Post-Game Talk: Leafs lose 4-3 in OT

[Nithoniniel]

Let's move on. My question was clear. Advanced stats as a term is widely used.

Different stats are probably more accurate.

You never posed a question. You made a claim that proponents say advanced stats to seem superior. Looking at some mainstream media intro articles doesn't support that.

[...] digging into the weeds to understand what each data point mean.

I doubt anybody that had an education in statistics or worked with it professionally would ever say something like this.

 

[Pookie]

You never posed a question. You made a claim that proponents say advanced stats to seem superior. Looking at some mainstream media intro articles doesn't support that.

You don't think mainstream media and the average HF poster who cites them would be considered a "proponent"?

 

[diceman934]

IIRC Babcock has criticized the shot clock numerous times. If the team's compiled data is different than the official data being used in formulating these stats, is the team still using the same publically available stat?

Lol ...please these are not even close to what has been discussed.

 

[Nithoniniel]

You don't think mainstream media and the average HF poster who cites them would be considered a "proponent"?

OK. Let's say that's so.

... So when you said that certain people used the term 'advanced stats' to seem superior, you are then talking about mainstream media? Because that's the topic here. Otherwise you are free to specify who exactly you think uses that term to be superior, and we can check whether that's true.

 

[Bluelines]

I doubt anybody that had an education in statistics or worked with it professionally would ever say something like this.

Well you would be wrong then because I just did.

This is what I'm talking about, this guy gets it. No metric is perfect. Be cognizant of flaws and benefits. When you are a weekend GM and don't understand the flaws and use analytics as the be all and end all, you show that you don't understand the metric.

There is value in numbers as long as the person who is quoting them uses them in the right context and with support of other data points that proves the hypothesis.

 

[Bluelines]

You don't think mainstream media and the average HF poster who cites them would be considered a "proponent"?

I think they call themselves Level 73 uber Hockey gods, able to whip out a useless fan stat at the drop of a Funyun.

 

[Nithoniniel]

This is what I'm talking about, this guy gets it.

That's what everybody says. I've seen Zeke say it, I've seen every single high profile public analytics guy say it. I say it.

Everybody who deals with analytics know there are limitations to both accuracy and scope of what a metric tells you.

And let's just say that I seriously doubt your claim.

 

[Bluelines]

That's what everybody says. I've seen Zeke say it, I've seen every single high profile public analytics guy say it. I say it.

Everybody who deals with analytics know there are limitations to both accuracy and scope of what a metric tells you.

And let's just say that I seriously doubt your claim.

Which claim is that?

 

[Nithoniniel]

Which claim is that?

Here's a thought I get paid quite well to be an analyst.

Just a suggestion. People aren't just going to take your word for it around here. Especially when you talk about it in a way that is distinctly different from verified sources. You don't do yourself a favor by appealing to your own authority, whether or not it is true.

 

[Pookie]

OK. Let's say that's so.

... So when you said that certain people used the term 'advanced stats' to seem superior, you are then talking about mainstream media? Because that's the topic here. Otherwise you are free to specify who exactly you think uses that term to be superior, and we can check whether that's true.

We don't need to do that.. particularly in a game thread. I'm not debating the use of "different" stats with McCurdy or Dubas. I'm posting here on HF boards.

I made a general comment that I "think" people like to use the term advanced stat as it allows them to feel as though on a pedestal... as if they have some special insight that others don't.

I still recall one fellow posting results of a study claiming that Corsi/Fenwick were predictive of wins... and when I cut and pasted the part of the study that actually discussed wins and prediction and it was found to be pretty low... explains the results about 18% of the time for a limited number of games in a season... you'd think that a fan of the stats would have simply said... "oh, my mistake. Not completely useless but I guess if that's what the study I'm referencing says, I need to acknowledge the limitation."

Now, this fellow is well known probably has some likes. (I can't see him anymore as he's on ignore).

