The Advanced Stats Thread Episode VI: RIP To Our Databases

[nyr2k2]

I think its more that people don't understand it than they disagree with it...people see all of these charts and graphs with pretty colors but don't fully understand the meaning behind the charts and how they are determined. and most people aren't willing to take the time to learn it like its a statistics class. so for the non-analytics people you see a chart and might not know if it is good or bad...

corsi has become more commonly accepted because people know its shots for vs shots against. at least at high level people understand what they are looking at...

as someone that wants to understand them more but is way too lazy to actually research it, i think the analytic community would be better served focusing on a handful of key stats/charts and trying to make the understanding of those stats common knowledge instead of trying to come up with new stats that will just confuse the non-analytic people even more

I normally steer clear of this thread but I checked in and this post caught my eye.

I jumped on the advanced stats in baseball bandwagon really early on. Like, I remember discussing it on Usenet back in the day. I liked that stuff because I could understand it.

On HF, a lot of times I see a chart with a bunch of colors and lines or dots or whatever presented without much (if any) explanation and I have literally no idea what it means. It's just...there. And then I'm like **** that, I don't have the time to look up a bunch of acronyms or read a paper to figure out what this data indicates. The HERO chart or whatever it is, I like that one because even if I don't understand the underlying metrics, I see bars comparing things and even I can understand what that indicates.

So, I WANT to understand this stuff more, I just feel like a lot of people want to share the data without giving the "uneducated" among us a proper explanation of what it means. This isn't everyone, it's just something I see a lot.

 

[Machinehead]

I normally steer clear of this thread but I checked in and this post caught my eye.

I jumped on the advanced stats in baseball bandwagon really early on. Like, I remember discussing it on Usenet back in the day. I liked that stuff because I could understand it.

On HF, a lot of times I see a chart with a bunch of colors and lines or dots or whatever presented without much (if any) explanation and I have literally no idea what it means. It's just...there. And then I'm like **** that, I don't have the time to look up a bunch of acronyms or read a paper to figure out what this data indicates. The HERO chart or whatever it is, I like that one because even if I don't understand the underlying metrics, I see bars comparing things and even I can understand what that indicates.

So, I WANT to understand this stuff more, I just feel like a lot of people want to share the data without giving the "uneducated" among us a proper explanation of what it means. This isn't everyone, it's just something I see a lot.

Hockey stats are actually significantly easier to understand than what you're used to in baseball.

Most of our charts just kind of suck. (Silverfish makes good stuff)

And yes, many people just throw data without explaining it.

If you ever see something like this, bring it over. That's what we're here for

 

[silverfish]

I normally steer clear of this thread but I checked in and this post caught my eye.

I jumped on the advanced stats in baseball bandwagon really early on. Like, I remember discussing it on Usenet back in the day. I liked that stuff because I could understand it.

On HF, a lot of times I see a chart with a bunch of colors and lines or dots or whatever presented without much (if any) explanation and I have literally no idea what it means. It's just...there. And then I'm like **** that, I don't have the time to look up a bunch of acronyms or read a paper to figure out what this data indicates. The HERO chart or whatever it is, I like that one because even if I don't understand the underlying metrics, I see bars comparing things and even I can understand what that indicates.

So, I WANT to understand this stuff more, I just feel like a lot of people want to share the data without giving the "uneducated" among us a proper explanation of what it means. This isn't everyone, it's just something I see a lot.

The problem with this is, is that it takes a lot of time to explain something every time you post it. And, in the beginning, when I was doing this, it was often met with a lot of vitriol. Which I guess I understand. But, it made wanting to spend the time explaining everything tiresome and tedious knowing that there would be a lot of pushback. Which was unfortunate for me, as I felt bad for the posters who were curious and wanted to learn. I did my best to direct them to this thread. Some came, some did not.

Which is why I like that this thread exists because I hope it's a place where the curious can come ask the questions they want answered and everyone who is "anti-advanced-stats" can just ignore it. Sometimes that's how it works, sometimes not. This is an internet forum, after all

Most of our charts just kind of suck. (Silverfish makes good stuff)

[Always looking for and appreciate all feedback if it's something that I'm making, btw]

 

[nyr2k2]

All right, the next time I find something that looks interesting that I don't understand, I'll bring it up here.

 

[Revel]

Hockey stats are actually significantly easier to understand than what you're used to in baseball.

Most of our charts just kind of suck. (Silverfish makes good stuff

I agree that Hockey Advanced Stats are not that difficult to understand and that it is the graphic presentations that suck. Much too busy most of the time. So much so...that I prefer to view numeric tables.

