I do not propose to rate posters - read please

3rdside

Hall of Famer
Joined
Nov 4, 2002
Posts
1,531
Reaction score
202
Location
London, UK
I don't know why you didn't finger me as that quotation's author, but I'm happy to stand by it.

I did initially when replying to the quote but starting a new thread i didn't want to appear like i was clowning arguably the most credible poster here. Which makes the quote even more baffling considering how much certainty you ascribe to something that hasn't happened yet.

Your idea is cute, but you need to standardize how Dragic's "score" will be calculated. Otherwise people will just spin his mixed performance so as to fit their own narratives.

Popular vote would be the best proxy. Take out the top 25%, bottom 25% and average the rest. Of course you could do something exactly quantifiable; in Dragic's case predict his PER, or if you wanted to look at other areas; number of wins in a season is the obvious one, conference position after 82 games etc.

The problem with this (besides being a complete time suck), is that if it any way uses numbers, it will be wrong.

Like any analysis it will take time but the results are generally worth it for the insight. And I'm telling you numbers will tell you something meaningful so long as you get enough data.

Bad idea, bbb.net had something like that and it accomplishes nothing. In the end its a popularity thing and has nothing to do with credibility.

Not sure how popularity comes into this?

Credibility is earned through reputation. You have to post to do that. The guy with 221 posts over a decade calling out a quote by a guy with 16K posts over the same time period on the issue of credibility ... that's weak sauce.

I didn't call him out firstly and not sure you can convince me that quantity is better than quality secondly. In Eric's case, sure, but it's not quantity alone that does it.

For the record I tend to agree with your statement Eric about Dragic, just not with the level of certainty you have. The stats below for Nash and Dragic might support my view - both players have played 5 years and both were 22 years plus change coming into their first season.

Nash

SEASON GP MIN REB AST STL PF TO PTS AST/MIN PTS/MIN AST / TO
'96-'97 65 10.5 0.9 2.1 0.3 1.4 1.0 3.3 0.20 0. 31 2.10
'97-'98 76 21.9 2.1 3.4 0.8 1.9 1.3 9.1 0.16 0. 2 2.62
'98-'99 40 31.7 2.8 5.5 0.9 2.5 2.1 7.9 0.17 0.25 2.62
'99-'00 56 27.4 2.2 4.9 0.7 2.2 1.8 8.6 0.18 0.31 2.72
'00-'01 70 34.1 3.2 7.3 1.0 2.3 2.9 15.6 0.21 0.46 2.52

SEASON GP MIN REB AST STL PF TO PTS AST/MIN PTS/MIN AST / TO
'08-'09 55 13.2 1.9 2.0 0.5 1.6 1.3 4.5 0.15 0.34 1.54
'09-'10 80 18.0 2.2 3.0 0.6 1.6 1.6 7.9 0.17 0.44 1.88
'10-'11 22 17.2 2.5 2.5 0.6 1.6 1.0 7.7 0.15 0.45 2.50
'10-'11 48 17.8 1.8 3.1 0.8 1.9 2.0 7.4 0.17 0.42 1.55
'11-'12 66 26.5 2.5 5.3 1.3 2.5 2.4 11.7 0.20 0.44 2.21


Coming off the 5th season, Nash and Dragic's Ast / Min and Pts / Min are similar, only the Ast /TO is there much of a difference - in Nash's favour offset by the greater number of steals Dragic makes.

Summary:

i) Gospel statements sound strange and no one will convince me - level of intellect, number of posts, or otherwise - that I should believe them regardless of the level of confidence in which they are stated.

ii) Having a reference, even if not that strong, with which to gauge the credibility of a strongly worded statement would be a good thing. Conversely, those who make strongly worded statements may be less inclined to do so if, statistically, their forecasts were consistently off.

iii) It would be straight up readable to see some stats over a 5 / 10 year time frame; we've been here that long, some sort of record for the next ten wouldn't be a bad thing.
 

elindholm

edited for content
Joined
Sep 14, 2002
Posts
27,447
Reaction score
9,604
Location
L.A. area
For the record I tend to agree with your statement Eric about Dragic, just not with the level of certainty you have.

