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

[Ail]

Going from 788 to 741 is a steep drop to take willingly.

Is it? The difference in those two numbers isn't even a 7th round pick according to the chart, so how exactly are we quantifying what 788 vs 741 is?

 

[silverfish]

Anyone have a good read on underrated volume shooters/goal scorers going into this draft, or do I have to put that research together myself? I wish there was an easy way to just query the major junior league stats sites. I'm considering this 'advanced stats' talk because I'm curious about SOG/GP and I know once that `/` comes into play, people tend to lose their shit (ie. points per 60).

Off the top of my head, I know whichever team spends a fourth-rounder on Jeremy McKenna is going to get, at worst, a guy who carries their AHL team. 36 goals and over 300 SOG but he's going to go in the 4th round because he's 5'9". Would put money on him putting up 50g in the Q next year.

 

[Oscar Lindberg]

@silverfish one of the mid round guys I like who is a very solid goal scorer is Sampo Ranta

He would need work on his defensive side of his game to make to the NHL, but he has a great shot and was one of the best rookie goal scorers in the USHL last year

 

[silverfish]

@silverfish one of the mid round guys I like who is a very solid goal scorer is Sampo Ranta

He would need work on his defensive side of his game to make to the NHL, but he has a great shot and was one of the best rookie goal scorers in the USHL last year

Mid-round? He's the 18th ranked NA skater.

Or did you mean mid-round as the middle of the first round and I was wrong to read mid-round as meaning like a 3rd or 4th rounder. My bad.

 

[Oscar Lindberg]

Mid-round? He's the 18th ranked NA skater.

Or did you mean mid-round as the middle of the first round and I was wrong to read mid-round as meaning like a 3rd or 4th rounder. My bad.

Mid round meaning like 3ish round. He may go as high as the late second.

When I first heard about him he was hovering around late third early fourth

I looked at his shots, he didn’t take an enormous amount. About 163 for the season. He’s not a volume shooter like the guy you mentioned in your previous post, but he can wire the puck

 

[Filthy Dangles]

So, if I read the chart correctly....the 1st OA == 9th OA + 23 OA.

Is there a single soul in the front office executive world that would even consider that for a second without laughing in your face?

 

[aufheben]

Shouldn't there be a huge drop off in value at around the middle of the 1st round?

 

[Irishguy42]

Shouldn't there be a huge drop off in value at around the middle of the 1st round?

Yes.

https://myslu.stlawu.edu/~msch/sports/Schuckers_NHL_Draft.pdf

 

[Doctyl]

So, if I read the chart correctly....the 1st OA == 9th OA + 23 OA.

Is there a single soul in the front office executive world that would even consider that for a second without laughing in your face?

Nope, which is the point here. The drop off from 1st to 9th isn’t worth the overpayment that the 1st overall pick would command.

 

[Filthy Dangles]

Nope, which is the point here. The drop off from 1st to 9th isn’t worth the overpayment that the 1st overall pick would command.

Agreed but it can be both; it's not worth overpayment the industry has established and that the chart might be undervaluing the picks (at least in certain draft years)

Like, it's obviously insanely different if it's the year Crosby or McDavid go 1st OA vs if it's Ryan Murray or Erik Johnson draft year.

The draft is obviously a crapshoot but certain years you get really special players that you can project more easily, there's a lot of variance.

 

[Doctyl]

Agreed but it can be both; it's not worth overpayment the industry has established and that the chart might be undervaluing the picks (at least in certain draft years)

Like, it's obviously insanely different if it's the year Crosby or McDavid go 1st OA vs if it's Ryan Murray or Jack Johnson draft year.

The draft is obviously a crapshoot but certain years you get really special players that you can project more easily, there's a lot of variance.

I think that chart just uses averages and doesn’t take into account specific cases like McDavid or Dahlin.

 

[Mac n Gs]

I think that chart just uses averages and doesn’t take into account specific cases like McDavid or Dahlin.

It also needs to be updated with new data. There's been a lot more value at later picks with how prospects are getting better.

I think we'd have to look into something like the leaked draft value sheet from Tampa. That had some interesting insight on it

 

[Irishguy42]

I believe the chart has been updated more recently. I think you'd have to dig through Tierney's Twitter TL. Not sure.

 

[Machinehead]

Is it? The difference in those two numbers isn't even a 7th round pick according to the chart, so how exactly are we quantifying what 788 vs 741 is?

Considering you gave up two picks for one in an event that's almost entirely probability, to LOWER your quality, yes it is.

 

[Machinehead]

Caps CF% in the playoffs is very, very good.

How are they all of a sudden outplaying teams when they haven't for 164 regular season games?

 

[FireGerardGallant]

Caps CF% in the playoffs is very, very good.

How are they all of a sudden outplaying teams when they haven't for 164 regular season games?

They prob coasted through the regular season and are improve their play when it matters most

 

[Machinehead]

They prob coasted through the regular season and are improve their play when it matters most

I never bought into that but I suppose it's possible.

I guess we'll really find out the rest of this series. They got killed that period.

 

[DanielBrassard]

Pens also missing Hags and Malkin to be fair. 1st round was a bit surprising though.

 

[silverfish]

Caps CF% in the playoffs is very, very good.

How are they all of a sudden outplaying teams when they haven't for 164 regular season games?

What is it on a game by game basis? One huge game in a sample this tiny can inflate numbers.

Gut check says they win the CF% battle 3/7 games so far, but when they win it, they win it.

 

[Machinehead]

What is it on a game by game basis? One huge game in a sample this tiny can inflate numbers.

Gut check says they win the CF% battle 3/7 games so far, but when they win it, they win it.

So far it's the opposite. They've won all but two and got lobotomized in those two.

Could chalk it up to Columbus just being ass, and like I said, we'll see what the Pens do.

 

[silverfish]

Are you sure?

Should note that this is 5v5 unadjusted

 

[Machinehead]

I may have gotten my signals crossed.

 

[silverfish]

Slowly but surely working through, what will surely be a terrible, performance projection of Neal Pionk.

Unless of course it shows that he will be great, then I will pretend I never did anything. If it's analysis that proves you're wrong, then delete all traces of it. The point of analytics is to prove what you already know. Straight up, b.

 

[Mac n Gs]

Slowly but surely working through, what will surely be a terrible, performance projection of Neal Pionk.

Unless of course it shows that he will be great, then I will pretend I never did anything. If it's analysis that proves you're wrong, then delete all traces of it. The point of analytics is to prove what you already know. Straight up, b.

Won’t it just tell us that he’ll get blitzed in shot attempts if he’s used in a similar fashion next season? There’s no way he’ll be getting 18 min of 5v5 time and 23 mpg next season unless Shattenkirk/ADA both evaporate

 

[silverfish]

Won’t it just tell us that he’ll get blitzed in shot attempts if he’s used in a similar fashion next season? There’s no way he’ll be getting 18 min of 5v5 time and 23 mpg next season unless Shattenkirk/ADA both evaporate

What I'll be doing is comparing him to players who performed similarly to him in their 22 year-old season using a mix of relative, individual, and context metrics. Then evaluating how the players similar to him in their 22 year-old season advanced throughout the rest of their career.

It's my old MILLER model, but I needed to re-work for d-men.

It's also not very scientific, and should be taken with millions of grains of salt and lots and lots of skepticism.