U20 tournament - Junior Evaluation Camp (Sieloff, Gillies, Monahan, Poirier & Janko)

[Oil Stain]

Originally Posted by MrMastodonFarm

So, I think this is the thread you're talking about, Stain. It played out a little different then you recall.

No. That's not it. That has about 12 responses. The other was very long. I'm not going to go to in depth defending the guy. While Mudcrutch has some interesting statistical takes his delivery could use some work. He has a similar posting style as you do, and its hard to defend someone that comes off as arrogant a good deal of the time.

[Textcritic]

Originally Posted by Oil Stain

...In hindsight, I think there would be a sizable segment of the fanbase that is now leaning more towards MC79's side of the arguement...

Yes and no. If one only paid attention to the points-ledger, it would appear to support MC79's take on Phaneuf, and the reason for his early success. However, this sort of analysis fails to take into consideration the impact that Phaneuf's injuries had in his sophomore season, and fail quantify why he never appeared to be the same player since then. Was he sheltered early on? Absolutely. Does this tell the whole story about why Phaneuf failed to become a franchise defensman in his prime? Not even close, as there were mitigating factors.

[Lanny_McDonald]

Originally Posted by BACKCHECK!!!

You want to convince people that this is voodoo? Show us that there is no statistical correlation between NHL-E and a player's point production in the NHL. Or better yet, don't bother, because it gets done by many people every year, and SPOILER ALERT: there is a correlation.

I'll tell you what. I'll let you list the stats nerds employed by NHL teams in their player development side of the business, because that's what we are talking about, then I'll make you look silly by listing all the scouts who work on that side of the business. No one relies on NHL-E as an evaluation tool because it doesn't tell you anything about a player's abilities at the game of hockey. When you are evaluating players you are looking for improvements in specific skill areas, not just their scoring. In fact, many time players are given specific things to work on that can impact their scoring, which again makes NHL-E a useless statistic when trying to evaluate player development. NHL-E can't measure improvements in skating, or defensive positioning, or strength on the boards, or any other specific skill a player has to improve upon to play in the NHL. Only the evaluative skills of scouts and player development personnel can do that. Not some nerd with a spreadsheet. Hell, find me three articles, not from shady bloggers, quoting NHL teams referencing NHL-E and I'll change my tune on the measure.

[Oil Stain]

Originally Posted by Textcritic

Yes and no. If one only paid attention to the points-ledger, it would appear to support MC79's take on Phaneuf, and the reason for his early success. However, this sort of analysis fails to take into consideration the impact that Phaneuf's injuries had in his sophomore season, and fail quantify why he never appeared to be the same player since then. Was he sheltered early on? Absolutely. Does this tell the whole story about why Phaneuf failed to become a franchise defensman in his prime? Not even close, as there were mitigating factors.

Yea, I can certainly see how his injuries could have hampered his career. Phaneuf really relied on his explosive skating when he first came into the league. I think his average hockey sense would have held him back from being a franchise defenseman even without the injuries, but we'll never know for certain.

[GreenLantern2814]

Originally Posted by MrMastodonFarm

I'd like to see that thread, think you can dig it up?

There was a plethora of guys here, and media types who were disgusted with Phaneuf's play on his side of the blueline. Similar to PK Subban and Kris Lateng this season. I would imagine the backlash was more a typical homer backlash to an Oiler guy bringing down a Flames guy on a Flames forum.




Stats guys were some of the first to rang the alarm bell about Schremp?

That's an interesting take. Anyone who ever saw him skate knew it would be a tough road ahead for Schremp. Then once you knew his poor attitude it was written on the wall. You didn't need to be a stat guy to know that.

Pk shut down Crosby and ovechkin in the playoffs as a 20 year old rookie playing most of the season in the AHL. He can be over aggressive at times, but don't act like he's a liability in his own zone.

[GreenLantern2814]

Originally Posted by dying4acup

Jankowski pick was a boom or bust HS pick. To equate to the NBA, he could be Kevin Garnett, or he could be kwame brown. 1 year after the draft told us nothing. 2 years after the draft told us little. 3 yrs after the draft told us a lot.

I wasn't sure I believed this at the draft. I sure don't believe it now. A 22nd overall pick with size, skill, skating and good hockey sense is not 'top 6 or bust'. He is a player who will play in the NHL. Whether he's a number one franchise player or a good #3C, he's got everything he needs to be a contributing player in the league. And at 22nd overall, I think that's exactly what you want from a prospect.