Well obviously it's a prediction, not a guarantee in the literal sense. I assumed that was obvious from the context. I'm not sure what probability I'd put on it, maybe 85-90%.

The stats below for Nash and Dragic might support my view

No they don't. First of all, Nash already had excellent shooting mechanics -- just look at his FT% in college and early during his NBA career. Also, Nash already knew how to protect the ball with his dribble while keeping his head up, which Dragic can't do and hasn't spend any off-seasons getting better at. This is how Nash could become so expert and weaving through a defense, probing for weaknesses, while Dragic panics easily while dribbling and can be pressured into turnovers (or unloading the ball like a hot potato) quite easily. In short, Nash makes defenses pay for overplaying him, while Dragic can't, and that's because Nash has (and had, even early in his career) fundamental tools that Dragic lacks and hasn't bothered to acquire.

But even more important, Nash is a unique phenomenon. This is what people tend to forget. No other PG in the history of the league has had his best years in his early 30s, or has played as well in his late 30s as he did in his mid 20s. It's like if you buy a seat in a baseball stadium deep in the bleachers, because a guy you know once caught a 440-foot tape measure home run at that exact spot, and you're telling people, "Sure, I can catch a home run here." Yes, it did happen once, as a thousand-to-one shot, but that's not much of a reason for believing it can happen again.

Nash as a precedent does not support any argument for any other player.

i) Gospel statements sound strange and no one will convince me - level of intellect, number of posts, or otherwise - that I should believe them regardless of the level of confidence in which they are stated.

No, of course not, don't believe it if you don't want to, or if it doesn't feel right to you. We'll know soon enough. It was the Suns front office's responsibility to try to predict the future, not ours. My guess is that they made a mistake, and that we'll all be in agreement on that within a year or two.

ii) Having a reference, even if not that strong, with which to gauge the credibility of a strongly worded statement would be a good thing.

Well I'm not sure what you mean here. If other posters chimed in and said, "This guy is usually right," that wouldn't mean anything. My reasons for being pessimistic about Dragic come from having watched him play quite a bit and from having seen that certain organic weaknesses in his game have not gotten one shred better from when he entered the league. He hasn't demonstrated a hunger to improve, and that's why he has struggled so far when he gets competition even from a teammate for playing time, never mind competition from an opponent for being effective on the floor.
 
Last edited:

CardsFan88

ASFN Addict
Joined
May 28, 2002
Posts
7,510
Reaction score
4,465
Sorry, number systems are flawed because they use numbers. Statistics is false. It is not even a legitimate study, it is quite the opposite. Colleges would be better off ditching it 100 percent.

The whole theory behind sample size is wrong, and sorry there is no way to equal out the variables because they are infinite.

Look, statistics aren't science, they're a bunch of bull. I can label a million examples of how badly they've screwed up society, but then it'd turn into a P&R discussion.

The simple fact is, there is NO LEGITIMATE way to use numbers for this purpose. None. All the college professors are wrong....they sell nothing more than snake oil.

Statistical theory, is not a theory, it is made up fantasy land. Getting more data will not make it 'better', because the theories behind such idiotic thinking are completely wrong. Bigger sample sizes doesn't mean squat. 0 or 1.

It should be a crime to teach statistics as valid, let alone charge to teach it. It isn't real. It only makes people believe they are doing something, when really whittling wood is more useful.

There is no such thing as 'confidence' in statistics. This is made up bull. Quit thinking like a 'statistician', they're useless and wrong about everything they get their grubby idiot hands on.

Even outside of it, 'confidence' is just herd thinking, predicated by another useless profession.

Instead you want this absolutely fantasy land number system, based on nothing valid, to be the sole rational of whether a poster is good or not? A freaking number that in no way resembles reality?

Here's the truth. Statistics are about a fantasy land with fractional numbers between 0 and 1. The entire rational of statistical theory is between 0 and 1. It is about trying to make something up to pretend something attributable to a fake number is equal to something else.

The real world, and everything inside of it, is only 0 or 1. Something either is or isn't, and the number of these is and isn'ts are literally infinite.

