Player Discussion Josh Archibald

[McDNicks17]

You look at raw corsi for example almost every player near the top has favourable zone starts while almost every player near the bottom has unfavourable zone starts every single year. On numbers like that it has a huge impact.

I think that's more correlation than causation. Better players generally get better zone starts.

I found the study I was remembering: Beware of What Zone Starts are Telling You, Part II: Shot Metrics

 

[Samus44]

Do you have a link to the charts?

From what I'm seeing, Sheahan is 12th on the Oilers forwards in QoC TOI% and a 40% defensive zone start% isn't all that extreme.

Yeah i linked it a couple comments up it's bolded in blue. I also posted it just now for everyone to see.

 

[McDNicks17]

View attachment 322263

That clearly shows a valuable player playing top competition with poor zone starts and surviving.

Thanks. I didn't see your other post with it.

I'm not a big fan of relative CF%.

For instance, it says Benson, Neal, Gagner, Granlund and Jurco are the top five forwards on the Oilers.

 

[Samus44]

I think that's more correlation than causation. Better players generally get better zone starts.

I found the study I was remembering: Beware of What Zone Starts are Telling You, Part II: Shot Metrics

Depends on the team and how the coach uses his players. Coaches want offensive guys to score but usually they aren't your most dependable checkers, that's why a Ryan O'Reilly is so damn good since he can do both.

Edit: I should add i didn't read the article. Too lazy. I'm confident in what i'm seeing and in my experience playing to know that they make a difference. Having dependable players who can win draws and not get scored on matters big time, it frees up your quality offensive players to only worry about scoring.

 

[Missing smitty]

All I know is they better well sign him, because my jersey is going in to get his number on it soon and I don't want to cancel it

 

[MessierII]

I think that's more correlation than causation. Better players generally get better zone starts.

I found the study I was remembering: Beware of What Zone Starts are Telling You, Part II: Shot Metrics

And here’s a quote directly from that article.

“The influence of shift start locations on Corsi-for percentages is actually quite large:

“

The impact is huge and the ridiculous math equation he tries to use to “adjust” is just that. Ridiculous. The fact is that zone starts have a massive impact on raw stats like corsi and fenwick.

 

[McDNicks17]

And here’s a quote directly from that article.

“The influence of shift start locations on Corsi-for percentages is actually quite large:

View attachment 322265“

The impact is huge and the ridiculous math equation he tries to use to “adjust” is just that. Ridiculous. The fact is that zone starts have a massive impact on raw stats like corsi and fenwick.

You have to read the whole article.

The whole point of it is saying zone starts have a significant effect on the first 10-15 seconds of ice time after a face off, but that ice time makes up such a small percentage of the player's total ice time that it has almost no statistical effect over the course of a season for players outside the extreme.

 

[Stoneman89]

No coincidence that since Archie and Shea have been here, our PK has been standout. Both of them on their way to 10-15 goals which is more than acceptable. And Archie is now showing a penchant for scoring some important goals in the clutch. That bodes well for playoff performance. Wasn't impressed with him at the beginning of the season, but he's really embraced the role Tip has given him and run with it.
I don't want them to give the farm away for him, but I'm sure Holland will work out something reasonable in terms of amount and term. 1.3 per for 2-3 years is a decent start.

 

[Samus44]

Thanks. I didn't see your other post with it.

I'm not a big fan of QoC relCF%.

For instance, it says Benson, Neal, Gagner, Granlund and Jurco are the top five forwards on the Oilers.

When i look at that chart i see a reflection of exactly how i'm seeing those players used. It's clear as day watching who's being used to check and score and the chart reflects that. I didn't even bother to look up player usage before my initial comment today yet lo and behold when i did the chart reflected exactly what i'm seeing, that suggests to me that relCorsi is very effective in displaying true QoC. Tippett is a very intentional coach.

But it's a fair point still and perhaps comparing the two charts might be even more beneficial. Here's one that i believe is using raw corsi.

Edmonton Oilers Player Usage Chart | leftwinglock.com

I'd suggest this one looks less accurate by my eye but still relevant. Perhaps the truth is in the middle? It's tough because stats are far from perfect evaluators given their limitations in providing context. I'd imagine if you asked Tippett how he's using them he'd tell you they're a checking line, hence he often starts them against other teams top lines to begin the game.

 

[Samus44]

Thanks. I didn't see your other post with it.

I'm not a big fan of relative CF%.

For instance, it says Benson, Neal, Gagner, Granlund and Jurco are the top five forwards on the Oilers.

Also I look at that chart and i can understand why those guys are top 5. All are used in a heavy offensive role except Granlund who faces weaker competition.

 

[McDNicks17]

When i look at that chart i see a reflection of exactly how i'm seeing those players used. It's clear as day watching who's being used to check and score and the chart reflects that. I didn't even bother to look up player usage before my initial comment today yet lo and behold when i did the chart reflected exactly what i'm seeing, that suggests to me that relCorsi is very effective in displaying true QoC. Tippett is a very intentional coach.