Doesn't make mistakes all the time, I'm sure. Folks lap up his numbers. I'm sure they are nice numbers. But in that instance, he showed he'd rather stay on the pedestal (falsely) than elaborate and provide context to his numbers. This is where I developed my "pedestal" theory.

100% respect for anyone that offers an insight and can also acknowledge the limitation of it so we have context to form an opinion. That's why I asked about the t test to start this.

Happy to turn the discussion back to the game itself.

 

[zeke]

What's your opinion on game scores?
I see a few of the good analytics guys posting it on twitter.

Season average

Matthews 1.17
JVR 0.76
Nylander 0.75
Rielly 0.73
Marner 0.72
Marleau 0.63
Hyman 0.55
Bozak 0.55
Gardiner 0.44
Brown 0.36
Hainsey 0.34
Leivo 0.30
Polak 0.29
Moore 0.28
Martin 0.27
Borgman 0.26
Komarov 0.18

not a huge fan in previous years but these models get improved every year so i'd have to study up on that one.

this list passes the smell teat at least.

in general for me most of the consolidated stats don't do a great job of accounting for quality of competition, and imo this is the biggest problem with most analytics so far.

 

[Pookie]

I think they call themselves Level 73 uber Hockey gods, able to whip out a useless fan stat at the drop of a Funyun.

What's a Funyun?

I know of the food but is this a Twittergrambook thing?

 

[Bluelines]

Just a suggestion. People aren't just going to take your word for it around here. Especially when you talk about it in a way that is distinctly different from verified sources. You don't do yourself a favor by appealing to your own authority, whether or not it is true.

Who hurt you Nit to make you so cynical? If your gut is calling BS on me, well you can stop trusting your gut.

I double pinky swear, I do get paid quite well, my bonus last year was more than most people make in 6 months, my concentration is in analytics. If you don't believe me that is your prerogative, its really no skin off my back if you don't believe the truth.

 

[Nithoniniel]

I made a general comment that I "think" people like to use the term advanced stat as it allows them to feel as though on a pedestal... as if they have some special insight that others don't.

Yes, and then you supported that by quoting mainstream media. You then asked me whether or not mainstream media could not be included in that group of proponents. But it's fun to see you put this much work in to try to twist things around after painting yourself in a corner.

This is where I developed my "pedestal" theory.

So you not only make claims about a group of people based on one person, it's a person that in fact doesn't really like the term to begin with either. Wow.

And I remember that discussion. I remember people trying to explain it to you. And I remember you going about your usual intellectually dishonest ways to avoid having to acknowledge it.

 

[Bluelines]

What's a Funyuns?

I know of the food but is this a Twittergrambook thing?

It's the "go to" stoners snack in North America

 

[zeke]

Here's a thought I get paid quite well to be an analyst, I'm not anti analytics I actually love analytics, I'm anti stupid fan stats, its people like you use stupid fan stats to perch yourself up as a person of knowledge, people like you who don't fully understand the depth of the stats you drool out, it annoys people like me who have spent decades honing my craft. I actually make a living digging into the weeds to understand what each data point mean. I work with people like you who "think" they understand the metrics but all it does it make my job harder because I have to take valuable time away from useful work to explain that an apple is not always an apple.

If nothing your "analysis" are entertaining, carry on.

it's funny because you repeatedly rely on legit stupid meaningless stats in your arguments, and blindly argue against analytics which any well-paid analyst in any field knows are actual simple, basic, necessary tools that everyone and every industry uses.

p.s. the plural of analysis is analyses, mr.well-paid analyst.

 

[Nithoniniel]

Who hurt you Nit to make you so cynical? If your gut is calling BS on me, well you can stop trusting your gut.

If you give me no reason to believe you, I won't. That's not being cynical, that's healthy skepticism.

Also, this:

[...]argue against analytics which any well-paid analyst in any field knows are actual simple, basic, necessary tools that everyone and every industry uses.

 

[Bluelines]

Yes, and then you supported that by quoting mainstream media. You then asked me whether or not mainstream media could not be included in that group of proponents. But it's fun to see you put this much work in to try to twist things around after painting yourself in a corner.