Again, I understand where you are coming from, the quest to find better truth, but the fact remains your methodology is completely wrong and can never work.
 
Last edited:

elindholm

edited for content
Joined
Sep 14, 2002
Posts
27,447
Reaction score
9,604
Location
L.A. area
Sorry, number systems are flawed because they use numbers. Statistics is false. It is not even a legitimate study, it is quite the opposite. Colleges would be better off ditching it 100 percent.

The whole theory behind sample size is wrong, and sorry there is no way to equal out the variables because they are infinite.

Look, statistics aren't science, they're a bunch of bull. I can label a million examples of how badly they've screwed up society, but then it'd turn into a P&R discussion.

The simple fact is, there is NO LEGITIMATE way to use numbers for this purpose. None. All the college professors are wrong....they sell nothing more than snake oil.

Statistical theory, is not a theory, it is made up fantasy land. Getting more data will not make it 'better', because the theories behind such idiotic thinking are completely wrong. Bigger sample sizes doesn't mean squat. 0 or 1.

It should be a crime to teach statistics as valid, let alone charge to teach it. It isn't real. It only makes people believe they are doing something, when really whittling wood is more useful.

There is no such thing as 'confidence' in statistics. This is made up bull. Quit thinking like a 'statistician', they're useless and wrong about everything they get their grubby idiot hands on.

Even outside of it, 'confidence' is just herd thinking, predicated by another useless profession.

Did you come up with eight different ways to write the same paragraph, then forget to choose just one and post all eight by accident?

Let me guess -- a LaRouche supporter?
 

CardsFan88

ASFN Addict
Joined
May 28, 2002
Posts
7,510
Reaction score
4,465
Did you come up with eight different ways to write the same paragraph, then forget to choose just one and post all eight by accident?

Let me guess -- a LaRouche supporter?

ROFL, there are many avenues to hit statistics. It's a false god. Sometimes it takes awhile, as I continuously think of a better way to approach something, or how to include more. Sometimes you re-read the original persons quote and something else pops out at you. That sort of thing. The main point is to get the message across.

I support what's real, not fake fraud stuff. If someone else holds that opinion, and I feel they are correct, you can deem that support or not. I know who has been right and who has been wrong. Consistently. It ain't the statisticians who have been right.

For this discussion, since the whole premise of creating a number system based on statistical theory as it's validation, it is just another manifestation of fraud, but on a teeny, tiny, white lie scale.

Maybe I should hold back. Or I realize that you show statistical nonsense no quarter, anywhere, at any time. That and why would I want to see someone waste their time so fruitlessly? If you saw someone trying to dig to china, wouldn't you say...hey...it ain't gonna happen. Well.....

Fact remains even my statistics teacher at ASU, spent the whole semester, talking about how bull it was.

Besides, it's pretty simple to blow holes through statistics, it messes up everything it touches. All one needs to do is to see it's failure everywhere it has been applied. The scary thing? Look at how much is entrusted to it. Scarier than the scariest movie of all time by a million.
 
Last edited:

elindholm

edited for content
Joined
Sep 14, 2002
Posts
27,447
Reaction score
9,604
Location
L.A. area
I notice you dodged my question about whether you are a LaRouche supporter.

Statistics can be misinterpreted and misapplied just like any other evidence, but to say that the whole field is bunk is simply ignorant. Either you misunderstood your professor or he was exaggerating to make a larger point that you didn't get.

No other conversation will be fruitful, since rambling conspiracy theorists as a population -- oops, there's a statistic for you -- are incapable of rational discourse.
 

CardsFan88

ASFN Addict
Joined
May 28, 2002
Posts
7,510
Reaction score
4,465
I notice you dodged my question about whether you are a LaRouche supporter.

Statistics can be misinterpreted and misapplied just like any other evidence, but to say that the whole field is bunk is simply ignorant. Either you misunderstood your professor or he was exaggerating to make a larger point that you didn't get.

No other conversation will be fruitful, since rambling conspiracy theorists as a population -- oops, there's a statistic for you -- are incapable of rational discourse.

