Habs
- BingoLady

...yea...

i'd still rather be winning
- daeth

in the long run you want the better underlying stats instead of the wins now..sounds dumb but building wins through something sustainable is better.. we saw it last year leafs started the season 10-4 or something but were getting killed...over 80 games those things tend to even out ..we saw

that said..its really early..hard to draw any conclusions

in the long run you want the better underlying stats instead of the wins now..sounds dumb but building wins through something sustainable is better.. we saw it last year leafs started the season 10-4 or something but were getting killed...over 80 games those things tend to even out ..we saw

that said..its really early..hard to draw any conclusions
- senstroll

first off, I'm very new to these "new" stats, like corsi and fenwick(?),

But from what I've gleaned here and other places is that Corsi doesn't track anything other than possession, they don't include turnovers, giveaways, passes, or defensive end zone play.

And like any other single stat, cannot be looked at alone.

And as far as NHL hockey is such a fluid game, VS MLB and the NFL where there are set plays and set defences, and therefore stats like corsi and fenwick(?) do not relate as well to the NHL in general.

So I don't know if there should be as much weight placed of them as in the other major sports.

first off, I'm very new to these "new" stats, like corsi and fenwick(?),

But from what I've gleaned here and other places is that Corsi doesn't track anything other than possession, they don't include turnovers, giveaways, passes, or defensive end zone play.

And like any other single stat, cannot be looked at alone.

And as far as NHL hockey is such a fluid game, VS MLB and the NFL where there are set plays and set defences, and therefore stats like corsi and fenwick(?) do not relate as well to the NHL in general.

So I don't know if there should be as much weight placed of them as in the other major sports.
- Doppleganger

All 100% true. Which is why most of your own fanbase looked like fools last year referencing it endlessly like it was more important than the scoreboard.

Additionally, giveaways, turnovers, missed shots, hits, and several other ones, the NHL's tracking of them is near useless, given how much they differ from rink to rink.

All 100% true. Which is why most of your own fanbase looked like fools last year referencing it endlessly like it was more important than the scoreboard.

Additionally, giveaways, turnovers, missed shots, hits, and several other ones, the NHL's tracking of them is near useless, given how much they differ from rink to rink.
- prock

almost as foolish as leaf fans like you, who claimed they weren't important at all, at that the leafs could totally sustain their "winning ways" despite getting DOMINATED and posting some of the worst possession stats ever.

almost as foolish as leaf fans like you, who claimed they weren't important at all, at that the leafs could totally sustain their "winning ways" despite getting DOMINATED and posting some of the worst possession stats ever.
- sensarmy_11

I've never said they weren't relevant at all. I've said a million times, they're just another stat. One small piece of the puzzle. No more or less relevant than the vast number of other stats out there.

I do see statements like the one you just mentioned in the Sens threads these days though.

I've never said they weren't relevant at all. I've said a million times, they're just another stat. One small piece of the puzzle. No more or less relevant than the vast number of other stats out there.

I do see statements like the one you just mentioned in the Sens threads these days though.
- prock

Not in the Sens threads, in the Sens forum on this site.

first off, I'm very new to these "new" stats, like corsi and fenwick(?),

But from what I've gleaned here and other places is that Corsi doesn't track anything other than possession, they don't include turnovers, giveaways, passes, or defensive end zone play.

And like any other single stat, cannot be looked at alone.

And as far as NHL hockey is such a fluid game, VS MLB and the NFL where there are set plays and set defences, and therefore stats like corsi and fenwick(?) do not relate as well to the NHL in general.

So I don't know if there should be as much weight placed of them as in the other major sports.
- Doppleganger

data tracking and analytics is improving for Hockey and I hope lots of NHL teams think like you do since the Leafs are fully onboard with them now.

cosri is just shot attempts.. having more than your opponent we can see through these stats correlates to winning more. Its not hard stuff

the stats are evolving and getting better....just hope your team is onboard

data tracking and analytics is improving for Hockey and I hope lots of NHL teams think like you do since the Leafs are fully onboard with them now.

cosri is just shot attempts.. having more than your opponent we can see through these stats correlates to winning more. Its not hard stuff

the stats are evolving and getting better....just hope your team is onboard
- senstroll

My problem with this statement is that there are a LOT of stats that “correlate to winning more”. In fact, probably 90% of stats correlate to winning or losing more in some fashion. Why do they have to be so “advanced”? I’m sure simple shots for and shots against, ratios of the two, etc, are probably pretty close to the same as CORSI, with all the additional crap they pile in. And that’s at a team level.

