Kovi

[SterioDesign]

Stats don't mean anything if they are used with small sample sizes. That's why the "shift of your life" example is invalid

 

I know that and that's my point. Every play is different and there's too many variables and luck that even with a bigger sample it's not giving you the whole story.

[Triumph]

Sterio:  This is the last time I'll be explaining my position on statistics and how they help my understanding of the game to you.  If you continue to bait me, I'll be adding you to my ignore list.  I don't want to have to do that, because I do think you're very curious about why teams win games and you have insight about the games you're watching.

 

I don't like using shots either.  I wish I could use goals only.  But hockey seasons aren't long enough to use goals - we've seen that goal differential is much less predictive of future results than Fenwick % when a game is close.  Teams get hot for half a season with great goaltending and shooting luck.  We're seeing it right now with Anaheim.  But there's too much randomness involved in scoring a goal at the NHL level to just consider goals without making gigantic errors in player and team evaluation.  

 

Whatever you're seeing in hockey and soccer, that's at a much lower level than the NHL - the reason why I think Fenwick % is valid at the NHL level is because players as unique as you're describing at your level tend not to make it in the NHL.  Yeah, there are guys like Jason Blake who probably shoot too much and guys like Sergei Kostitsyn who probably pass too much.  These players have to be accounted for, certainly.  It's not perfect.  But if I showed you the Fenwick tied numbers from the last 5 years of the NHL for each team, I'd be showing you pretty much the standings in the NHL. There's some imperfections because of goaltending differences, and certainly those have to be accounted for. Teammates and game situation also have to be accounted for - we know that teams who trail outshoot teams who are ahead.

 

Corsi/Fenwick isn't perfect.  I wish the NHL counted scoring chances.  I wish we had more data about puck touches in the defensive zone, dump-ins versus carrying the puck into the offensive zone, etc.  But we're learning a lot of things about how team s win hockey games, and winning teams control the game by having the puck more than the other team.  That translates into a higher Corsi/Fenwick.  Most of the game is played 5 on 5 - it's real hard to win just through special teams and goaltending.

 

Anyway this was a great article about the subject published today:  http://www.mc79hockey.com/?p=5701 

Edited March 26, 2013 by Triumph

[Derlique]

I know that and that's my point. Every play is different and there's too many variables and luck that even with a bigger sample it's not giving you the whole story.

I'm on my phone, so I may not do the best job explaining this. According the advanced statistics, the Ducks are overachieving big time. They have a negative shots for: shots against ratio and all if their top 6 are shooting a ridiculous %. Now, they are on a three game losing streak. Shooting % is certainly a good stat. Loktianiov and Henrique will definitely come back to earth because no one in the history of the league has maintained such a % throughout their career. I think of anyone who has played over 500 games, Tanguay has the highest and he's at like 18%.

[SterioDesign]

Thats a good article Triumph and Smantzas you're right.

 

I think at this point it's now clear that we all know that stats can demonstrate stuff / be misleading and everything So my question... why do some people keep on bringing it on every single arguments? im serious.

 

seems like any regular talk that goes like this like...

 

"yeah loktionov has been pretty good lately he's really impressive and he's producing"

 

we're 95% sure of getting a reply with "loktionov is currently shooting "some %" and it's not likely that he'll keep putting up those numbers"

 

"Some team is going well lately they are 8-1-3"

 

we'll have a reply... "that team is shooting some % now and are running on luck and good goaltending" 

 

it's like... 75% of your posts needs a shooting % or any kind of stats thrown in it, just for the sake of it. While it's clear that on a small sample (and for most talk it is) it can be misleading, so whats the point?

[ATLL765]

Sterio:  This is the last time I'll be explaining my position on statistics and how they help my understanding of the game to you.  If you continue to bait me, I'll be adding you to my ignore list.  I don't want to have to do that, because I do think you're very curious about why teams win games and you have insight about the games you're watching.