But I do view graphics on occasion. What I like about Silverfish is that he is open and inviting when it comes to creating his stuff. He looks for feedback and often incorporates it in the hopes that people will be more drawn into the analytic world.

My interest in Advanced Stats waxes and wanes. When it's waxing, it's usually because of something Silverfish has presented.

P.S. - He has even created high contrast presentations so us "color challenged" folk can follow along. It's much appreciated!

 

[silverfish]

I agree that Hockey Advanced Stats are not that difficult to understand and that it is the graphic presentations that suck. Much too busy most of the time. So much so...that I prefer to view numeric tables.

But I do view graphics on occasion. What I like about Silverfish is that he is open and inviting when it comes to creating his stuff. He looks for feedback and often incorporates it in the hopes that people will be more drawn into the analytic world.

My interest in Advanced Stats waxes and wanes. When it's waxing, it's usually because of something Silverfish has presented.

P.S. - He has even created high contrast presentations so us "color challenged" folk can follow along. It's much appreciated!

You're the reason the post-game chart dumps are 5 charts instead of 4

Think my DFS charts wouldn't work for you at all - https://imgur.com/a/AE44p

But seriously, I'm happy to help and always want feedback. I make the charts to post on here, so it only makes sense to make sure that the end-user is getting what they want

 

[Filthy Dangles]

Alright. Let's start a serious discussion about some adv stats.

What kind of stats could you exploit to gain an edge in DFS Hockey? Besides the obvious of looking for players who are shooting well over/under their career Shooting %'s and stuff like that.

 

[silverfish]

Alright. Let's start a serious discussion about some adv stats.

What kind of stats could you exploit to gain an edge in DFS Hockey? Besides the obvious of looking for players who are shooting well over/under their career Shooting %'s and stuff like that.

Man, if you find out, let me know. I've tried all sorts of things on this and lost $ testing them

With DFS, at least on DraftKings, I don't think there's anything more important that playing players who shoot a lot, and who play on the PP. (0.5 points per shot on goal).

SA16 might have some knowledge on this, but I'm not sure if he dabbles in DFS hockey.

 

[Filthy Dangles]

Man, if you find out, let me know. I've tried all sorts of things on this and lost $ testing them

With DFS, at least on DraftKings, I don't think there's anything more important that playing players who shoot a lot, and who play on the PP. (0.5 points per shot on goal).

SA16 might have some knowledge on this, but I'm not sure if he dabbles in DFS hockey.

Volume shooters who are playing against bad shot suppression teams?

Obviously, volume shooters are usually superstar players but I'd say maybe a lowkey volume shooter who isn't a star player (thus would be cheap and valuable on DraftKings) playing against a bad shot suppression team like TOR/NYR/PIT could be valuable and rack up some points.

 

[Synergy27]

Man, if you find out, let me know. I've tried all sorts of things on this and lost $ testing them

With DFS, at least on DraftKings, I don't think there's anything more important that playing players who shoot a lot, and who play on the PP. (0.5 points per shot on goal).

SA16 might have some knowledge on this, but I'm not sure if he dabbles in DFS hockey.

I've never played daily hockey fantasy, but I'd imagine that any edge is going to be very hard to come by given that goals are random events right?

A big problem I have with predictive analytics in these really complex and nuanced sports is that there is no modeling dumb luck. I understand that luck evens out over the long term, but how does that help in DFS?

 

[SA16]

Man, if you find out, let me know. I've tried all sorts of things on this and lost $ testing them

With DFS, at least on DraftKings, I don't think there's anything more important that playing players who shoot a lot, and who play on the PP. (0.5 points per shot on goal).

SA16 might have some knowledge on this, but I'm not sure if he dabbles in DFS hockey.

Oh hockey is my number 1 DFS sport. I'm probably top 5-10 in NHL volume on DK at least in cash games. I'm an NHL/MLB/NFL guy. The NBA is bad and should feel bad.

The essence of it is pretty much what you said. Forwards who shoot a lot and play on the PP. Play the elite D. And cheap D who get PK time and block a ton of shots while also shooting (ex. de Haan/Hamonic/Gudas/Matt Irwin if an injury are some of the best values). And most importantly guys who get a lot of TOI or you expect to get more due to a change in role (namely Irwin went from like a 12 min/game guy to 20+ when either Subban or Josi were out).

Honestly most of the other stuff is noise and you really don't need any of the complex stuff.

I've never played daily hockey fantasy, but I'd imagine that any edge is going to be very hard to come by given that goals are random events right?