[MrMastodonFarm]

Originally Posted by Oil Stain

No. That's not it. That has about 12 responses. The other was very long. I'm not going to go to in depth defending the guy. While Mudcrutch has some interesting statistical takes his delivery could use some work. He has a similar posting style as you do, and its hard to defend someone that comes off as arrogant a good deal of the time.

Do you think you could find the other thread?

Honestly though, I don't think it matters. If mudcrutch79 thought Phaneuf was the 80th best blueliner in the NHL, and used stats based analysis to "prove" it then it should be consistent. If he thought it in that thread, then he'd think it in others too. A silly claim at the time that only looks worse now.

I'm also not asking you to defend him, I'm asking you to show what you said above. mudcrutch79 didn't say Phaneuf was in the 40-50 range as you said he did, he said he was in the 80th range, almost doubling the position. You using this instance to show Calgarypuck posters don't understand stats based analysis is just wrong. It's a very poor example.

So that's the first example you used, which seems totally bunk, let's move on to the second one. I still haven't the foggiest idea where you got the whole stat guys were the first to ring the bell against Schremp. This is the first I've heard of this. I'd love to see some of this stuff. Taking Oiler fans and Flames fans out of the discussion, he was a guy people seemed to like or hate as a prospect. I'm not saying stat guys didn't use there algorithms to get a peg on Schremp but I highly doubt they were the first ones to do that. All you had to do was watch him skate and you knew there would be trouble.

Cute shot at me towards the end though, class act as always Stain.

Lastly, I'll quote Rubicant, who won the thread with not only the funniest, but truest thought on the matter.

Originally Posted by Rubicant

Statistics should be used in the same manner as a drunk uses a light post - for support rather than illumination.

Nailed it.

[sun]

Originally Posted by Oil Stain

No. That's not it. That has about 12 responses. The other was very long. I'm not going to go to in depth defending the guy. While Mudcrutch has some interesting statistical takes his delivery could use some work. He has a similar posting style as you do, and its hard to defend someone that comes off as arrogant a good deal of the time.

lol subtle.

[opendoor]

Originally Posted by New Era

If there was any value to them teams would be firing their scouting staffs and hiring rooms full of stat nerds.

That's line of thinking makes about as much sense as some Luddite in the '80s saying if there was any benefit in using video in coaching that teams would be firing their regular coaches en masse and hiring audiovisual nerds (with the logical implication of course being that since coaches weren't being replaced by nerds, that video had no utility in coaching sports). Just a terrible and idiotic argument in both cases.

You realize it's quite possible for different tools to complement one another, right? Teams can combine traditional scouting methods with strong statistical models in order to achieve even better results, which is what pretty much any decently run organization is already doing.

[Day Tripper]

Originally Posted by kirant

Curious: How much data did Vollman use?

AHL: 124 players
KHL: 28 players
Elitserein: 31 players
Czech Extraliga: 8 players
SM-Liiga: 9 players
National League A: 12 players
WCHA: 32 players
CCHA: 16 players
ECAC: 8 players
Hockey East: 16 players
OHL: 45 players
QMJHL: 18 players
WHL: 41 players

As you can see, there's a trade-off between sample size and avoiding outdated information.

[Day Tripper]

Originally Posted by New Era

rooms full of stat nerds.

Originally Posted by New Era

list the stats nerds employed by NHL teams

Originally Posted by New Era

some nerd with a spreadsheet

You may not agree with these methods, but is there really any need to be so blatantly anti-intellectual?

[Lanny_McDonald]

Originally Posted by opendoor

You realize it's quite possible for different tools to complement one another, right? Teams can combine traditional scouting methods with strong statistical models in order to achieve even better results, which is what pretty much any decently run organization is already doing.

Yeah, that's the point. We're being told in this thread that NHL-E says Jankowski is a total bust because his numbers at Providence were not "great." This contrary to information available from professional scouts and even those that saw him at development camp. I'm a big proponent of statistical analysis, but only if the methodology is appropriate. I find the method weak and the data unreliable.

Originally Posted by Day Tripper

You may not agree with these methods, but is there really any need to be so blatantly anti-intellectual?