There is a bit of cause and effect there to me.

Additionally, like many other stats, they’re context driven to some extent. Different styles of play, team composition, etc, will result in different results of different statistics. Additionally, different roles on a team can alter CORSI measures rather drastically.

I've never said they weren't relevant at all. I've said a million times, they're just another stat. One small piece of the puzzle. No more or less relevant than the vast number of other stats out there.

I do see statements like the one you just mentioned in the Sens threads these days though.
- prock

They're more useful than giveaways and takeaways for two reasons:

1) like you said, those are doled out completely differently depending on the rink
2) A ton of giveaways could either indicate a player is careless with the puck, or it could indicate simply that they possess the puck a whole heck of a lot. Guys like Thornton and Kessel have high giveaway totals because there's almost an 100% chance they're going to touch the puck every possession. They're cornerstones. Who are the full-time guys with the fewest giveaways? Guys like Matt Martin and Lance Bouma.

My problem with this statement is that there are a LOT of stats that “correlate to winning more”. In fact, probably 90% of stats correlate to winning or losing more in some fashion. Why do they have to be so “advanced”? I’m sure simple shots for and shots against, ratios of the two, etc, are probably pretty close to the same as CORSI, with all the additional crap they pile in. And that’s at a team level.

There is a bit of cause and effect there to me.

Additionally, like many other stats, they’re context driven to some extent. Different styles of play, team composition, etc, will result in different results of different statistics. Additionally, different roles on a team can alter CORSI measures rather drastically.
- prock

There are three possibilities:
1) Shots for, shots against are less accurate than shot attempts for, shot attempts against.

2)Shots for, shots against are just as accurate as shot attempts for, shot attempts against

3) Shots for, shots against are more accurate than shot attempts for, shot attempts against

If 1) is true, obviously corsi is a better measure.
If 2) is true, corsi is still a better measure, because it generates considerably more data. More data to draw from means more accuracy.
So unless 3) is true, it seems to me that corsi is a better indicator of success than merely looking at shots for, shots against.

There's probably an answer to this question by the way, I just haven't peeled through Yost's back catalogues.

By the way, I agree that whatever the answer, neither corsi nor any other stat is a be-all and end-all and people who make that claim were probably picked last in gym.

They're more useful than giveaways and takeaways for two reasons:

1) like you said, those are doled out completely differently depending on the rink
2) A ton of giveaways could either indicate a player is careless with the puck, or it could indicate simply that they possess the puck a whole heck of a lot. Guys like Thornton and Kessel have high giveaway totals because there's almost an 100% chance they're going to touch the puck every possession. They're cornerstones. Who are the full-time guys with the fewest giveaways? Guys like Matt Martin and Lance Bouma.
- Morris

Yes, I should have said “no more or less relevant than many of the vast number of other stats out there. Giveaways and takeaways are a bit of a joke.

It’s funny how you can look at some teams, and if you believed those stats, they “giveaway” the puck and a fraction of the rate in their own building as they do in everyone else’s. Same as takeaway.

It would be a little more useful to look at those stats in away games only, as they would tend to average out over a larger sample size, but it’s still a joke.

And yes, they’re very role/context driven stats too.

If 2) is true, corsi is still a better measure, because it generates considerably more data. More data to draw from means more accuracy.

- Morris

Only if all data is considered equal. Missed shots, blocked shots, etc, run into the same difficulties as giveaways, takeaways, hits. Mind you, to a lesser extent, but there is a little more subjectivity than in straight shots for/against(which, even that, there is some subjectivity in).

Only if all data is considered equal. Missed shots, blocked shots, etc, run into the same difficulties as giveaways, takeaways, hits. Mind you, to a lesser extent, but there is a little more subjectivity than in straight shots for/against(which, even that, there is some subjectivity in).
- prock

that's a good point. If a dman trying to keep it in turns and flings a puck in the general direction of the net, it's a lot different than a point shot that hits the post.

