Post-Game Talk: #10 - 10/29/19 | lightning @ RANGERS

[ohbaby]

There is only one Corsi measurement.
There are no "other Corsi figures"
You don't know what you're talking about.

Ugh! You know what plural means. Corsi figures as in more than one, you know like there are more than one player on a team... Geez

 

[ohbaby]

I hope we do it again on Saturday.

Consistency will be the problem this year. Usually a problem with a young team.

 

[Irishguy42]

Consistency will be the problem this year. Usually a problem with a young team.

I'm not expecting the same level, but even a vague resemblance would be welcome.

 

[ohbaby]

I'm not expecting the same level, but even a vague resemblance would be welcome.

Well at least it's good to know we are capable of some pretty awesome efforts when we put our minds to it.

 

[Machinehead]

Not sure why I had Sacco on the brain there.

He also coached the Avs, I believe.

 

[Shesterkybomb]

Analytics are what has Toronto unable to get over the hump because analytics dont measure game to game intimidation, opponent size and physicality, or basically anything that has to do with what a player does vs a bigger more physically intimidating player not to mention the style change that occurs in playoffs. It just points you to skilled players who get a lot of points during the regular season or a guy who plays safe hockey. Claude Lemieux would be an analytical nightmare during the regular season. There isnt a team that wouldnt take his playoff abilities. Analytics are a tool but they dont tell much of a story game to game vs different opponents.

 

[ReggieDunlop68]

I agree, but Twitter and sometimes HF keep telling me that Smith is bad because he's a defenseman playing as a forward.

Or something like that...

I thought people were saying something like "man he was terrible as a defensemen, so they moved him to forward as a last resort, and he's doing sort of ok on a sort of ok team".

 

[ReggieDunlop68]

Analytics are what has Toronto unable to get over the hump because analytics dont measure game to game intimidation, opponent size and physicality, or basically anything that has to do with what a player does vs a bigger more physically intimidating player not to mention the style change that occurs in playoffs. It just points you to skilled players who get a lot of points during the regular season or a guy who plays safe hockey. Claude Lemieux would be an analytical nightmare during the regular season. There isnt a team that wouldnt take his playoff abilities. Analytics are a tool but they dont tell much of a story game to game vs different opponents.

The real truth about the analytics is it's just more bullshit for fans to waste too much time on.

 

[Kaapo Hollweg]

You can ask that for the entire team. No one looked good that boston game.

Sure, but everybody talked about possibly benching him instead of Staal after Lindgren came up. So I was just wondering how he responded to that.

 

[romba]

is it possible to discuss a player not playing well without HF crying "WHIPPING BOY!!!" in unison?

Is world peace achievable?

 

[GoAwayGiannone]

The real truth about the analytics is it's just more bull**** for fans to waste too much time on.

Surely, that's why billion dollar organizations are "wasting their time" by employing entire firms dedicated to analytics.

 

[ReggieDunlop68]

Surely, that's why billion dollar organizations are "wasting their time" by employing entire firms dedicated to analytics.

They use different ones than the fans.

 

[Fitzy]

Whipping boy to me means disproportionate or undeserved critisisms i.e. Roszival.

Staal gets critiqued because his performance to contract ratio is bottom five in the league

 

[romba]

xGF is a little bit better. That said, you can still inflate xGF just by doing a lot of useless corsi.

Although I will say that 10% is probably about accurate tbh. That's a .900 save percentage. You're not holding a job in the NHL with a .900 save percentage.

So yeah, if an NHL goalie is stopping it 9/10 times, it's probably a scoring chance.

I feel like many fans think of scoring chance as like a 30% or higher chance of going in. But that's pretty much a half open net these days with the modern goalie's ability to recover and get across in time (unless it's roofed, more often than not). I'd say a 10% chance of going in is a bit low but not that crazy. A few percentage points make the difference for me, but this is very subjective.

 

[SA16]

I am not so sure Staal gets criticized because of his contract. If he was on a league min deal he should still be criticized the same. It's not like he is playing well but is overpaid like say...Derek Stepan maybe.

 

[SA16]

The league average shooting percentage at 5v5 this year is 8.3%. The league average shooting percentage at 5v5 for forwards is 10.1% If you have a stat declaring something 10% as a scoring chance your stat is wrong. You're basically saying the average shot is a scoring chance. Absolute nonsense. Not saying the stat doesn't correlate to future goals better than something else but the name of your stat is misleading.

 

[Machinehead]

The league average shooting percentage at 5v5 this year is 8.3%. The league average shooting percentage at 5v5 for forwards is 10.1% If you have a stat declaring something 10% as a scoring chance your stat is wrong. You're basically saying the average shot is a scoring chance. Absolute nonsense. Not saying the stat doesn't correlate to future goals better than something else but the name of your stat is misleading.

