GDT: Devils @ Sabres 1/30/18 7PM

[NJDfan1711]

5 minutes ago, mfitz804 said:

I thought my post was borderline obvious, so much so I almost didn't post it. Apparently not lol. 

Seriously.  So did I lol

[SterioDesign]

7 minutes ago, Triumph said:

Yes, great, the 'I'm not racist, but' of hockey statistics.

I'm not even against hockey statistics. I'm just saying and will always say that they do not tell the whole story. Like you said they can show tendencies and confirm what you're seeing from the eye test in order to analyse some situations but way too much stuff that can't be translated in numbers is happening in hockey. Hockey is not baseball.

 

 

Edited January 31, 2018 by SterioDesign

['7']

You have to strike a balance between where to draw the line on advanced stats, and not lose the human/personal/chemistry touch of the game. if you're strictly mathematical and base everything on it then you could lose the team, you'll compile players that should be doing X but not taking into account certain intangible toxic elements, all of a sudden you have mayhem in the locker room, lack of motivation, cancerous players.

Sometimes it may actually be worth carrying a guy who's perhaps not a statistical darling because he keeps the team together, he's like another coach, he keeps at risk guys on the straight and narrow

 

Edited January 31, 2018 by '7'

[SterioDesign]

9 minutes ago, Triumph said:

Yes, great, the 'I'm not racist, but' of hockey statistics.  You'll forgive me for thinking that those 6 words don't mean a whole lot when the rest of the post is running down people who use numbers.

Also there's a difference with "people who use numbers" and what i said which was "people obsessed with numbers". 

[mfitz804]

11 minutes ago, Triumph said:

Yes, great, the 'I'm not racist, but' of hockey statistics.  

Dude, some of my best friends like hockey statistics!!

[Triumph]

3 minutes ago, SterioDesign said:

I'm not even against hockey statistics. I'm just saying and will always say that they do not tell the whole story. Like you said they can show tendencies and confirm what you're seeing from the eye test in order to analyse some situations but way too much stuff that can't be translated in numbers is happening in hockey. Hockey is not baseball.

 

 

You just want to have the same arguments again and again, don't you?  Absolutely no one said they told the whole story - they don't even tell the whole story in baseball.  But once again, I will remind you that 15 years ago, everyone thought baseball wasn't baseball.  That you couldn't measure all this stuff.  That GMs knew best because they played the game.  And guess what - now almost no GMs in baseball are former baseball players.  But talking to those GMs, who did not play the game, they will all tell you - there is still a HUGE need for scouts and eye test and all of that stuff - but you've got to have the numbers helping guide those eyes to where they should be, and vice versa.  

[SterioDesign]

1 minute ago, Triumph said:

You just want to have the same arguments again and again, don't you?  Absolutely no one said they told the whole story - they don't even tell the whole story in baseball.  But once again, I will remind you that 15 years ago, everyone thought baseball wasn't baseball.  That you couldn't measure all this stuff.  That GMs knew best because they played the game.  And guess what - now almost no GMs in baseball are former baseball players.  But talking to those GMs, who did not play the game, they will all tell you - there is still a HUGE need for scouts and eye test and all of that stuff - but you've got to have the numbers helping guide those eyes to where they should be, and vice versa.  

 

Well exactly you need both lol Again how many time am i supposed to say that i was talking about the people OBSESSED with stats that simply cannot analyse a game without looking at numbers or judge a player 100% based on his numbers.

and baseball will never compare to hockey as there's way way way less factors going in a play and affecting numbers.

[Triumph]

10 minutes ago, SterioDesign said:

Well exactly you need both lol Again how many time am i supposed to say that i was talking about the people OBSESSED with stats that simply cannot analyse a game without looking at numbers or judge a player 100% based on his numbers.

and baseball will never compare to hockey as there's way way way less factors going in a play and affecting numbers.

You're fighting a strawman.  We both agree that such a person, if he or she even exists, is not worth listening to.  But I am always going to look at the numbers after a game, just to see if what I was seeing matched up with what was actually going on.

[Steven M.]

There is obviously merit to both, but more often than not, the numbers usually correlate to the play. I saw a post on here trashing Mike Brown when he put a chart up of our expected goals for throughout the entire season. Now, I know everyone likes to trash Mike Brown, but the chart is a legitimate one. However, the chart showed that we were playing below the median for the first half of the season more often than not, and since around the time of the Vatanen trade, we have been playing above it. Which likely correlates to how many shots we used to give up versus how much we now give up. Everyone proceeded to trash him saying it didn't mean sh!t because in the beginning of the season we were winning and now we're not. To be honest, that doesn't tell the whole story and reeks of 2013-14 when Schneider was obviously far and away the superior goalie but there was still the crowd that wanted to ride Marty for the entire season because for some strange reason they'd win with him. When you look at the rest of that chart and where the teams were it often accurately depicted each team's success. 