I'd like to see an impartial group do a study on whether the increased data (with worse data) of Corsi is a net win over the less data (and better data) of SOG

that's a good point. If a dman trying to keep it in turns and flings a puck in the general direction of the net, it's a lot different than a point shot that hits the post.

I'd like to see an impartial group do a study on whether the increased data (with worse data) of Corsi is a net win over the less data (and better data) of SOG
- Morris

I’ve seen people try to do analysis of how strongly it correlates to winning.

I’d love to see the same analysis done beside things like shot differentials, goal differentials, etc. i.e. studies on correlation of ALL stats to winning.

I suspect that, as I’ve said, you can take many, if not most stats, and find either a negative or positive correlation to winning with them. Better teams will have better stats. Seems pretty obvious to me. CORSI, among many other stats, included. Why does having a good CORSI mean so much more than the rest of the stats that behave the same way? In my opinion, it doesn’t.

I think people should do more analysis on what it tells you about a team, and not necessarily just “itz gunna make us winz more”. And to me, more importantly, on an individual player, what does it tell you about that player, and not just assume “itz cuz he’s the bestest”.

I’ve seen people try to do analysis of how strongly it correlates to winning.

I’d love to see the same analysis done beside things like shot differentials, goal differentials, etc. i.e. studies on correlation of ALL stats to winning.

I suspect that, as I’ve said, you can take many, if not most stats, and find either a negative or positive correlation to winning with them. Better teams will have better stats. Seems pretty obvious to me. CORSI, among many other stats, included. Why does having a good CORSI mean so much more than the rest of the stats that behave the same way? In my opinion, it doesn’t.

I think people should do more analysis on what it tells you about a team, and not necessarily just “itz gunna make us winz more”. And to me, more importantly, on an individual player, what does it tell you about that player, and not just assume “itz cuz he’s the bestest”.
- prock

Yeah, looking at what teams/players have good corsi% doesn't really tell you anything about what to DO to get good numbers.

I like looking at zone entries and exits. I know specifically in watching the oilers that an inability to skate the puck out of our zone or into their zone is a big problem. Other than that, things you count when watching the game like how many corner battles a team wins, how successful a dman in pinching, etc. probably end up correlating pretty well to corsi anyway, but are much more useful than "they were at 53% when the score was close!"

first off, I'm very new to these "new" stats, like corsi and fenwick(?),

But from what I've gleaned here and other places is that Corsi doesn't track anything other than possession, they don't include turnovers, giveaways, passes, or defensive end zone play.

And like any other single stat, cannot be looked at alone.

And as far as NHL hockey is such a fluid game, VS MLB and the NFL where there are set plays and set defences, and therefore stats like corsi and fenwick(?) do not relate as well to the NHL in general.

So I don't know if there should be as much weight placed of them as in the other major sports.
- Doppleganger

A lot of the sites that track corsi/fenwick also do track other stats, like zone starts, quality of competition(based on the competition's corsi), relative corsi(isolating one player to see his effectiveness, even if the team is very good or very bad at possesion), and how successful players are at zone exits and zone entries.

While it doesn't look at zone time per say, it uses shot attempts as a proxy for possession, and if you're getting shots towards, and not allowing the other team to get many towards yours, odds are you're going to be spending significantly more time in the other team's zone.

There are teams that can win in spite of poor corsi. Teams with fantastic goaltending, and high-end forwards, for example. Toronto, for example had a very good record through 120+ games, in spite of bad possession stats. The problem was, when the goaltending stumbled, everything fell apart.

So while Corsi doesn't really tell you what team is going to win more, it does reveal which team is playing a better system overall(IMO)... and what teams are more resistant to unforeseen circumstances. If Toronto had been a good possession team, odds are Bernier going down wouldn't have hurt them as much as it did.

that's a good point. If a dman trying to keep it in turns and flings a puck in the general direction of the net, it's a lot different than a point shot that hits the post.

I'd like to see an impartial group do a study on whether the increased data (with worse data) of Corsi is a net win over the less data (and better data) of SOG
- Morris

But it doesn't matter that one shot was a high scoring opportunity and the other wasn't ever going to be a goal(outside of insane bounces)... Corsi doesn't care about that... it's just looking at possession. Knowing that one team is shooting tells you that they have possession of the puck. Doesn't matter where they're shooting from or what they're shooting at.