Actually I didn't dodge your question at all. Since there are rules on the forum, I was trying to make a very concise point. I support what's REAL, something that was obviously missed. It was quite clear where I stand, even if I don't use your term, support, because I believe in ideas over people. I focus on facts, over what a person says using a forked tongue and sophisticated answers. Supporting people is folly. Focusing on ideas that a person holds is a much cleaner way of determining things.

I know statistics is false, based on many sources, but you can't seem to accept that reality.

Is there something wrong with said person? If you think yes, you are patently wrong on so many levels. You are using it as a qualifier, or disqualifier in this case, one that is blatantly transparent, and wholly ignorant.

No I didn't misunderstand my professor. Where do you think 0 or 1 came from? That was the #1 thing he wanted us to remember about statistics. Am I wrong that entire course of study is based on everything inbetween? Nope.

The professor played a game, and became very rich at it. But tried to instill in us to let OTHERS interpret what the numbers mean, because every interpretation is inherently wrong. I don't agree with playing the game, I believe in ending it.

In this case, someone wants to spend time fruitlessly creating nothing valid by manipulating numbers. I understand the desire, I am not knocking anyone on that. The methodology however, is completely wrong. Knowing how false in reality the statistical profession is, I cannot see someone trying to develop a fake game as I'll say from scratch, and think it is somehow legitimate.

Statistics are always wrong. They may APPEAR to be right, like a broken clock twice a day, but never because the output somehow recreated the universe in numbers and spit the correct output out. Misapplied means anywhere, because there is no legitimacy in the entire field. NONE. Just ask the weather man.

Actually the funny thing is, where did I bring up a conspiracy theory?

Let alone the fact remains, history is littered with conspiracies. Guess you forgot history. Or choose to ignore it.

But what about the rational and sophisticated present? There can't possibly be conspiracies amongst humans which invented conspiracies.

Well just go down to your local courthouse....you'll find.
Conspiracy to.....x, y, and z.
Commit Murder
To defraud
The list is endless, everywhere in this country. Court is in session every working day revolving around them.

So the fact is conspiracies, in whatever form, are actually quite common, not that anything in my posts contain any. I didn't mention a single conspiracy, that is on you for interjecting it. One which the entire premise is wrong by the way, based on history and the present.

The simple fact is statistics are fake. They will never be real. They will never be science. They are manipulation of numbers. They are pseudo in every sense of the word.

It's all good man, believe what you want to believe. But the facts on the matter are clear. If you choose to ignore that, fine. If that somehow makes ME the one that is either not civil, or incapable of having discourse, then you're wrong.

We can argue a dead man isn't dead. But the fact remains a dead man is dead. Just like statistical theory is fraudulent by it's very nature, evidenced by it's mess ups everywhere it is used. Use common sense, and tell me if you can tell me how many total variables there are in the universe and whether or not a statistical model has incorporated them all, and been correct 100 percent of the time, and stood the test of time, forever on it. You can't, because it doesn't exist. It never will. This is an easy concept.

We're not talking about a touchdown is 7 points. That is not statistical theory. We're talking about manipulation of numbers of a few variables out of an immeasurable amount, and trying to claim there is validity in them. That is what you are defending. That is what it ALWAYS is.

Again it's all good though, we still have a team in common, and my explanations would be the same no matter who the poster was.
 
Last edited:

elindholm

edited for content
Joined
Sep 14, 2002
Posts
27,447
Reaction score
9,604
Location
L.A. area
The simple fact is statistics are fake. They will never be real. They will never be science.

One characteristic of LaRouche followers -- if you don't like the word "supporters" -- is that they hide behind convoluted rhetoric without ever quite getting to the point they're ostensibly making. You have the patter down well, and if you aren't in fact one of his followers, you could pass for one easily.

But with the above, you actually went out on a limb and made a real claim: That statistics aren't science. And on that point, you are unfortunately unequivocally wrong.

I will offer a simple example, which is that of radioactive decay. A radioactive sample contains a large number of nuclei, each of which will eventually decay, splitting into two or more smaller parts and releasing energy in the process. (Hopefully we can at least agree on this.)

