The Advanced Stats Thread Episode VII: An Ode to the Sanity of Silverfish

[YoSoyLalo]

See that’s something that makes perfect sense for us and Chicago with tweaking. My original proposal was dumb until I edited it

It was dumb and you should be ashamed of yourself IMHO.

 

[Filthy Dangles]

Analytics aren't shot share. Shot share is overrated as a statistic, anyway. People need to get that out of their head that analytics = corsi

Analytics are basically anything. "Should I start my backup tonight" "Should I use my timeout right now" "Pull the goalie now or wait" etc.

 

[DanielBrassard]

Saad would be an awesome trade target.

 

[silverfish]

So I've been thinking a lot about finding 'the next Nate Schmidt' ie a guy who the "analytics community" knows is good, but the "traditional" hockey world hasn't really picked up on yet because he plays on his team's third pairing. I ran my similarity model on him for the 1617 year... top comparables:

1. Schultz - 1617
2. Shattenkirk - 1415
3. Jurcina - 0910
4. Wideman - 0809
5. Hickey - 1314
6. Kindl - 1213
7. Jensen - 1617
8. Franson - 1112
9. Babchuk - 0809
10. Stuart - 0809

Not bad company.

Effective use of something like this could be to then run Nate Schmidt this year, and see if any "low-key" names pop. Running that, I'm eyeballing Jan Rutta. So if we're trading for Saad, let's make CHI throw him in, too, even though it sort of feels like he didn't have the most impressive season.

Ah, f*** it, just sign Franson.

 

[Irishguy42]

So I've been thinking a lot about finding 'the next Nate Schmidt' ie a guy who the "analytics community" knows is good, but the "traditional" hockey world hasn't really picked up on yet because he plays on his team's third pairing. I ran my similarity model on him for the 1617 year... top comparables:

1. Schultz - 1617
2. Shattenkirk - 1415
3. Jurcina - 0910
4. Wideman - 0809
5. Hickey - 1314
6. Kindl - 1213
7. Jensen - 1617
8. Franson - 1112
9. Babchuk - 0809
10. Stuart - 0809

Not bad company.

Effective use of something like this could be to then run Nate Schmidt this year, and see if any "low-key" names pop. Running that, I'm eyeballing Jan Rutta. So if we're trading for Saad, let's make CHI throw him in, too, even though it sort of feels like he didn't have the most impressive season.

Ah, **** it, just sign Franson.

Pretty interesting stuff.

But to me, all this really just means that we're not leaving room for Nakladal.

 

[silverfish]

I'd also like to point out that I changed one of the parameters from primary points to just total points, since this is d-men, not forwards, and Neal Pionk's number 1 comparable is no longer Cody Ceci.

It's Marc Staal (09-10). Ceci is #2 by 1.4 points.

 

[Mac n Gs]

So I've been thinking a lot about finding 'the next Nate Schmidt' ie a guy who the "analytics community" knows is good, but the "traditional" hockey world hasn't really picked up on yet because he plays on his team's third pairing. I ran my similarity model on him for the 1617 year... top comparables:

1. Schultz - 1617
2. Shattenkirk - 1415
3. Jurcina - 0910
4. Wideman - 0809
5. Hickey - 1314
6. Kindl - 1213
7. Jensen - 1617
8. Franson - 1112
9. Babchuk - 0809
10. Stuart - 0809

Not bad company.

Effective use of something like this could be to then run Nate Schmidt this year, and see if any "low-key" names pop. Running that, I'm eyeballing Jan Rutta. So if we're trading for Saad, let's make CHI throw him in, too, even though it sort of feels like he didn't have the most impressive season.

Ah, **** it, just sign Franson.