 

I don't like using shots either.  I wish I could use goals only.  But hockey seasons aren't long enough to use goals - we've seen that goal differential is much less predictive of future results than Fenwick % when a game is close.  Teams get hot for half a season with great goaltending and shooting luck.  We're seeing it right now with Anaheim.  But there's too much randomness involved in scoring a goal at the NHL level to just consider goals without making gigantic errors in player and team evaluation.  

 

Whatever you're seeing in hockey and soccer, that's at a much lower level than the NHL - the reason why I think Fenwick % is valid at the NHL level is because players as unique as you're describing at your level tend not to make it in the NHL.  Yeah, there are guys like Jason Blake who probably shoot too much and guys like Sergei Kostitsyn who probably pass too much.  These players have to be accounted for, certainly.  It's not perfect.  But if I showed you the Fenwick tied numbers from the last 5 years of the NHL for each team, I'd be showing you pretty much the standings in the NHL. There's some imperfections because of goaltending differences, and certainly those have to be accounted for. Teammates and game situation also have to be accounted for - we know that teams who trail outshoot teams who are ahead.

 

Corsi/Fenwick isn't perfect.  I wish the NHL counted scoring chances.  I wish we had more data about puck touches in the defensive zone, dump-ins versus carrying the puck into the offensive zone, etc.  But we're learning a lot of things about how team s win hockey games, and winning teams control the game by having the puck more than the other team.  That translates into a higher Corsi/Fenwick.  Most of the game is played 5 on 5 - it's real hard to win just through special teams and goaltending.

 

Anyway this was a great article about the subject published today:  http://www.mc79hockey.com/?p=5701 

You clearly didn't get it either. You bait others just as much with you tone. You make others feel as if once they disagree with you, you feel as if we are below you. So either stop being all snobby about your stats or stfu about it.

 

And again, I believe your stats in many cases do show trends that are otherwise lost through the eye test which has limited memory. I just think you don't need to be be so condescending every time you talk about it.

Edited March 26, 2013 by ATLL765

[Triumph]

You clearly didn't get it either. You bait others just as much with you tone. You make others feel as if once they disagree with you, you feel as if we are below you. So either stop being all snobby about your stats or stfu about it.

 

And again, I believe your stats in many cases do show trends that are otherwise lost through the eye test which has limited memory. I just think you don't need to be be so condescending every time you talk about it.

 

Feel free to ignore me if you like, it's not difficult to do.  Sterio gives as good as he gets, he's called me out either actually or by implication in multiple threads.  If he continues I'll ignore him and the board will likely be better for it, but I hope he doesn't.

 

Sterio:  I don't control what other people post - I don't think I've posted anything about Loktionov's shooting percentage.  It's more about managing expectations, I guess.  Loktionov is real good but he's not likely to be a 30 goal scorer anytime soon, that's all.

Edited March 26, 2013 by Triumph

[Pepperkorn]

Stats are cool because as the article said

honestly, a lot of the time, the data says things that hockey people would believe anyway, with the added benefit of providing some scale as to how significant some phenomenon or tactic is.

 

I was initially turned off to stats because someone with ZERO intuitive power had grabbed on to stats and often attacked me for my intuitive proclamations.  Time has proven me correct . 

 

Tri is different - yeah he wants to be right - we all do, the Virgos anyhow, and he has slowly gone to stats to justify his proclamations.  Maybe because I've read his stuff for about a decade now, I see his stats stuff not know-it-all as much as testing the waters - he gets bolder and bolder as stats qualify his intuition.  

 

I don't see why you need to be irritated with the little stats tossed in.  Consider it your own pet peeve not some irritating sh!t some Devils fans does.  It's not tri's job to make you happy, you know?  What he's saying is harmless - maybe even meaningless.  Relax

 

[edit - you know... or how Tri said it for himself ]

Edited March 26, 2013 by Pepperkorn

[CarpathianForest]

This really shouldn't be a gang up on Triumph thread. Everyone know he and I don't get along well. However, his stat analysis is valuable and he contributes to the forum. I learned to just get past his "Holier than Thou" attitude and value his analysis since he typically has something important to add.