I enjoy looking at advanced stats, but one of the big virtues of them is that data can be compiled on every game without having watched it.

The difference between amateur blogger and guy who gets hired by an NHL team is when you see a trend, and then go back and pore over game film to see how that trend came to be.

I enjoy looking at advanced stats, but one of the big virtues of them is that data can be compiled on every game without having watched it.

The difference between amateur blogger and guy who gets hired by an NHL team is when you see a trend, and then go back and pore over game film to see how that trend came to be.
- Morris

and you know these guys are loving it now that they're getting hired up by teams... not just because of the paycheques, but because they get to see a lot more information now, and have a say in new ways of tracking teams, and gathering info.

But it doesn't matter that one shot was a high scoring opportunity and the other wasn't ever going to be a goal(outside of insane bounces)... Corsi doesn't care about that... it's just looking at possession. Knowing that one team is shooting tells you that they have possession of the puck. Doesn't matter where they're shooting from or what they're shooting at.
- Feeling Glucky?

I know that. I'm saying there are some instances where a player makes what one person could construe as a shot attempt and another may not.

Example 1: What one person would consider a missed shot and another would consider dumping it into the corner.

Example 2: What one person would consider a pass that hit a defender's skate another would consider a blocked shot.

The people who count missed and blocked shots at the rink are the same people who botch hits, giveaways and takeaways and vary wildly from each other. It's not like all the advanced statisticians are counting themselves. They're collecting data from these people.

Yeah, looking at what teams/players have good corsi% doesn't really tell you anything about what to DO to get good numbers.

I like looking at zone entries and exits. I know specifically in watching the oilers that an inability to skate the puck out of our zone or into their zone is a big problem. Other than that, things you count when watching the game like how many corner battles a team wins, how successful a dman in pinching, etc. probably end up correlating pretty well to corsi anyway, but are much more useful than "they were at 53% when the score was close!"
- Morris

And, in hockey, you need a mix of all kinds of different skills to become a winning team. Which is why, perhaps looking for a player with excellent stats in that area, as a priority over CORSI, goals, assists, hits, etc, etc, may be smart. i.e. look to fill a hole in the team. In my opinion, that’s what GMs should be doing. Whether it be by position, or stat, or whatever. Not just one stat. Where are the holes in the team, and what player can contribute to fixing that.

I’d love to see someone assemble a team of CORSI kings from around the league. I wonder if they’d even be a winning team. I personally think it would shut the CORSI obsessed fans up.

and you know these guys are loving it now that they're getting hired up by teams... not just because of the paycheques, but because they get to see a lot more information now, and have a say in new ways of tracking teams, and gathering info.
- Feeling Glucky?

Well they probably have a chance to eliminate some of the recording errors I mention in the post above now that they're getting paid to do it.

And, in hockey, you need a mix of all kinds of different skills to become a winning team. Which is why, perhaps looking for a player with excellent stats in that area, as a priority over CORSI, goals, assists, hits, etc, etc, may be smart. i.e. look to fill a hole in the team. In my opinion, that’s what GMs should be doing. Whether it be by position, or stat, or whatever. Not just one stat. Where are the holes in the team, and what player can contribute to fixing that.

I’d love to see someone assemble a team of CORSI kings from around the league. I wonder if they’d even be a winning team. I personally think it would shut the CORSI obsessed fans up.
- prock

The LA Kings.

And, in hockey, you need a mix of all kinds of different skills to become a winning team. Which is why, perhaps looking for a player with excellent stats in that area, as a priority over CORSI, goals, assists, hits, etc, etc, may be smart. i.e. look to fill a hole in the team. In my opinion, that’s what GMs should be doing. Whether it be by position, or stat, or whatever. Not just one stat. Where are the holes in the team, and what player can contribute to fixing that.

I’d love to see someone assemble a team of CORSI kings from around the league. I wonder if they’d even be a winning team. I personally think it would shut the CORSI obsessed fans up.
- prock

Do you mean people who by traditional metrics weren't that good but by Corsi are great?

Because guys like Toews, Parise, Kopitar have good corsi numbers