As for further anecdotal evidence: last year, Wood was getting caved in every shift like an AHLer. We all saw that with our eyes, and the numbers backed it up. This year, he's playing much better, and the numbers back it up. Zacha: playing much better of late by the eye test despite not burying it. The numbers back it up.

Counterpoint: sometimes when you only rely on the numbers, it doesn't necessarily back anything up. The numbers said that Beau Bennett belonged in our top six every night. For whatever reason the coaching staff disagreed. When he'd get his opportunities, he wouldn't seize on them. Upon leaving our organization, he ends up going to St. Louis and can't stick there either. I think the point is that the numbers OFTEN back up the eye test. So should you rely on them completely like MB does? Probably not. Should you (for the most part) disregard them like SD does? Probably not. The happy medium lies somewhere in between. I think it's pretty obvious how the league sees it though. More often than not the good general managers are embracing analytics. I don't think a GM that pays attention to this stuff trades Hall for a Larsson. I don't think a GM paying attention to this stuff trades a 3rd round pick for Zac Rinaldo. 

Just my opinion!

[CommonDreads]

3 hours ago, Martyisth3b3st said:

Maybe someone with fancy stats can weigh in here, but from the eye-test I'd say he's been one of the 3 or 4 best forwards on the team night-in and night-out ever since he got put on a line with Palms and Wood. He had multiple high-danger chances again last night. Eventually some of these pucks are going to go in the back of the net, but I thought Zacha looked like the 3rd best forward on the team last night.

And if you compare last night with the first few weeks of the season where he looked unbelievably outmatched and overwhelmed, I'd say "absolutely flourishing" is an apt description of his play. 

Obviously some games are better than others but he's been generally good, he had a valley against Nashville with a 31.58% Corsi (I will also point out in that game just to illustrate bad games can happen even to fan favorites, Jesper Bratt sat at a 26.09 in that game) but Zacha also had a ridiculously good possession game against Boston with an 85.71% Corsi. Ups and downs happen for a kid his age, so it's nice to see him have games like he did against Boston where he can dominate the pace of a game.

[SterioDesign]

8 minutes ago, Steven M. said:

There is obviously merit to both, but more often than not, the numbers usually correlate to the play. I saw a post on here trashing Mike Brown when he put a chart up of our expected goals for throughout the entire season. Now, I know everyone likes to trash Mike Brown, but the chart is a legitimate one. However, the chart showed that we were playing below the median for the first half of the season more often than not, and since around the time of the Vatanen trade, we have been playing above it. Which likely correlates to how many shots we used to give up versus how much we now give up. Everyone proceeded to trash him saying it didn't mean sh!t because in the beginning of the season we were winning and now we're not. To be honest, that doesn't tell the whole story and reeks of 2013-14 when Schneider was obviously far and away the superior goalie but there was still the crowd that wanted to ride Marty for the entire season because for some strange reason they'd win with him. When you look at the rest of that chart and where the teams were it often accurately depicted each team's success. 

As for further anecdotal evidence: last year, Wood was getting caved in every shift like an AHLer. We all saw that with our eyes, and the numbers backed it up. This year, he's playing much better, and the numbers back it up. Zacha: playing much better of late by the eye test despite not burying it. The numbers back it up.

Counterpoint: sometimes when you only rely on the numbers, it doesn't necessarily back anything up. The numbers said that Beau Bennett belonged in our top six every night. For whatever reason the coaching staff disagreed. When he'd get his opportunities, he wouldn't seize on them. Upon leaving our organization, he ends up going to St. Louis and can't stick there either. I think the point is that the numbers OFTEN back up the eye test. So should you rely on them completely like MB does? Probably not. Should you (for the most part) disregard them like SD does? Probably not. The happy medium lies somewhere in between. I think it's pretty obvious how the league sees it though. More often than not the good general managers are embracing analytics. I don't think a GM that pays attention to this stuff trades Hall for a Larsson. I don't think a GM paying attention to this stuff trades a 3rd round pick for Zac Rinaldo. 

Just my opinion!

Well exactly you gotta mix both or put the stats in context. And that's the main thing, numbers don't often give you a context and it's incredibly important in order to really understand the numbers

[DevsMan84]

I just find it funny when anytime the results does not match the evidence/observations the numbers crowd always blames it on either good or bad luck.

[CommonDreads]

9 minutes ago, DevsMan84 said:

I just find it funny when anytime the results does not match the evidence/observations the numbers crowd always blames it on either good or bad luck.

I don't really consider myself a numbers guy, rather I just find them a useful tool to understand better what's happening on the ice but I'll trust my eyes first, but good vs bad luck I do find to be a valid card to play sometimes. Hockey simply has so many more variables than other sports, so I don't think we'll ever see a true moneyball implementation in hockey, and there is validity to numbers not lining up or being skewed because of a few good or bad breaks here and there.