It is well known, confirmed thousands of times over, that you can characterize the time scale of this process using something called a "half-life." The half-life is the amount of time that will elapse before approximately half of the atoms in the sample decay. One must say "approximately," but because the number of atoms is so large, in fact the variance on a percentage basis is small.

It is impossible to predict when a given nucleus will decay. Nor is it possible to predict which half of them will decay within the half-life. And yet you can predict with confidence that, sure enough, roughly half of the atoms will have indeed decayed during that interval.

That is pure statistics, and it is also science. And it is fact. It would be easy to come up with lots of other examples, of course, but this one is vivid and clear.

Now, from your signature file, I surmise that what really frightens you about statistics is the extent to which some economists seem to rely on them in order to advocate for certain economic policies. And I agree that that is a questionable enterprise. But that is not the fault of the statistics; but instead of their misuse.

We can argue a dead man isn't dead. But the fact remains a dead man is dead.

This is a classic LaRouchean tactic -- appealing to a rhetorical truism as though it is somehow germane to the discussion. But we aren't arguing about whether a dead man is dead.

Use common sense, and tell me if you can tell me how many total variables there are in the universe and whether or not a statistical model has incorporated them all, and been correct 100 percent of the time, and stood the test of time, forever on it. You can't, because it doesn't exist. It never will. This is an easy concept.

Yes, it is an easy concept, and it reveals how little you understand about statistics. Statistics do not predict certainties, but instead probabilities. If such-and-such a weather pattern has produced rain in 800 out of the last 1000 times that it has been present, then one may predict with reasonable confidence that the probability of rain is 80%. But that is not a certainty.

We're talking about manipulation of numbers of a few variables out of an immeasurable amount, and trying to claim there is validity in them. That is what you are defending.

I'm not defending any such thing, and putting words in the other party's mouth is another LaRouche tactic that I can't stand.

Look, I've tried to have discussions with your kind before. The discussions always fail because I can't keep the other person on a single point of contention long enough to refute it. LaRouchies bob and weave, throw out bizarre distractors ("A dead man is dead"), appeal to grand irrefutable statements ("You can't tell me how many total variables there are in the universe"), and yet somehow never reveal exactly what it is that they're trying to convince you of. It is, I admit, an intriguing phenomenon, and a "debate" style that is a skill in its own right, if ultimately fruitless.

Again it's all good though, we still have a team in common, and my explanations would be the same no matter who the poster was.

Yes, it should be obvious that this isn't personal. I'm just trying to get a clearer handle on why your thinking is so muddled. When did they get to you?
 

Cheesebeef

ASFN IDOL
Supporting Member
Joined
Jan 2, 2003
Posts
91,306
Reaction score
68,280
uh... is this like the intellectual version of the fights me and Chap get into?
 

CardsFan88

ASFN Addict
Joined
May 28, 2002
Posts
7,510
Reaction score
4,465
One characteristic of LaRouche followers -- if you don't like the word "supporters" -- is that they hide behind convoluted rhetoric without ever quite getting to the point they're ostensibly making. You have the patter down well, and if you aren't in fact one of his followers, you could pass for one easily.

Actually I made the point quite clear. 0 or 1. Probability, confidence intervals, correlation, never actually hit 1, and no matter how high they go up it's still a .00000000001 or a .999999999 or somewhere in between or even further out in the fractions of a number. They never actually hit 0 or 1. This is reality. Which means in reality, it never hits anything. This is quite simple and easy. They never tell you 0 or 1. Thus they never tell you anything real. If they could tell you 0 or 1, and are as constant and reliable as decaying nuclei, I'd say, hey statistics tell you something. But they don't.

Statistics, are man made, and they are guesses, nothing more. Guesses can be right, but the process of guessing isn't science, even if it sometimes it first involves a guess from an observation. The science isn't the guess. It's the result of whether the process observation cannot be disproved. Don't work it backwards. Science isn't guessing. Science is the end result process irrespective of whatever guesses came before it. Think of it this way. Whatever is .97999 can have it's model altered to be .98999 or then again to .99999. It's just manipulation of numbers, variables, adding or subtracting one. Trying to justify it.