A big problem I have with predictive analytics in these really complex and nuanced sports is that there is no modeling dumb luck. I understand that luck evens out over the long term, but how does that help in DFS?

Goals are often random that's why you focus on the guys who are on the ice a lot and do stuff (shoot/block) and that helps build a nice basis of points (ex. Burns like always get you 5 shots+blocks, erik johnson around 4, Ovy is usually good for 4+ shots a game) so when they don't score you still get points. You absolutely never play guys like Joe Thornton, Henrik, Sedin, Niklas Kronwall since their range of scores are basically in 2's since they get assists and do absolutely nothing else.

There is a lot of luck on a daily basis. That's why you try to get a portfolio of a lot of different games (I get as many H2H's as I can against as many different opponents I can), have a conservative bankroll management strategy, and play on as many days as possible to help mitigate variance and realize your long run expectation. DFS NHL is very simple and people hugely overcomplicate it.

 

[Kane One]

Can we at least keep adding definitions to the OP? Sort of like that huge list in the Visiting NYC thread.

 

[silverfish]

Oh hockey is my number 1 DFS sport. I'm probably top 5-10 in NHL volume on DK at least in cash games. I'm an NHL/MLB/NFL guy. The NBA is bad and should feel bad.

The essence of it is pretty much what you said. Forwards who shoot a lot and play on the PP. Play the elite D. And cheap D who get PK time and block a ton of shots while also shooting (ex. de Haan/Hamonic/Gudas/Matt Irwin if an injury are some of the best values). And most importantly guys who get a lot of TOI or you expect to get more due to a change in role (namely Irwin went from like a 12 min/game guy to 20+ when either Subban or Josi were out).

Honestly most of the other stuff is noise and you really don't need any of the complex stuff.



Goals are often random that's why you focus on the guys who are on the ice a lot and do stuff (shoot/block) and that helps build a nice basis of points (ex. Burns like always get you 5 shots+blocks, erik johnson around 4, Ovy is usually good for 4+ shots a game) so when they don't score you still get points. You absolutely never play guys like Joe Thornton, Henrik, Sedin, Niklas Kronwall since their range of scores are basically in 2's since they get assists and do absolutely nothing else.

There is a lot of luck on a daily basis. That's why you try to get a portfolio of a lot of different games (I get as many H2H's as I can against as many different opponents I can), have a conservative bankroll management strategy, and play on as many days as possible to help mitigate variance and realize your long run expectation. DFS NHL is very simple and people hugely overcomplicate it.

This is all very interesting and I'll be referring back to this post in a few weeks

No stats for tonight's game, btw. AHL arenas

 

[SA16]

PM if you have any questions during the year I don't think this is supposed to become the gamblers thread.

Lots of preseason games haven't had stats (and as an aside this is why DK/FD don't run preseason NHL contests...FD tried t last year and had to cancel them when they realized there were no stats).

 

[silverfish]

Rangers @ Devils, 9/23

 

[Machinehead]

McDonagh-Shatt great.

Staal a disease.

"What's AV Going to Do?!" would be an awesome gameshow.

Howie Mandel: "Who's going to be on the ice late in a close game???? We'll find out after these messages."

(but it's Staal tho)

 

[silverfish]

Desjardins was brought here on the suggestion of Brad Richards, right? Feel like I heard that during preseason game 1.

Nothing like hiring former players. These guys really think the game at a top level. Totally aren't biased towards their friends or guys they've played with before. No, no no.

Get Desjardins the **** outta here.

 

[ManUtdTobbe]

Yeah, not a great start for Richards here. Players need to wipe their biases clear when they step into a management role or they will be awful.

 

[Levitate]

Eh he's a PTO and a camp body. I'll be concerned if they give him a real contract (I doubt they will considering what other guys have shown).

 

[silverfish]

A bunch of post got deleted and everything is different and I'm scared.

 

[Mac n Gs]

Andrew Berkshire's top center article is *Italian chef kissing fingers*

Oh, also, Derek Stepan is 23rd when factoring in offense, transition, and defense, so f*** off haters.

 

[Revel]

A bunch of post got deleted and everything is different and I'm scared.

I'm a little creeped out myself.

 

[nyr2k2]

Me too, man, me too. This is way more scary than when HF switched from whatever to old HF Rivals ran on to vBulletin.

 

[Revel]

I had literally hundreds of threads on ignore, and everything neatly organized. Now nothing is on ignore. What a mess. This is going to take days to clean up.

 

[SA16]

I see we've got the typical "upgrade" that makes everything HUGE and requires insane amount of scrolling to see the same amount of info we had before.