That number factors in a lot of really bad shooters though. Good forwards can and do shoot higher than that and they're shooting higher because they're creating better chances.

You do bring up a valid point. I was just making the point that I don't think 10% for the low bar of what's considered a scoring chance is outrageous, but that is the low bar.

And a lot of analytics people have taken issue with scoring chances because they have a binning issue. For those who don't know (I know you do), a "binning" issue in statistics refers to putting very different results into the same broadly defined bin. That is to say, a 10% scoring chance is in the same bin as a 30-40% scoring chance, so you're absolutely right to say it's misleading.

I just think there's a tendency to overrate or inflate what constitutes a scoring chance. Not every scoring chance is a particularly good one.

 

[Machinehead]

Whipping boy to me means disproportionate or undeserved critisisms i.e. Roszival.

Staal gets critiqued because his performance to contract ratio is bottom five in the league

Exactly this.

Being really bad and people thinking you're really bad is not a whipping boy.

 

[SA16]

That number factors in a lot of really bad shooters though. Good forwards can and do shoot higher than that and they're shooting higher because they're creating better chances.

You do bring up a valid point. I was just making the point that I don't think 10% for the low bar of what's considered a scoring chance is outrageous, but that is the low bar.

And a lot of analytics people have taken issue with scoring chances because they have a binning issue. For those who don't know (I know you do), a "binning" issue in statistics refers to putting very different results into the same broadly defined bin. That is to say, a 10% scoring chance is in the same bin as a 30-40% scoring chance, so you're absolutely right to say it's misleading.

I just think there's a tendency to overrate or inflate what constitutes a scoring chance. Not every scoring chance is a particularly good one.

I guess I would say if it's not a "good" scoring chance then I personally would not consider it a scoring chance. To me HDCF is more of actual scoring chances and SCF is basically just shots excluding dmen shooting from far out.

 

[Machinehead]

I guess I would say if it's not a "good" scoring chance then I personally would not consider it a scoring chance. To me HDCF is more of actual scoring chances and SCF is basically just shots excluding dmen shooting from far out.

I think that's a fair take. One can't miss chance moves the needle more than 7 or 8 decent ones. That's why it's so hard to quantify hockey.

I think the next wave of analytics is going to be puck-based and player-based stuff.

Soccer has a similar problem where we all kind of know what a scoring chance is but it's very difficult to put in concrete terms and impossible to rank one better than the next. So what they're doing instead, is focusing on the ball and the player. It's hard to quantify chances, but we can quantify things that would reasonably lead to or prevents chances -- passes that beat defenders, dribbles that beat defenders, areal battles, tackles and interceptions as well as tackles and interceptions in prime areas, defenders drawn out of position.

It would be cool to see the same in hockey. It's hard to say how good a "chance" is but if we can say how many board battles a player won or how many times a game his passes beat a defenseman, we can reasonably conclude that he's driving chances.

 

[Shesterkybomb]

Surely, that's why billion dollar organizations are "wasting their time" by employing entire firms dedicated to analytics.

The Blues have the least amount of money spent on analytics.

 

[romba]

I think that's a fair take. One can't miss chance moves the needle more than 7 or 8 decent ones. That's why it's so hard to quantify hockey.

I think the next wave of analytics is going to be puck-based and player-based stuff.

Soccer has a similar problem where we all kind of know what a scoring chance is but it's very difficult to put in concrete terms and impossible to rank one better than the next. So what they're doing instead, is focusing on the ball and the player. It's hard to quantify chances, but we can quantify things that would reasonably lead to or prevents chances -- passes that beat defenders, dribbles that beat defenders, areal battles, tackles and interceptions as well as tackles and interceptions in prime areas, defenders drawn out of position.

It would be cool to see the same in hockey. It's hard to say how good a "chance" is but if we can say how many board battles a player won or how many times a game his passes beat a defenseman, we can reasonably conclude that he's driving chances.

What you say is definitely true and would be wonderful to track but once the puck and players are tracked via chip we should be able to quantify scoring chances themselves better. Exact release point, nearest defender distance from puck, traffic between the goalie and release point, puck trajectory, and speed of the shot should be quantifiable. Or at least I hope it will be!

 

[duhmetreE]

seeing Strome's career as anything other than a consistent tweener disappointment is rose colored glasses lmao

There's nothing consistent about Strome until he came here. He's a middle 6 player possibly figuring it out.

Maybe I was never as high on Stepan as others but they're in the same tier for me.

 

[will1066]

I thought people were saying something like "man he was terrible as a defensemen, so they moved him to forward as a last resort, and he's doing sort of ok on a sort of ok team".

You don't need to carry as much weight around playing forward than you do defense

 

[will1066]

There's nothing consistent about Strome until he came here. He's a middle 6 player possibly figuring it out.

Maybe I was never as high on Stepan as others but they're in the same tier for me.

He consistently gets a point a game