I said I agree with ideas, which means I don't follow people or support people per se. I agree with certain ideas that can be attributed to various people. But belief in PEOPLE instead of IDEAS, is a derivative that can and usually does lead to a disassociation from facts.

Which means people don't need to understand or even think about the idea or find out what is right or wrong, valid or invalid, they'll just believe the person to have figured it out. Then even more disassociation from the idea singularly, is a group of people that might give each other a name. Then they'll believe because the group says so. The best way is to understand the idea. The merits of an idea is what you throw your support behind, if it factual, based on actual science, and overall is real. So if I 'support' this group or that group, would actually be incorrect. I support specific ideas that happen to be fostered by various groups.

Thus I was just being honest that I don't 'support' a person. It's quite clear the rationale behind it. The people you describe to have a problem with, actually have ideas that are real. Maybe some of them don't understand. Any group has people that follow people or groups instead of ideas. Maybe you talk to them. I don't know.

I'm not using any tactic. No one has ever taught me any tactic. I never went to some spin school or something lol. I just try to explain something because I feel someone isn't seeing something, and thus multiple directions tends to isolate an idea. Metaphorically boxes it in. Draws a frame. Something. So if I come at a problem multiple directions it is just to put a better picture in the reader's head and/or give another singular one they can grasp onto.

Like....When you buy a hamburger are you supporting McDonalds, or are you eating because you are hungry. You are eating because you are hungry, it just so happens that at that meal, you are 'supporting' McDonalds.

Of course you are 'supporting' McDonalds with your money. But that wasn't why you bought McDonalds. You were hungry, and various factors, maybe distance, maybe taste, in this case random fast food purchase as an example of a routine process was why I used it metaphorically.

So change the metaphor from eating and McDonalds to an idea that a person/group fosters, and hopefully you can understand what I mean.

But with the above, you actually went out on a limb and made a real claim: That statistics aren't science. And on that point, you are unfortunately unequivocally wrong.

I will offer a simple example, which is that of radioactive decay. A radioactive sample contains a large number of nuclei, each of which will eventually decay, splitting into two or more smaller parts and releasing energy in the process. (Hopefully we can at least agree on this.)

It is well known, confirmed thousands of times over, that you can characterize the time scale of this process using something called a "half-life." The half-life is the amount of time that will elapse before approximately half of the atoms in the sample decay. One must say "approximately," but because the number of atoms is so large, in fact the variance on a percentage basis is small.

It is impossible to predict when a given nucleus will decay. Nor is it possible to predict which half of them will decay within the half-life. And yet you can predict with confidence that, sure enough, roughly half of the atoms will have indeed decayed during that interval.

That is pure statistics, and it is also science. And it is fact. It would be easy to come up with lots of other examples, of course, but this one is vivid and clear.


Let's see, actually the description of the rate of nuclei isn't statistics. You can describe it as a process that can be described numerically. You are confusing the two. It isn't however a statistical model. You see man made number manipulation of any sort, isn't a natural process of science. Both can attempt to describe something numerically. Only one is actually real.

It isn't a natural pattern governed by the laws of the universe. You can attempt to recreate something by number manipulation in a similar pattern, or develop a brand new one, but it doesn't make it science. That is the pattern of decaying nuclei. It isn't if Steve Nash is better than Joe Johnson, or something in value will go up or down, or whether it will rain tomorrow based on x, y, and z.

You're confusing natural patterns with man made ones, and equating the two as equal. They are assuredly not. That is the main point. Joe Shmoe at Harvard manipulating numbers whether it be for whatever profession, or in this case for fun to see what player is better or poster or whatever, is not the same as decaying nuclei. This is crystal clear. Science isn't man made. Man can only discover science. It cannot create it. Only learn ways to access it and benefit from that. But the underlying Science has always been there waiting to be accessed.

I just don't see how it's not obvious. Decaying nuclei is not Joe Shmoe's guesses, even if both can be described numerically.

Now, from your signature file, I surmise that what really frightens you about statistics is the extent to which some economists seem to rely on them in order to advocate for certain economic policies. And I agree that that is a questionable enterprise. But that is not the fault of the statistics; but instead of their misuse.