So you not only make claims about a group of people based on one person, it's a person that in fact doesn't really like the term to begin with either. Wow.

And I remember that discussion. I remember people trying to explain it to you. And I remember you going about your usual intellectually dishonest ways to avoid having to acknowledge it.

Nit some credit to you, I dont always agree with you but you debate the right way. Kudos. This is sincere BTW, JIC you think I'm being sarcastic.

 

[Sniper99]

Who hurt you Nit to make you so cynical? If your gut is calling BS on me, well you can stop trusting your gut.

I double pink swear, I do get paid quite well, my bonus last year was more than most people make in 6 months, my concentration is in analytics. If you don't believe me that is your prerogative, its really no skin off my back if you don't believe the truth.

you can double pinky swear all you want. Do adults really "pinky swear" anymore. You look like one of those posters in the trades boards that make bad proposals and then they say they have "good sources" yet wont prove anything. It really doesn't look good on you.

 

[Menzinger]

If I'm searching for a FO guy... I'm probably going to look at FO% as a stat and then reaction times in a "combine" setting.

If I'm searching for a shot blocker, I'm going to peak at that stat.

My point was that you can't use them unless you understand the context and what you are measuring. In this thread, the insinuation was that a decline in a Corsi number was a bad thing.

No context given as to whether the data was statistically significant and whether the player has been used differently. His output in December was a point or two shy of that in November. Is that a big deal?

When using it as a rank order to compare our team vs others, we have no context. We make all those assumptions (ie. All teams playing the same system, all the time, every shift, with every talent... ) and conclude that we are either good or bad relative to how we rank on the chart. That's just not reality.

I’m not sure what type of context you’re looking for when it comes to corsi - I doubt there’s a a single coach in the league who would say it’s a good thing that the other team controls the puck more.... different systems can produce positive possession numbers.

As for number accuracy it’s is best thought of a broad indicator rather than anything else 50% + good, 55%+ elite, sub 45% terrible.

 

[Bluelines]

If you give me no reason to believe you, I won't. That's not being cynical, that's healthy skepticism.

LOL, you looking for my 2016 tax return? Let me just give you my SIN number, my three account numbers for my investments and my drivers licence number too.

Not for nothing but I usually only call BS when I have proof of said BS. Your gut is wrong in this case.

Cheers.

 

[Nithoniniel]

LOL, you looking for my 2016 tax return? Let me just give you my SIN number, my three account numbers for my investments and my drivers licence number too.

Not for nothing but I usually only call BS when I have proof of said BS. Your gut is wrong in this case.

Cheers.

No, I'm just suggesting that you don't make claims when you can't back them up. It's what a lot of people do when they don't have any good arguments, so it will hurt your argument more than help it.

 

[Bluelines]

you can double pinky swear all you want. Do adults really "pinky swear" anymore. You look like one of those posters in the trades boards that make bad proposals and then they say they have "good sources" yet wont prove anything. It really doesn't look good on you.

Kinda hard to prove I had a discussion with that scout, next time I'll take out my phone and ask him if its cool if I record him, people love being candid in today's day of social media. Kinda hard to prove I make a living doing what I said I do without giving up sensitive personal information. The pinky swear was kinda a playful joke, maybe it went over your head.

Hey if you are cynical too, that's on you not me.

 

[Menzinger]

Read any analytics blog or podcast and folks who follow them totally reject the term “advanced stats” or “fancy stats”, they’re just new stats. They’re given that label by folks in the media to help differentiate them from the “traditional ones”, but in a couple years they’ll be so integrated that any separation between old and new will disappear, we’ll just call them “stats”

 

[Sniper99]

Kinda hard to prove I had a discussion with that scout, next time I'll take out my phone and ask him if its cool if I record him, people love being candid in today's day of social media. Kinda hard to prove I make a living doing what I said I do without giving up sensitive personal information. The pinky swear was kinda a playful joke, maybe it went over your head.

Hey if you are cynical too, that's on you not me.

nope not at all.