Vancouver led the league in points at Christmas, Can they continue this run into the New Year?

[Bank Shot]

Or maybe the guy is sitting in a high-percentage area that suits his style of play even more, and for him 37% is his new norm. Maybe he's even underperforming what he really should be doing in that spot. Without context for why the shooting percentage is 37% and what we should really expect on average and why a player might be better/worse than average, no one knows whether it's truly sustainable or not and what we should expect going forward.

Maybe, maybe, maybe.

Sure lots of things are possible, but 99 times out of 100 when a guy playing the same role in the NHL that he did for the last 5 seasons and his shooting % doubles, it reverts to normal shortly after.

Problem is, every time someone goes on a heater, the fans of that team think that player will be the exception. They almost never are. All the guys shooting way over career average on the Nucks right are extremely unlikely to be the exception.

So many Oilers fans wanted to keep Klim Kostin last year. Some fans pointed out his extremely high shooting percentages as a reason he wasn't worth $2 million. Well there were a bunch of reasons why that didnt apply to him, and he would excel.

Now Detroit is paying 2 bills for 3 points in 20 games or whatever.

 

[Ted Hoffman]

Maybe, maybe, maybe.

Sure lots of things are possible, but 99 times out of 100 when a guy playing the same role in the NHL that he did for the last 5 seasons and his shooting % doubles, it reverts to normal shortly after.

I don't disagree by and large with your comment, but I would never structure it as an absolute. Or even a "99 times out of 100." See my comment about putting numbers in proper context.

Spoiler: Multi - and I do mean multi-million dollar idea being given away for free here:

All of teh analytics folks who churn out statistics like it's chocolate bars coming off the factory line at Hershey's? You do a whole lot of work, churn out stats, put out a whole bunch of stuff, make pretty charts and graphs for everyone to see, and talk breathlessly about guys' whatever stats you want to spew ... and you make all of your stats them in isolation, lob out "regression to the mean" like you're getting paid every time you use it, make trends off 1 or 2 data points, and then get confused why your aggregate projections don't unfold as expected.

Not one of you ever looks at exceptional or unusual data points and stops to ask a simple question: why is that?

When the first one of you starts digging below the surface, starts trying to answer the why behind the stats, and starts realizing what's really driving results and what's mere statistical noise [double spoiler: this is at both an individual and a team level] - and even more importantly, figures out how to implement that in real ways that improve results beyond "get obviously better players, have obviously better results," - you're going to make yourself a holy mother f***ing goddamn shitload of money, because this won't be some sports-specific thing you've stumbled across. Until then, relatively speaking you're spinning your wheels in the local parking lot trying to see if you can generate more smoke and a longer streak of rubber than everyone else, thinking you're about to win F1.

 

[Bank Shot]

I don't disagree by and large with your comment, but I would never structure it as an absolute. Or even a "99 times out of 100." See my comment about putting numbers in proper context.

It is very likely 99 out of 100 though. You can find hundreds of players with one year of crazy high shooting percentage compared to their career average.

Can you find me 20 players that suddenly shot up in shooting percentage and then maintained that for the rest of their career?

Heck find me 10. I'd be happy with 5 honestly.

 

[RobsonStreet]

Without any other context than simply looking at the figure relative to some average, sure - it will probably go down.

With context? Maybe it will still go down, back to the "average" figure. [Which pretends that the figure we cite is the true, correct mean going forward - but that's a discussion I'll save for another time.] Maybe the guy spent a shitload of time this offseason working on shooting accuracy and pinpointing corners, and he's been able to apply it - so, it might still go down but not as much as one might think. Or maybe the guy is sitting in a high-percentage area that suits his style of play even more, and for him 37% is his new norm. Maybe he's even underperforming what he really should be doing in that spot. Without context for why the shooting percentage is 37% and what we should really expect on average and why a player might be better/worse than average, no one knows whether it's truly sustainable or not and what we should expect going forward.

That's really the problem with every "regression to the mean" claim: it pretends to know what "the mean" is at any and all given times, assumes that all of the known information is baked into a couple numbers that can be compared, and no further context need be given to those numbers. And frequently, it only assumes "regression to the mean" and not an overshoot/undershoot in future performance which might bring about yet another "regression to the mean" claim but now in the opposite direction.