Use Colin Miller from 1617 and Pesce from 1516 if you can, please and thank you

 

[silverfish]

Use Colin Miller from 1617 and Pesce from 1516 if you can, please and thank you

Top five for each:

Miller:

Krug - 1617
White - 0708
Martinez - 1112
Muzzin - 1314
Rundblad - 1415

Pesce:

Brian Lee - 0809
Spurgeon - 1112
Savard - 1112
Murphy - 1516
McBain -1011

 

[Machinehead]

Shit I forgot about Brian Lee

 

[silverfish]

**** I forgot about Brian Lee

So did hockey

 

[TomasHertlsRooster]

Tim Heed is the next defenseman that blows up after having strong analytics and getting buried by a questionable coaching staff. His treatment by Peter DeBoer and the Sharks coaching staff this year was incredibly strange and didn’t really make much of any sense at all. He will probably get traded this summer for some minimally valuable asset and will be fantastic for the next team that acquires him.

 

[YoSoyLalo]

Tim Heed is the next defenseman that blows up after having strong analytics and getting buried by a questionable coaching staff. His treatment by Peter DeBoer and the Sharks coaching staff this year was incredibly strange and didn’t really make much of any sense at all. He will probably get traded this summer for some minimally valuable asset and will be fantastic for the next team that acquires him.

Yeah we’ve had some discussions on here about Heed.

Would love to be the team that scoops him up.

 

[silverfish]

Tim Heed is the next defenseman that blows up after having strong analytics and getting buried by a questionable coaching staff. His treatment by Peter DeBoer and the Sharks coaching staff this year was incredibly strange and didn’t really make much of any sense at all. He will probably get traded this summer for some minimally valuable asset and will be fantastic for the next team that acquires him.

I like you.

Have we done Heed on the model, yet? I think so...

Drewiskie - 1213 (47.2)
Dallman - 0708 (20.99) [We've def done him because I remember Dallman kept coming up lol who]
ME Vlasic - 1314 (18.1)
Shattenkirk - 1415 (9.5)
Commodore - 0708 (7.89)

gimme gimme gimme

 

[TomasHertlsRooster]

Yeah we’ve had some discussions on here about Heed.

Would love to be the team that scoops him up.

Joakim Ryan is another guy who is downright fantastic, but he probably stays. His WOWYs with Brent Burns are laughably hilarious to the degree you’d almost think they were a glitch or an error in the number counting if you didn’t know better.

I like you.

I’m glad! HFSharks is a lot of fun but not always the most active and we’re obviously super biased towards most players. I like this thread a lot.

While we’re talking Sharks D; Marc-Edouard Vlasic and Justin Braun are totally cooked and the Sharks should trade both immediately while they still have a strong reputation.

 

[Ola]

Tim Heed is the next defenseman that blows up after having strong analytics and getting buried by a questionable coaching staff. His treatment by Peter DeBoer and the Sharks coaching staff this year was incredibly strange and didn’t really make much of any sense at all. He will probably get traded this summer for some minimally valuable asset and will be fantastic for the next team that acquires him.

The problem for Heed is opportunity. Burns takes so much of the ice time Heed needs to get to optimize his game.

Silverfish- Schmidt was always very impressing in Washington, many of the names on that compareable lists are pretty awful IMO, Franson for example.

I think the big problem is that teams aren’t moving these younger players. They hang on to them, miss use them and what not — but I recon they are smart enough to understand that trading them might look whoever pulling the trigger less than great.

 

[mike14]

Silverfish, what is the 'model' you are referring to, or at least what metrics does it take into account?
Sorry if you've explained it before, I went looking back through the thread but couldn't find anything.
Thanks!

 

[silverfish]

Silverfish, what is the 'model' you are referring to, or at least what metrics does it take into account?
Sorry if you've explained it before, I went looking back through the thread but couldn't find anything.
Thanks!

It's silverfish gorilla math that uses these metrics: p/60, iCF60, ixGF60, relTCF/A60, relTxGF/A60, relxFSh/Sv, TOI/GP, relZSr, xGFQoT/C to try and establish a list of players that are similar to one player you are evaluating.

 

[Machinehead]

Joakim Ryan is another guy who is downright fantastic, but he probably stays. His WOWYs with Brent Burns are laughably hilarious to the degree you’d almost think they were a glitch or an error in the number counting if you didn’t know better.



I’m glad! HFSharks is a lot of fun but not always the most active and we’re obviously super biased towards most players. I like this thread a lot.

While we’re talking Sharks D; Marc-Edouard Vlasic and Justin Braun are totally cooked and the Sharks should trade both immediately while they still have a strong reputation.