But it's a fair point still and perhaps comparing the two charts might be even more beneficial. Here's one that i believe is using raw corsi.

Edmonton Oilers Player Usage Chart | leftwinglock.com

I'd suggest this one looks less accurate by my eye but still relevant. Perhaps the truth is in the middle? It's tough because stats are far from perfect evaluators given their limitations in providing context. I'd imagine if you asked Tippett how he's using them he'd tell you they're a checking line, hence he often starts them against other teams top lines to begin the game.

I feel like most QoC stats are flawed. Relative ones, especially. If Gagner has a 52% CF beating up on 3rd/4th lines. QoC relCF% is going to say he's a tougher match up than a McDavid or Drai at 51% CF.

QoC TOI% makes the most sense to me since logically the best players are playing the most and the worst playing the least for the most part.

 

[CycloneSweep]

We need to keep him and Sheahan. I always feel anytime we let a GOOD 4th line role player go, it hurts us.

 

[MessierII]

You have to read the whole article.

The whole point of it is saying zone starts have a significant effect on the first 10-15 seconds of ice time after a face off, but that ice time makes up such a small percentage of the player's total ice time that it has almost no statistical effect over the course of a season for players outside the extreme.

That initial spike has a huge impact on the numbers. He admits that right off the hop.

 

[Jimmi McJenkins]

We need to keep him and Sheahan. I always feel anytime we let a GOOD 4th line role player go, it hurts us.

This is basically how I feel, instead of looking for replacement, keep the effective guys you have. As long as you pay them at a reasonable rate.

 

[SwedishFire]

View attachment 322263

That clearly shows a valuable player playing top competition with poor zone starts and surviving.

That chart says to me : dump Chiasson, sign Patrik Russell

 

[780il]

Keep him and Sheahan. Excellent 4th liners.

 

[CycloneSweep]

This is basically how I feel, instead of looking for replacement, keep the effective guys you have. As long as you pay them at a reasonable rate.

Our 2nd line is elite, love the 4th line, 1st line needs work. Our third line needs work.

Ideally you find a way to move Neal, if not...

New - McDavid - Neal/Kassian
RNH - Drai - Yamamoto
New - New - Neal/Kassian
New - Sheahan - Archibald

That should be Holland's makeup for next year. Fill in the new guys with new guys or graduating prospects like Benson.

Either way, Archibald needs to stay.
Ideally you kee

 

[LTIR]

He should be one of our shooters in shootouts. Both his snipes yesterday were goal scorer goals

 

[McDNicks17]

That initial spike has a huge impact on the numbers. He admits that right off the hop.

Yes. It has a huge impact for the 20 seconds after a face off.

The point of the article is that for a player like Archibald who has two more defensive zone draws than offensive draws at 5v5, that means we're talking about ~40 seconds of impacted ice time out of 516:47 minutes.

 

[Jimmi McJenkins]

Our 2nd line is elite, love the 4th line, 1st line needs work. Our third line needs work.

Ideally you find a way to move Neal, if not...

New - McDavid - Neal/Kassian
RNH - Drai - Yamamoto
New - New - Neal/Kassian
New - Sheahan - Archibald

That should be Holland's makeup for next year. Fill in the new guys with new guys or graduating prospects like Benson.

Either way, Archibald needs to stay.
Ideally you kee

Well the left side of Sheahan and Archibald will be an eventually healthy Nygard. I agree, if you could move Neal that would be terrific, I still wonder if you could trade him for Turris and have Turris fit as a #3C (work out the money of course)

 

[Samus44]

That chart says to me : dump Chiasson, sign Patrik Russell

How so?

 

[MessierII]

Yes. It has a huge impact for the 20 seconds after a face off.

Those 20 seconds are very likely to be the most eventful of the shift. Shot attempts and high quality chances are most likely to occur in that time frame.

Classic advanced stats argument. Just average everything out to make the important stuff seem less important. Same argument I’ve been having for 11 years now.

 

[McDNicks17]

Those 20 seconds are very likely to be the most eventful of the shift. Shot attempts and high quality chances are most likely to occur in that time frame.

Classic advanced stats argument. Just average everything out to make the important stuff seem less important. Same argument I’ve been having for 11 years now.

I'm not arguing against the significance of those 20 seconds. The only thing I'm arguing against is someone saying his overall numbers are poor because of zone starts.

 

[MessierII]

I'm not arguing against the significance of those 20 seconds. The only thing I'm arguing against is someone saying his overall numbers are poor because of zone starts.

It’s a big factor for sure.

 

[LaGu]

Can we change the title please??

I know I am tired, but it's the third time in two days I come on hfoil and think that he's been resigned for $1M...