I prefer "randomness" to "luck" but I'll completely admit, that's nitpicking by invoking a statistical concept in place of some fate-based concept.


This is not wrong, but I feel like it's a bit backwards. People familiar with statistical concepts and probability distributions say "the bottom is true, but the top is what's most likely to happen." People who are bad at statistics see the bottom and say "see, anything can happen" and then when a low-probability scenario occurs, it's "proof" that "you so-called experts know nothing about statistics" and then becomes an argument for why that last outcome is not just highly predictive of what will happen next, it's "proof" that the same event will happen next; the further the observed event falls into the tail of the probability distribution, the more everyone should believe it will definitely happen again.

Keep in mind that top is the most likely to happen only if the shaded area under the curve is more than half the total auc.

While it’s true some people take the outrageous position that “anything can happen and numbers are useless”, most serious adults seem to make a different error: they misunderstand that though the mean may be the single most likely outcome, it is in itself extremely unlikely relative to all other alternatives. Ignore estimates of uncertainty at your peril.

More about this:

 

[Bank Shot]

Without any other context than simply looking at the figure relative to some average, sure - it will probably go down.

With context? Maybe it will still go down, back to the "average" figure. [Which pretends that the figure we cite is the true, correct mean going forward - but that's a discussion I'll save for another time.] Maybe the guy spent a shitload of time this offseason working on shooting accuracy and pinpointing corners, and he's been able to apply it - so, it might still go down but not as much as one might think. Or maybe the guy is sitting in a high-percentage area that suits his style of play even more, and for him 37% is his new norm. Maybe he's even underperforming what he really should be doing in that spot. Without context for why the shooting percentage is 37% and what we should really expect on average and why a player might be better/worse than average, no one knows whether it's truly sustainable or not and what we should expect going forward.

You are ignoring years of data if you choose to believe that a guy shooting 23% is somehow sustainable is the problem.

Alex Tanguay has the highest career shooting percentage in the last 35 years at 18.6%.

To expect someone to maintain at 23%+ for any kind of reason, is like a 10,000 to 1 moon shot in terms of probability.

We have a pretty good idea of what we should expect going forward for players with somewhat lengthy track records. They generally tend to hang roughly +/- 5% within their career average, and almost all top six NHL forwards in modern history hang within a pretty tight range for career shooting percentage.

To pretend that we don't have any kind of idea of what is likely or expected, is just false.

 

[Mad Dog Tannen]

Nerds!

 

[Javaman]

Still shocked and very happy the Canucks remain a top 5 team in the league nearly halfway through the season.

I know lots of people are aware of the big names (Hughes, Miller, Pettersson, Boeser, Demko) but I think what's really turned the tide for Vancouver this year is the performance of the depth players.

They're winning games even when the big 5 aren't at the top of their game. That's the most encouraging sign for me up to this point.

 

[Fishy McScales]

Deeply sorry that English isn't my first language mate and that as such, my word choices won't always be the best.

You still went far enough to read the most important stuff about the do's and don't's.

Global is not incorrect to use the way you did it.

 

[Ted Hoffman]

You are ignoring years of data if you choose to believe that a guy shooting 23% is somehow sustainable is the problem.

I've tried to point out understanding the importance of putting data in proper context twice, you're arguing against it because you continue to point to absolutes.

Have fun continuing that. I'm out.

 

[Ted Hoffman]

Keep in mind that top is the most likely to happen only if the shaded area under the curve is more than half the total auc.

While this is true, I was only considering the distribution presented, which most people are at least semi-familiar with. I really didn't think putting together a 20,000 word explanation to cover all of those aspects for every possible probability distribution - especially something (very) skewed like a gamma or lognormal distribution where parameters are selected to cause sufficient skewness in the pdf - was going to be remotely interesting for everyone else.

Maybe I'll scratch out a week for that when I have nothing else to do.

 

[Tobias Kahun]

I've tried to point out understanding the importance of putting data in proper context twice, you're arguing against it because you continue to point to absolutes.

Have fun continuing that. I'm out.

Hes not pointing to absolutes, hes pointing to what generally happens.

