The Advanced Stats Thread Episode VI: RIP To Our Databases

[Raspewtin]

I've read most of them and those are still more polite

You missed me and silverfish killing each other over Girardi being on the team three years ago, literally every day.

I'm serious when I say we likely fought about Girardi every day of the 2015 offseason.

Those were the days.

 

[Mac n Gs]

You missed me and silverfish killing each other over Girardi being on the team three years ago, literally every day.

I'm serious when I say we likely fought about Girardi every day of the 2015 offseason.

Those were the days.

What a time. My favorite arguments were when that poster, whose name I can't remember for the life of me, argued for dumping Girardi for Wiz

 

[DanielBrassard]

What a time. My favorite arguments were when that poster, whose name I can't remember for the life of me, argued for dumping Girardi for Wiz

Blue Blooded. I miss that dude.

 

[lilphildub]

You missed me and silverfish killing each other over Girardi being on the team three years ago, literally every day.

I'm serious when I say we likely fought about Girardi every day of the 2015 offseason.

Those were the days.

And look at you guys being so close and agreeing on things now #bromance

 

[Machinehead]

Things that are funny to me:

Stat-line prediction threads. You can go in there, and spout off whatever numbers you want with no basis, and nobody cares. But if someone went in there with an analytical study that said, here's what I expect for players this year based on historical performance, similar players, and age trends. Well, people would lose their minds.

For the past week or so, I've only been posting in two threads on HFNYR; this one, and the football thread. Lurking everywhere else. It's been nice.

When I do those I think 25% about the numbers and just do like 75% on gut.

I mean, it's a fun thread, but you're right that it's weird how nobody will call you out on gut numbers, but people react poorly to concrete numbers.

 

[Machinehead]

The idea that there must be a good reason behind your TOI is up there with the Earth being flat.

Dan Girardi alone murders that idea in a dark alley but the list is long.

The real reason is that coaches play favorites. All of them. Every. Single. One.

 

[Machinehead]

And just as I'm going on about how there's no logical authority among those in charge, Glass gets a tryout.

 

[silverfish]

You missed me and silverfish killing each other over Girardi being on the team three years ago, literally every day.

I'm serious when I say we likely fought about Girardi every day of the 2015 offseason.

Those were the days.

Exactly what I was referencing when I said earlier that the first few iterations of this thread were bloodbaths. Fun, though. If nothing else, at least a debate rooted in realistic terms. Did a lot of learning early on.

And just as I'm going on about how there's no logical authority among those in charge, Glass gets a tryout.

Pretty good sign that DJ was hired to be a database engineer and not an analyst, one would hope, for DJ's sake. Does anyone remember what Glass's relSv% was?

 

[Igor Shestyorkin]

Exactly what I was referencing when I said earlier that the first few iterations of this thread were bloodbaths. Fun, though. If nothing else, at least a debate rooted in realistic terms. Did a lot of learning early on.



Pretty good sign that DJ was hired to be a database engineer and not an analyst, one would hope, for DJ's sake. Does anyone remember what Glass's relSv% was?

Pretty good.

 

[BlueshirtBlitz]

What do you stat-heads think about the chances for Hank being a Vezina finalist this year? Even with age-decline, I remember plenty of charts regarding shots in the slot and tables showing Hank's numbers (not sure if there are consensus goalie metrics yet, I think I remember Machinehead sharing one) were being dragged down by the number of high danger shots he was taking compared to league average. Have to expect, or hope, that won't be the case with Shatty and more ice-time for Skjei.

 

[Machinehead]

If Hank wants to win the Vezina, he has to be better on the penalty kill. His PK metrics are deplorable.

*whispers*

Lundqvist is the reason our penalty kill has sharply declined the last 2-3 years.

 

[DanielBrassard]

Exactly what I was referencing when I said earlier that the first few iterations of this thread were bloodbaths. Fun, though. If nothing else, at least a debate rooted in realistic terms. Did a lot of learning early on.



Pretty good sign that DJ was hired to be a database engineer and not an analyst, one would hope, for DJ's sake. Does anyone remember what Glass's relSv% was?

There were some great debates during that period of time for sure. I kind of miss it lol.

 

[Igor Shestyorkin]

Doesn't matter how many high-danger chances Girardi and Staal let up, they had good +/-.

 

[Mac n Gs]

Finally starting my journey learning R. Gonna be cool since I'll be able to directly apply it my job right away

 

[silverfish]

Finally starting my journey learning R. Gonna be cool since I'll be able to directly apply it my job right away

Jealous that you can apply things you learn in R in your 'real world' job. That's going to do wonders for the learning process.

Bring everything you learn here

 

[PlamsUnlimited]

Gonna post here since stats are kind of bleeding into roster thread but since a lot of sites are getting privatized or taken down due to hiring, would anyone want to try to do some of our own for the Rangers with me? Little project I want to try.

 

[silverfish]

Gonna post here since stats are kind of bleeding into roster thread but since a lot of sites are getting privatized or taken down due to hiring, would anyone want to try to do some of our own for the Rangers with me? Little project I want to try.

If you're gonna take the time to make one for the Rangers specifically, you should just do the whole league. With the right coding, all it should take is more time to process

 

[PlamsUnlimited]

If you're gonna take the time to make one for the Rangers specifically, you should just do the whole league. With the right coding, all it should take is more time to process

I am still learning coding. Some of it is hard to remember for a newbie.

This would be a massive undertaking for me so that's why I want to start small I will probably need your help

 

[romba]

Yeah I'd agree with that, albeit it's weird to see his goal:assist splt be 1:15. Which when you think about it, it's ****ing bonkers that he has 40 playoff games and 1 goal. He and Hayes suck in the playoffs. Regression is a *****

Hayes and Miller don't suck in the playoffs because of regression. Regression should be random with regard to peaks and valleys but they seem to do it at a specific short period each year. That implies an independent cause for the repression relating to the specific time period. Potentially something like 'tightening defensive adjustments against higher level of competition'.

Similar to the team. Start out on fire (offensive scheme capitalizes on teams still figuring out their defensive schemes and new personnel), jump into 1st with a cushion, ride said cushion in to playoffs while trying and failing to 'recapture' magic from first few months of the season. Enter playoffs and proceed to flip a coin each game for a win or loss, then play some golf when luck runs out. Rinses and repeat.

 

[silverfish]

I am still learning coding. Some of it is hard to remember for a newbie.

This would be a massive undertaking for me so that's why I want to start small I will probably need your help

I way oversimplified it anyway

Happy to be of help as I can be.

 

[PlamsUnlimited]

I way oversimplified it anyway

Happy to be of help as I can be.

I'm learning a lot still. Luckily python has free online tutorial books

 

[silverfish]

I'm learning a lot still. Luckily python has free online tutorial books

Oh that's right, you went with Python, not R. Turns out I probably won't be able to help

 

[PlamsUnlimited]

Oh that's right, you went with Python, not R. Turns out I probably won't be able to help

I'm still a noob so I don't think I've gotten into anything super hard yet. So there may be hope!

I hope I can do this since it'll be tons of tracking and more focus on just about everything.

 

[silverfish]

I'm still a noob so I don't think I've gotten into anything super hard yet. So there may be hope!

I hope I can do this since it'll be tons of tracking and more focus on just about everything.

This package may save you a lot of time: https://pypi.python.org/pypi/nhlscrapi/

I use the "R version" of this.

 

[TheTakedown]

Doesn't matter how many high-danger chances Girardi and Staal let up, they had good +/-.

I ****ing hated when people would use that argument... I am so over +/- that I removed it from our fantasy league this year