/60 number...:sarcasm:

regard 02-05-2013 02:17 PM

I imagine Tampa is up there because of the 1-3-1 they were running. And when they were running that in full Bergeron seemed to be thier favorite option to be the back one. All he had to to was wait for teams to chip it in then go retrieve it.

Thats my only contribution to this thread. :P

Shady Machine 02-05-2013 02:44 PM

Quote:

Originally Posted by vodeni (Post 59096009)
it is adjusted by the number of offensive or defensive starts, they count that in, also there is a relative number that takes into consideration a quality of the opponents that one player is up against while on ice, in other words, Brandon Sutter and mat Cooke although with the worst offensive Corsi number, get bounce back because the opposing five they face have very high Corsi number, quite complicated but IMO it can show some real trends in development of a player...
and yes its only 5 on 5

I stand corrected. Thanks for clarifying. Someday I will dig into all the details of it. If that's the case, I wonder how MAB had such a high number. He played well in those playoffs, but he isn't a defensive stalwart by any means.

Edit: Seems that you were wrong haha. It seemed odd that zone starts would be included.

Pancakes 02-05-2013 03:59 PM

Brandon Sutter starts 60% of his shifts in the defensive zone and he has not been on the ice for an even strength goal all season. He's the only Pen player with that distinction.

Jacob 02-05-2013 04:45 PM

Quote:

Originally Posted by vodeni (Post 59093147)
Despres has the highest Corsi score, followed by Nisky, Malking, Sid and Letang
it is alarming that our third line has a very low number, basically saying that they do not posses the puck and have a very little offensive production

http://www.behindthenet.ca/nhl_stati...4+45+46+63+67#

What do you make of it?

Do we need Corsi to tell us that? They have a combined 3 goals.

billybudd 02-05-2013 04:59 PM

http://www.behindthenet.ca/nhl_stati...4+45+46+63+67#


This is probably the better stat to look at. It's corsi adjusted for difficulty of opposition.

And yeah, there's a huge difference in this for Martin compared to what he was doing in last year's playoffs, as if it wasn't already blatantly obvious how much better he's been.

Sideline 02-05-2013 05:10 PM

Quote:

Originally Posted by Jacob (Post 59104763)
Do we need Corsi to tell us that? They have a combined 3 goals.

My day job involves a fair bit of statistical arbitrage in financial markets. I dream of finding situations where people believe in models like this.

Corsi is supposed to be a good way to figure out which 3rd liners are the most efficient. Someone that believes in Corsi thinks Tyler Kennedy is a better 2-way player than Pascal Dupuis. Kennedy is consistently a better Corsi player than Dupuis. Anyone that knows anything about hockey knows Dupuis is a clearly superior 2-way player.

Malkin4Top6Wingerz 02-05-2013 09:41 PM

Quote:

Originally Posted by Sideline (Post 59105989)
My day job involves a fair bit of statistical arbitrage in financial markets. I dream of finding situations where people believe in models like this.

Corsi is supposed to be a good way to figure out which 3rd liners are the most efficient. Someone that believes in Corsi thinks Tyler Kennedy is a better 2-way player than Pascal Dupuis. Kennedy is consistently a better Corsi player than Dupuis. Anyone that knows anything about hockey knows Dupuis is a clearly superior 2-way player.

Corsi is an accurate measure of two way play - said by nobody who has anything valuable to say on the subject of Corsi.

I have to run so I'll make this very brief. Corsi is a loose but ultimately statistically significant statistic which measures a players ability to influence goal differential at even strength. That is it. It has some outliers and needs to be context adjusted, but anybody arguing that it is meaningless should be ignored.

JordanStaal#1Fan 02-05-2013 09:44 PM

Quote:

Originally Posted by Malkin4Top6Wingerz (Post 59138059)
Corsi is an accurate measure of two way play - said by nobody who has anything valuable to say on the subject of Corsi.

I have to run so I'll make this very brief. Corsi is a loose but ultimately statistically significant statistic which measures a players ability to influence goal differential at even strength. That is it. It has some outliers and needs to be context adjusted, but anybody arguing that it is meaningless should be ignored.

Then, how do you explain M-A Bergerin leading the league in corsi during the playoffs? The guy is one of the WORST defensive player I've ever seen.

peskypenguin4 02-05-2013 11:26 PM

Quote:

Originally Posted by JordanStaal#1Fan (Post 59138329)
Then, how do you explain M-A Bergerin leading the league in corsi during the playoffs? The guy is one of the WORST defensive player I've ever seen.

I think you need to re-read his post.

wgknestrick 02-06-2013 08:21 AM

Quote:

Originally Posted by JordanStaal#1Fan (Post 59138329)
Then, how do you explain M-A Bergerin leading the league in corsi during the playoffs? The guy is one of the WORST defensive player I've ever seen.

Corsi is NOT a measure of defensive prowess. It is a measure of shots for/against. This is why Gogo was and still is a Corsi monster on D. It is better to play "defense" in the other team's zone. It also doesn't take into account any shot quality for/against. This is why Crosby's corsi is better than TK's corsi. This is also why the Pens had to part ways with Michalek for basically nothing. He was/is a poor defensively player.

The corsi rel qualcomp list is only the "adjustment" that is added/subtracted to the raw corsi numbers. It is not very significant. Zone starts adjustments (which unfortunately aren't accounted for anywhere on behindthenet) are absolutely critical. Nothing impacts your corsi numbers more than where you start your shifts.


All times are GMT -5. The time now is 05:20 AM.

vBulletin Copyright ©2000 - 2015, Jelsoft Enterprises Ltd.
HFBoards.com, A property of CraveOnline, a division of AtomicOnline LLC ©2009 CraveOnline Media, LLC. All Rights Reserved.