Anti-intellectual? Just the opposite. Just because someone plays around with a spreadsheet and manipulates numbers does not make them an intellectual. An intellectual is someone who exercises critical thinking skills. I do not see any application of critical thinking skills to punching numbers into a spreadsheet and believing that they are a representation of the complex systems that make up a team sport where the majority of the play is considered chaos and unmeasurable to analytics. What about philosophy and psychology, both huge factors in sports today? How are these humanistic factors measured? These are actually the greatest factors that separate players and allow them to succeed at the elite level, yet they aren't considered in the models. Thinking you can replace the experience of a scout, and the understanding of the human experience that a scout brings, with a spreadsheet is anti-intellectual. I wish some of these stats nerds would apply some critical thinking skills to their projects and recognize that much of what makes a hockey player a hockey player is not measurable on spreadsheet or able to be modeled in a lab. Stats tell only half the story, and that's a false equivalency as stats don't tell 1/10th of the story, and much more attention should be paid to qualitative data from people with actual expertise in the subject matter.

Here is how useless NHL-E. This is an analysis released prior to the draft. I guess every team at the draft got it completely wrong. Or maybe they know something about NHL-E that some are afraid to admit here?

[ThePrince]

Originally Posted by New Era

Yeah, that's the point. We're being told in this thread that NHL-E says Jankowski is a total bust because his numbers at Providence were not "great." This contrary to information available from professional scouts and even those that saw him at development camp. I'm a big proponent of statistical analysis, but only if the methodology is appropriate. I find the method weak and the data unreliable.



Anti-intellectual? Just the opposite. Just because someone plays around with a spreadsheet and manipulates numbers does not make them an intellectual. An intellectual is someone who exercises critical thinking skills. I do not see any application of critical thinking skills to punching numbers into a spreadsheet and believing that they are a representation of the complex systems that make up a team sport where the majority of the play is considered chaos and unmeasurable to analytics. What about philosophy and psychology, both huge factors in sports today? How are these humanistic factors measured? These are actually the greatest factors that separate players and allow them to succeed at the elite level, yet they aren't considered in the models. Thinking you can replace the experience of a scout, and the understanding of the human experience that a scout brings, with a spreadsheet is anti-intellectual. I wish some of these stats nerds would apply some critical thinking skills to their projects and recognize that much of what makes a hockey player a hockey player is not measurable on spreadsheet or able to be modeled in a lab. Stats tell only half the story, and that's a false equivalency as stats don't tell 1/10th of the story, and much more attention should be paid to qualitative data from people with actual expertise in the subject matter.

Here is how useless NHL-E. This is an analysis released prior to the draft. I guess every team at the draft got it completely wrong. Or maybe they know something about NHL-E that some are afraid to admit here?

Huh, that list looks pretty good to me. Based solely on statistics, it managed to determine the majority of first round picks. Obviously other factors could move players up or down, but like EVERYONE has been saying, NHL-E isn't supposed to be used as the sole method of scouting, but rather as a complementary tool which can help to aid scouts determine the development of a player.

[Vulcan]

Again, I don't think anyone is saying NHL-E is the end all and be all, it's just another tool which will give an average stats
expectation for a player moving from one league up to the NHL.

[FlamesAddiction]

I think NHL-E is useful as long as people keep it in perspective. Even though the output is an exact number, it's still a pretty abstract concept and doesn't take into consideration a myriad of other factors that effect players in real life transition from one league to another.

If you think of the output more as a rating than an actual "equivalence", then it makes a lot more sense. I'm not talking about this thread in particular, but I'm not a fan when people hinge their argument's on it. It's kind of like using +/- to judge a player. It's one of those stats that has it's use only when kept within context.

[Tinordi]

Statistics are alot more than just providing a justification for your opinion. They're a bias free way of interpretting complex variables and outcomes. As I posted earlier in this thread, simple algorithms based on statistics are universally better predictors than human "expertise."

That tension is rearing its head in this thread where you have people saying that scouting and human interpretation are better predictive tools than something like NHL-E. This is based in a human conceit that we know more about the world around than we actually do. We seek out causal relationships that satisfy our biases not creating biases based on causal relationships. All of this is well documented over the past 4 decades of psychological research. Look up Paul Meehl.

What would be a useful test is to evaluate draft years using NHL-E as your main statistic to predict player outcomes. Would NHL-E perform better writ-large over the past 10 years for example? Is anyone aware of such an analysis?