You're welcome any time!

 

[TomasHertlsRooster]

The problem for Heed is opportunity. Burns takes so much of the ice time Heed needs to get to optimize his game.

Silverfish- Schmidt was always very impressing in Washington, many of the names on that compareable lists are pretty awful IMO, Franson for example.

I think the big problem is that teams aren’t moving these younger players. They hang on to them, miss use them and what not — but I recon they are smart enough to understand that trading them might look whoever pulling the trigger less than great.

The thing is, I actually don’t think that San Jose will keep Heed. Their treatment of him was very, very strange.

He started off playing like 25 of the first 30-35 games. Then, he was replaced by Dylan DeMelo, and literally only played 2 more games in the season; he was benched by the 2nd period of both games.

 

[silverfish]

The thing is, I actually don’t think that San Jose will keep Heed. Their treatment of him was very, very strange.

He started off playing like 25 of the first 30-35 games. Then, he was replaced by Dylan DeMelo, and literally only played 2 more games in the season; he was benched by the 2nd period of both games.

I've been pumping Heed here for a while. I was miserable when I realized he had another year on his contract.

We'll take him, seriously. Give him a good home.

 

[silverfish]

Made another change. Flipped ixG60 to ixFSh. Figured we use attempts for quantity and expected shooting percentage for quality.

Not only is Cody Ceci's 1516 season back to being Pionk's top comparable, but Cody Ceci's 1617 season is now the 2nd highest comparable. Marc Staal relinquishes his crown and falls to third (0910).

I've also been messing with the viz that complements it... it's not fully there. The line of best fit isn't weighted towards players with higher similarity scores, it's treating everyone as equal which is wrong. But I honestly have no idea how to index and weight a relative to teammates metric across multiple players like this. Inconsequential at this junction.

 

[Mac n Gs]

No, YOU’VE watched Colin Miller’s first goal last night five times and have grumbled at what could have been.

 

[silverfish]

No, YOU’VE watched Colin Miller’s first goal last night five times and have grumbled at what could have been.

I guess this is more 'spec' thread or 'around the nhl' (which is why I think we still need to rename this thread something more along the lines of like: "Advanced Stats, Hockey Theory, and Hockey Economics", but I digress) but I believe so many GMs ****ed up expansion.

If I was a GM last June, I'd have scoured other teams' lists and gotten my hands on players cheap.

Hey, Boston, want to lose Colin Miller for free? Of course not. Here's a b-level prospect you won't have to protect so you get something instead of nothing for him.
Hey Washington, want to lose Nate Schmidt for free? Of course not. Here's a b-level prospect you won't have to protect so you get something instead of nothing for him.

To prep for this, you buy-out Staal and free up his slot, and expose Nick Holden.

McDonagh - Miller
Skjei - Shattenkirk
Smith - Schmidt

oh my ****ing god.

 

[YoSoyLalo]

nice blueline, but it would still be coached by av

you think he wouldnt play kampfer over miller? pffft.

 

[Mac n Gs]

I guess this is more 'spec' thread or 'around the nhl' (which is why I think we still need to rename this thread something more along the lines of like: "Advanced Stats, Hockey Theory, and Hockey Economics", but I digress) but I believe so many GMs ****ed up expansion.

If I was a GM last June, I'd have scoured other teams' lists and gotten my hands on players cheap.

Hey, Boston, want to lose Colin Miller for free? Of course not. Here's a b-level prospect you won't have to protect so you get something instead of nothing for him.
Hey Washington, want to lose Nate Schmidt for free? Of course not. Here's a b-level prospect you won't have to protect so you get something instead of nothing for him.

To prep for this, you buy-out Staal and free up his slot, and expose Nick Holden.

McDonagh - Miller
Skjei - Shattenkirk
Smith - Schmidt

oh my ****ing god.

See the main issue is that those two were viewed as expendable, and that’s the problem in of itself. How/why are talented players like that being used in lesser roles when they’re obviously outperforming them?

How can you project/predict that they will perform against better players. How can you maximize player efficiency using the system the coach practices? Is that system actually efficient and how can it be improved?