[EdgeControl]

"I wanted to create a sense of the scope of the differences between good and bad X liners, so I took their GD/60 and multiplied it by the average ice time for that line and then by 82 games to convert things into goal difference over the course of a season"

 

goal differential????   Dam, this sure sounds a bit like +/- to me.  an indicator between good and bad X liners???

[Pepperkorn]

but +/- doesn't say it all.  Supplementary data - not redundant.

[1Guy3Cups]

I am very good at math.. just throwing it out there.. 

[Zubie#8]

Tri contributes a lot to this forum, instead of the usual whining moaning and complaining like others.

[EdgeControl]

but +/- doesn't say it all.  Supplementary data - not redundant.

its goal differential , and thats what the author was using as a criteria   ( how can that possibly be???? lol )  btw all those "good" players mentioned are also top +/-.players      theres a correlation  between all these numbers

[Triumph]

its goal differential , and thats what the author was using as a criteria   ( how can that possibly be???? lol )  btw all those "good" players mentioned are also top +/-.players      theres a correlation  between all these numbers

 

The criteria he used to judge the players was ice time.  Then he looked at things like goal differential.  Taken in the aggregate, or over a long career, yeah, +/- is meaningful.  But it has all sorts of garbage like empty net goals and short handed goals for and against in it, and it doesn't take into account goaltending.  That's why we have things like Fenwick that take out the goaltender.

 

I agree there's a correlation between these numbers - players with higher +/-s tend to be better players - but just like you can't say 'that player's Fenwick is good thus he's good', you can't say 'that player's +/- is good, therefore he's good'.

[EdgeControl]

The criteria he used to judge the players was ice time.  Then he looked at things like goal differential.  Taken in the aggregate, or over a long career, yeah, +/- is meaningful.  But it has all sorts of garbage like empty net goals and short handed goals for and against in it, and it doesn't take into account goaltending.  That's why we have things like Fenwick that take out the goaltender.

 

I agree there's a correlation between these numbers - players with higher +/-s tend to be better players - but just like you can't say 'that player's Fenwick is good thus he's good', you can't say 'that player's +/- is good, therefore he's good'.

yup, loved the marchand explaination that was good! lol

[Marshall]

If you're doubtful about underlying numbers/fancy stats/adv stats...Mark Fraser is a good example. We all know him, we all know what his playing is like and we all know he's not a very good defender. But he lead the league in +/- for a few weeks and he's still in the top 20. Mark Fraser hasn't gone and become a great defenseman all of a sudden, he is after all 8th in ice time among defenseman on that Toronto team. But the underlying numbers show why (absurd shooting percentage and SV%...which is falling now), along with your eyes, whilst the traditional numbers don't.

Edited March 27, 2013 by Marshall

[SterioDesign]

Feel free to ignore me if you like, it's not difficult to do.  Sterio gives as good as he gets, he's called me out either actually or by implication in multiple threads.  If he continues I'll ignore him and the board will likely be better for it, but I hope he doesn't.

 

Sterio:  I don't control what other people post - I don't think I've posted anything about Loktionov's shooting percentage.  It's more about managing expectations, I guess.  Loktionov is real good but he's not likely to be a 30 goal scorer anytime soon, that's all.

 

Loktionov was purely a made up example. It's about the way you try to use stats arrogantly in every single post you make.

 

Feel free to ignore me if you like, it's not difficult to do.  Sterio gives as good as he gets, he's called me out either actually or by implication in multiple threads.  If he continues I'll ignore him and the board will likely be better for it, but I hope he doesn't.

 

Sterio:  I don't control what other people post - I don't think I've posted anything about Loktionov's shooting percentage.  It's more about managing expectations, I guess.  Loktionov is real good but he's not likely to be a 30 goal scorer anytime soon, that's all.

It's not about what other people are posting. Personally in the last few months i've been a downer on the forum and i pissed off a few people, and some called me out on it. I obviously don't like to be seen as a downer/whinner im not like that usually. And it hits home, so then lately i'm working on it and trying not to sound like a downer all the time so that everyone is happy and don't see me as a negative poster. 