NHL-E is certainly not the be-all-end-all it's flawed in many ways described in an article I posted earlier in this thread. Maybe there's a better way to statistically evaluate prospects that's more complex than NHL-e that may take into account player age, size and strength of their team. But to just make blanket statements that all statistical analysis is flawed because you don't agree with the what the stats are saying is actually more grist for the mill that the stats are probably a better indicator than your expertise.

[Textcritic]

Originally Posted by New Era

...Anti-intellectual? Just the opposite. Just because someone plays around with a spreadsheet and manipulates numbers does not make them an intellectual. An intellectual is someone who exercises critical thinking skills. I do not see any application of critical thinking skills to punching numbers into a spreadsheet and believing that they are a representation of the complex systems that make up a team sport where the majority of the play is considered chaos and unmeasurable to analytics. What about philosophy and psychology, both huge factors in sports today? How are these humanistic factors measured? These are actually the greatest factors that separate players and allow them to succeed at the elite level, yet they aren't considered in the models. Thinking you can replace the experience of a scout, and the understanding of the human experience that a scout brings, with a spreadsheet is anti-intellectual. I wish some of these stats nerds would apply some critical thinking skills to their projects and recognize that much of what makes a hockey player a hockey player is not measurable on spreadsheet or able to be modeled in a lab. Stats tell only half the story, and that's a false equivalency as stats don't tell 1/10th of the story, and much more attention should be paid to qualitative data from people with actual expertise in the subject matter.

I'm fairly confident that virtually everyone posting in this thread recognises that analytics cannot completely replace first-hand professional scouting. It is pretty ridiculous for you to mount a rant against a straw-man like this here.

Originally Posted by New Era

Here is how useless NHL-E. This is an analysis released prior to the draft. I guess every team at the draft got it completely wrong. Or maybe they know something about NHL-E that some are afraid to admit here?

This is a non-sequitor.

You seem not to understand the function and place of NHL-E in its proper employment. I think with your link to this article that you are suggesting NHL teams pay no heed to NHL-E because the draft order was substantially different than how these numbers play out. This is to be expected because NHL-E is not (and as far as I know never has been) intended to be a predictor of absolute NHL potential, only a metric by which to measure how a player's PRESENT performance will translate at the NHL level. You seem to think that NHL-E is meant to predict FUTURE potential; even in Kent Wilson's article, he has clearly indicated that this is a broad metric that functions only in conjunction with a host of other mitigating factors (Ugh, I can't believe you have me stooping to the level of defending Wilson).

[Dienasty]

Originally Posted by ThePrince

Huh, that list looks pretty good to me. Based solely on statistics, it managed to determine the majority of first round picks. Obviously other factors could move players up or down, but like EVERYONE has been saying, NHL-E isn't supposed to be used as the sole method of scouting, but rather as a complementary tool which can help to aid scouts determine the development of a player.

I totally with everything you said here.

New era using your logic I can go to a scout and look at his prior draft picks and see a he got a few choices wrong (which is the case every draft) and come to the conclusion that traditional scouting does not work.

If you looked at that list you would see that the top 5 was spot on. (Mackinnon, Barkov, Jones, Drouin and Lindholm) are all at the top Monahan who went 6th was ranked 7th out of forwards.

What this didn't take into account was size, those are called statistical anomaly's, that if you read the discussion it the bottom he says "Petan is small" (I paraphrased) not to mention Domi is also undersized and is probably the reason that Monahan went ahead of him. Drouin probably fell due to position, and track record (only one elite season under his belt)

There is a strong correlation here and that is all you can ask from statistics, you are trashing the system because it isn't 100% accurate, and that is not what the tool was created for.

[MrMastodonFarm]

MOD edit: Not necessary.

[edslunch]

Originally Posted by Tinordi

What would be a useful test is to evaluate draft years using NHL-E as your main statistic to predict player outcomes. Would NHL-E perform better writ-large over the past 10 years for example? Is anyone aware of such an analysis?

NHL-E is based on ESP/G which is a reasonable indicator of a players true offensive impact, so you're basically saying rank players by ESP/G factoring in differences between leagues (something NHL-E is explicitly designed to do). Apart from dropping all defence men out of the first round I would shocked if this didn't produce good results, writ large.

Where I have a problem is when posters use this writ small as an absolute test for an individual players potential early in their development.