 

[RobsonStreet]

While this is true, I was only considering the distribution presented, which most people are at least semi-familiar with. I really didn't think putting together a 20,000 word explanation to cover all of those aspects for every possible probability distribution - especially something (very) skewed like a gamma or lognormal distribution where parameters are selected to cause sufficient skewness in the pdf - was going to be remotely interesting for everyone else.

Maybe I'll scratch out a week for that when I have nothing else to do.

You don’t need 20,000 words to say that the more extreme your prediction and/or the narrower the range of your prediction, the lesser the likelihood the observed value will fit with your guess. The underlying distribution doesn’t change this unless we’re talking about some very silly edge cases.

People that hate/reject stats often ignore the first part, people that learn just enough stats to get by in conversation often miss the second part and most every fantasy hockey forecaster magazine (that mostly reprints the previous year’s numbers as predictions) ignore both.

If you want to write a dissertation, don’t let anyone stop you, just make sure you get a PhD at the end.

 

[Buresflyingelbow]

Should have had 7 tonight

 

[Bank Shot]

I've tried to point out understanding the importance of putting data in proper context twice, you're arguing against it because you continue to point to absolutes.

Have fun continuing that. I'm out.

You are off in thought experiment lala-land.

Anyone shooting over 20% is extremely unlikely to continue that based on all historical data.

That is context.

 

[jackjohnson]

It is very likely 99 out of 100 though. You can find hundreds of players with one year of crazy high shooting percentage compared to their career average.

Can you find me 20 players that suddenly shot up in shooting percentage and then maintained that for the rest of their career?

Heck find me 10. I'd be happy with 5 honestly.

Canucks living in Oilers fans head rent free

You are off in thought experiment lala-land.

Anyone shooting over 20% is extremely unlikely to continue that based on all historical data.

That is context.

Yeah thats true but when will they drop off??

 

[snag]

It is very likely 99 out of 100 though. You can find hundreds of players with one year of crazy high shooting percentage compared to their career average.

Can you find me 20 players that suddenly shot up in shooting percentage and then maintained that for the rest of their career?

Heck find me 10. I'd be happy with 5 honestly.

Craig Simpson count? If so that's one...and a happy one i presume lol

I had no idea he had those %ages until I did an NHL stats check

Canucks living in Oilers fans head rent free

That's rich considering the thread you "jacked" and got locked tonight rofl

 

[jackjohnson]

Craig Simpson count? If so that's one...and a happy one i presume lol

I had no idea he had those %ages until I did an NHL stats check



That's rich considering the thread you "jacked" and got locked tonight rofl

All the while you are making posts in the Canucks thread.

 

[Tobias Kahun]

All the while you are making posts in the Canucks thread.

You comment in every Oilers thread.

Our ECHL D man got 3 points tonight.

Not bad.

 

[snag]

All the while you are making posts in the Canucks thread.

I am not acting like a douche in this thread, am I?

 

[Dust]

The good news for the Canucks is that if this was last year, they would have found a way to blow that game against New Jersey. That goes a long ways to solidifying them as a playoff team.

 

[jackjohnson]

The good news for the Canucks is that if this was last year, they would have found a way to blow that game against New Jersey. That goes a long ways to solidifying them as a playoff team.

So it took you this long to determine Canucks are a playoff team this year having 53 points in 39 games?? isnt it too soon?

 

[Sykur]

 

[Dust]

So it took you this long to determine Canucks are a playoff team this year having 53 points in 39 games?? isnt it too soon?

I'll try to be nice, but you should stick to the Vancouver boards. All of the other Canucks posters have that "he's not with me, I don't know him" look on their faces when you post.

 

[jackjohnson]

I'll try to be nice, but you should stick to the Vancouver boards. All of the other Canucks posters have that "he's not with me, I don't know him" look on their faces when you post.

I can express my view here if i like. Maybe you should post in the oilers board so you would not encounter opposing opinions about the oilers

 

[HockeyWooot]

This Canucks team is miles better than last year the one last year under Boudreau.

Whether they remain at this place in the standings by season end remains to be seen, but they have definitely made the jump from mediocre team to good.

The compete level, attention to details, accountability and overall culture has been the big change.

They still have some roster issues to address to take that next step, but they have enough difference makers that can help them win games. Sometimes it's the like of Hughes/Miller/Pettersson, sometimes Demko. Of late it's been the depth particularly the third line.