 

Now how many times did someone called you out for acting superior and trying to make people feel like they are dumb when they don't agree with what you think with that know-it-all attitude? It's not about what you said it's about the way you say things. And you've been called out for it by MANY PEOPLE, MANY TIMES. Everyone appreciate what you bring to the forum in term of analysis but fair to say most hate the way you bring it to the table most of the time.

 

Never crossed your mind to work on that? like.. bringing the same arguments without sounding like an arrogant mental bully? 

Edited March 27, 2013 by SterioDesign

[Pepperkorn]

That's exactly why you need supplemental data. No one says stats are stand-alone, be all end all. Marshall you're sort of making the whole case for corsi/fenwicks It also kind of seems like you ignored Tri's follow up post. I'd be hard pressed not to respond "Duuude -- didn't you even read what I just wrote?" I'd call you stupid because I'd know full well you didn't even look at what I wrote. Not because you were dumb but because I was frustrated with how you call me a jackass (maybe not in so many words- or few as the case may be) without even considering what I wrote.

 

My stats fear, is based on a worry that their use will change the game - take out the heart and soul. but that's just the ludite in me. People need to make peace with the fact that numbers are not soulless. Mathematics is the language of God. Art = math.

 

So many comments are just not getting it - not even attempting to get it. How can a guy trying to teach the value of stats even respond when no one honestly wants to understand - they just want to be offended. You guys aren't too stupid or lazy to understand -- no, it looks like you just want to be offended and launch a counter attack without even attempting to understand what got you upset to begin with. I'd call you stupid to try to get you to tell me what you know. Like calling a guy a nancy to get him to fight. Just give it a shot - try to understand what stats people tell you with their stats - read what people write rather than just calling it superior mumbo jumbo, you know?

Edited March 27, 2013 by Pepperkorn

[DevsMan84]

The way I see it, there two types of sports people:

 

Those who watch the game and use stats to fill in the blanks.  Then there are those to bury their heads in the stat sheets and watch the game a little to fill in the blanks.  I consider myself to be in the first group.

 

Stats are important and I do not think either side will deny that, but it is a folly to think that the stats show everything just as much as totally ignoring them.  However, I also feel the people who tend to the be in the second group have bloated but tender egos as I have encountered the same type of superiority complex whether it is for baseball, football, basketball or hockey.

 

Just my 2 cents.

[SterioDesign]

The way I see it, there two types of sports people:

 

Those who watch the game and use stats to fill in the blanks.  Then there are those to bury their heads in the stat sheets and watch the game a little to fill in the blanks.  I consider myself to be in the first group.

 

Stats are important and I do not think either side will deny that, but it is a folly to think that the stats show everything just as much as totally ignoring them.  However, I also feel the people who tend to the be in the second group have bloated but tender egos as I have encountered the same type of superiority complex whether it is for baseball, football, basketball or hockey.

 

Just my 2 cents.

 

agreed with everything you said, i've been saying that some stats are bullsh!t and useless lately, i'd re-phrase that saying that i'm simply against the way some are using them.

 

i'd also add that i'm strongly believing that a good amount of the 2nd group as not played a lot of the sport they are following and are missing that "experience" of some aspect of the game and are ignoring or downplaying the importance of intangibles. When you played sports you know how important stuff like that is over the course of a season. You can't see that through stats.

[DH26]

agreed with everything you said, i've been saying that some stats are bullsh!t and useless lately, i'd re-phrase that saying that i'm simply against the way some are using them.

 

i'd also add that i'm strongly believing that a good amount of the 2nd group as not played a lot of the sport they are following and are missing that "experience" of some aspect of the game and are ignoring or downplaying the importance of intangibles. When you played sports you know how important stuff like that is over the course of a season. You can't see that through stats.