[SterioDesign]

11 minutes ago, DevsMan84 said:

I just find it funny when anytime the results does not match the evidence/observations the numbers crowd always blames it on either good or bad luck.

i was juuuuuuust about to bring that up haha

Yeah when the numbers are not giving you the answer you want then just say either "it's good or bad luck" or "those numbers are unsustainable" 

[DevsMan84]

4 minutes ago, CommonDreads said:

I don't really consider myself a numbers guy, rather I just find them a useful tool to understand better what's happening on the ice but I'll trust my eyes first, but good vs bad luck I do find to be a valid card to play sometimes. Hockey simply has so many more variables than other sports, so I don't think we'll ever see a true moneyball implementation in hockey, and there is validity to numbers not lining up or being skewed because of a few good or bad breaks here and there.

Then what's the use of deep-diving on numbers when there are so many variables in hockey versus say baseball?

I also see the word luck thrown around by the hardcore numbers people to take said numbers all that seriously.

[CommonDreads]

1 minute ago, DevsMan84 said:

Then what's the use of deep-diving on numbers when there are so many variables in hockey versus say baseball?

I also see the word luck thrown around by the hardcore numbers people to take said numbers all that seriously.

For me it's a good tool in terms of player comparison over a large sample size, watching a game you can get a general feel of your thoughts on a player but numbers are a useful tool to see side by side comparisons of a player's impact on the ice beyond goals, assists, points, etc. It's a quantifiable thing you can point to and say, Player X has a more positive impact on possession than Player Y based on Z, which i think is useful in player analysis.

Again, not a means to an end, but rather a tool. I don't believe you can build a team just on a spreadsheet, but you can use numbers to further understand what you're watching.

[MadDog2020]

Here's one for the 'Vatanen sucks' crowd: https://mobile.twitter.com/DevilsInsiders/status/958793426446807040?ref_src=twsrc%5Etfw&ref_url=https%3A%2F%2Fs9e.github.io%2Fiframe%2Ftwitter.min.html%23958793426446807040


Sent from my iPhone using Tapatalk

[DevsMan84]

7 minutes ago, CommonDreads said:

For me it's a good tool in terms of player comparison over a large sample size, watching a game you can get a general feel of your thoughts on a player but numbers are a useful tool to see side by side comparisons of a player's impact on the ice beyond goals, assists, points, etc. It's a quantifiable thing you can point to and say, Player X has a more positive impact on possession than Player Y based on Z, which i think is useful in player analysis.

Again, not a means to an end, but rather a tool. I don't believe you can build a team just on a spreadsheet, but you can use numbers to further understand what you're watching.

The bolded part largely makes sense.

However how many times have we discussed/argued back and forth on here about why the data is showing player x having a positive impact yet the end-result is negative?

Oh yeah I forgot, it's just "bad luck"

[Triumph]

8 minutes ago, CommonDreads said:

For me it's a good tool in terms of player comparison over a large sample size, watching a game you can get a general feel of your thoughts on a player but numbers are a useful tool to see side by side comparisons of a player's impact on the ice beyond goals, assists, points, etc. It's a quantifiable thing you can point to and say, Player X has a more positive impact on possession than Player Y based on Z, which i think is useful in player analysis.

Again, not a means to an end, but rather a tool. I don't believe you can build a team just on a spreadsheet, but you can use numbers to further understand what you're watching.

The most critical thing to take away is that it's within a tiny framework.  The best team this season has 71% of the possible points, the worst has 33%.  And neither is probably as good or as bad as their record.  The edges in hockey are quite small.  That's why luck gets talked about so much - it gets talked about in baseball a lot too because the edges there are also tiny.  The best team isn't expected to beat the worst more than 3 times in 4.  All a team can do in hockey is try to work on their tiny edges while hoping not to be overwhelmed by the luck monster.

[Triumph]

4 minutes ago, DevsMan84 said:

The bolded part largely makes sense.

However how many times have we discussed/argued back and forth on here about why the data is showing player x having a positive impact yet the end-result is negative?

Oh yeah I forgot, it's just "bad luck"

It's very hard to make a determination because luck overwhelms skill in the short term.  You often need many seasons worth of data on a player to determine around where their 'true' shooting percentage is.  Many people want to conflate different types of luck (i.e. the fact that a given player can't score doesn't say anything about whether or not the players on his line can).  Plus the Devils have missed the playoffs for the last 5 seasons so most of the talk is from the negative end.

[EdgeControl]

On 1/30/2018 at 9:37 PM, Nicomo said:

Jokes aside, Kinkaid played well. Hall with a goal his first game back, too. Can’t ask for much more. 

Nicomo

 

get a philly game thread going.. you got us a win out of the break

[NJDfan1711]

41 minutes ago, EdgeControl said:

Nicomo

 

get a philly game thread going.. you got us a win out of the break

Second'd.

[Marshall]

As someone who has worked (and continue to do so) professionally in hockey with statistical analysis as one of the main tools...this thread sure is wild.

[NJDevils1214]

2 hours ago, Marshall said:

As someone who has worked (and continue to do so) professionally in hockey with statistical analysis as one of the main tools...this thread sure is wild.

Can I get in on this profession? Economics grad here.