 

No matter what hockey experience you have, it's exactly the same amount in the NHL as everyone else on this board and most of the time I've seen people play that card it's out of desperation 

[SterioDesign]

No matter what hockey experience you have, it's exactly the same amount in the NHL as everyone else on this board and most of the time I've seen people play that card it's out of desperation 

 

On my part, it's really really not out of desperation. NHL is simply a higher level, players are dealing with the same stuff and they are doing the same thing than any other sports. If anything it's even more important at that level cause it's easier to slide and lose yourself in that circus sometimes.

 

I've played with guys who we're not putting up points but they we're bringing so much, just the way they we're willing to suffer for the team, or the way they'd help the young guys giving them tips and pointers in any situations, or just the way they approach the game or how they play every shift giving everything they have. It really goes a long way. What Langs did for Zach and what Zach did for Henrique goes a long way and what Elias as been doing for years now.

 

But all that stuff is mostly stuff that the fans don't see or simply hear about it in interviews and stuff, but you need that to be successful. There's a reason Lou trust his vets so much, cause they are PACKED with intangibles. You also don't see too often a GREAT players simply based on their skills, great players are skills + some kind of intangibles... Toews, Iginla, Crosby, etc

[ATLL765]

This really shouldn't be a gang up on Triumph thread. Everyone know he and I don't get along well. However, his stat analysis is valuable and he contributes to the forum. I learned to just get past his "Holier than Thou" attitude and value his analysis since he typically has something important to add.

I agree, that was my whole point. I mentioned Sterio was giving it too, but Tri can just be so condescending and that bothers me is all. I would honestly like to learn more about the stats and Tri is someone who clearly understands how the stats work, but anyone else like me is gonna be turned off when they disagree with Tri and he makes them feel like he thinks they're idiots for having disagreed, rather than trying to use the lack of understanding as teaching tool. That's my thoughts.

[Colorado Rockies 1976]

I agree, that was my whole point. I mentioned Sterio was giving it too, but Tri can just be so condescending and that bothers me is all. I would honestly like to learn more about the stats and Tri is someone who clearly understands how the stats work, but anyone else like me is gonna be turned off when they disagree with Tri and he makes them feel like he thinks they're idiots for having disagreed, rather than trying to use the lack of understanding as teaching tool. That's my thoughts.

 

Don't want to pile on Tri, but you get what you get with him:  his presentation of less commonly-known methods of measuring hockey players (you may not agree or only partially agree with those methods, but at the very least, they spark debate and discussion), heartfelt and passionate insights, and occasional lack of bedside manners that unfortunately have a way of taking the focus where it should lie (on the point he's presenting), and instead turning it towards HOW he presented his point, or chose to counter someone else's point.  I think most posters here have learned to accept that part of Tri's posting manner, but I think there are other members here (at times I've been one of them) who wishes the more abrasive ways Tri sometimes chooses to post could be softened (preferably by Tri himself).  That being said, I don't think he's going to change, so as far as that part of it goes, if you get into a debate with Tri and he doesn't agree with your points or counterpoints, and you're in a sensitive frame of mind during the exchange, you're probably going ot be put off on occasion by what he has to say, or more specifically, how he says it.  I've had good and both frustrating dialogue with Tri over the years, and there have been times where I've passed on debating with him because I had a pretty good idea of how he might respond. 

 

All this being said, would the board be missing a lot if he wasn't a participant?  Absolutely.     

Edited March 27, 2013 by Colorado Rockies 1976

[Triumph]

I agree, that was my whole point. I mentioned Sterio was giving it too, but Tri can just be so condescending and that bothers me is all. I would honestly like to learn more about the stats and Tri is someone who clearly understands how the stats work, but anyone else like me is gonna be turned off when they disagree with Tri and he makes them feel like he thinks they're idiots for having disagreed, rather than trying to use the lack of understanding as teaching tool. That's my thoughts.

 

This is fair - I tend to get upset when people pick fights (as happened here).  Typically if someone's open to ideas, I will not do this, but I'm not perfect in this regard.  Part of it is also that I've had ongoing discussions about these matters for years and that while the arguments are new to some, they're old to me and many of the people here, and I tire of explaining myself as though I'm saying these things for the first time.  Devils731 is someone who's much better at this.