This is a classic LaRouchean tactic -- appealing to a rhetorical truism as though it is somehow germane to the discussion. But we aren't arguing about whether a dead man is dead.

You're so close on this note. The practice of utilizing statistics for any reason is as valid as any economist. The uses don't matter. You can get advocates for it for various economic schools of thought, or for when it rains, there are so many aspects it goes into. They are all guesses just the same, and none of them are like the decaying nuclei.

It's just the real world disasters that arise from it are more widely felt since it happens in integral processes to humanity. The point is, whenever it is used (man made), is as valid as any other man made use. They are pure guesses, that aren't science. Just certain places we feel the impact more than others. Some could be more often right than others, but they are still all guesses prone to major, and sometimes catastrophic errors.

Thus my point was with the analogy of dead people was simple, I was attempting to show you that Joe Shmoe isn't decaying Nuclei, no matter how much we argue it won't change that a man made model isn't a scientific process of decaying nuclei. There is a huge difference between the two. One is correct. One isn't. By it's nature man made models and probabilities will never give you a one.

Yes, it is an easy concept, and it reveals how little you understand about statistics. Statistics do not predict certainties, but instead probabilities. If such-and-such a weather pattern has produced rain in 800 out of the last 1000 times that it has been present, then one may predict with reasonable confidence that the probability of rain is 80%. But that is not a certainty.

You see rain 800 out of 1000 times doesn't mean anything on whether it actually is going to rain or not at any time in the future. I understand it quite well. The process will either initiate or not, irrespective of whatever the numbers state, based on the actual factors present in the process at hand, that a model is not derived of. You keep saying I don't understand it. I'm describing the limitations of the entire field. It's 0 or 1. 800 out of 1000 isn't either and never will be. Thus you cannot ever actually use the knowledge.

Thus any probability is inherently meaningless, even if it hints at something. Obviously the indicators show that something that causes rain is present and that it might occur, but the models guess is as good as anyone's as it doesn't accurately describe the process. Using this as a foundation for something is what it is always used for, otherwise it's just academic nothings. If we're talking in a classroom for just crap and giggles is one thing. But to ever apply it, is asinine.

An additional fact is that they rarely state that it is a guess. Most people don't describe it as a guess that might either happen or not. They say 800 out of 1000 means something highly likely. Just like .999 that something isn't going to happen means it will almost never be likely to happen. But those numbers are always wrong. Because future isn't the past. We're talking about separate events, and statistics pretends they are linked. They try to teach you six sigma, and try to tell you to prepare for that level, which is really just an out. The fact remains that six sigma events are breached with regularity that makes the whole idea of six sigma moot. It happens WAY more often.

Either way, when you substitute an actual scientific process, with a man-made guess you've only introduced error, not eliminated it. Put it in a course, and it supposedly is an legitimate out for failure. Well it met Six sigma standards. This is what we base our world off of, is crap like this. It's failure built in at every level.

I'm not defending any such thing, and putting words in the other party's mouth is another LaRouche tactic that I can't stand.

Look, I've tried to have discussions with your kind before. The discussions always fail because I can't keep the other person on a single point of contention long enough to refute it. LaRouchies bob and weave, throw out bizarre distractors ("A dead man is dead"), appeal to grand irrefutable statements ("You can't tell me how many total variables there are in the universe"), and yet somehow never reveal exactly what it is that they're trying to convince you of. It is, I admit, an intriguing phenomenon, and a "debate" style that is a skill in its own right, if ultimately fruitless.

I wasn't trying to put words in your mouth, so if I did, I apologize.

Actually most people use analogies. The fact remains that no man-made model will never contain in it, every possible variable, in proper 'weighting' that can describe a process exactly in a way that gives the future. The human mind has not deciphered basketball, and physics, and the universe it resides in enough to describe, especially backwards by guessing, a model that numerically describes the process of figuring out better players or posters, or whatever.

We've hitched our wagon to guesses based on number manipulation and introduced error based on mathematics as a rule, not an exception. Using models to find the probability of rain is inferior to actually scientifically determining the processes of rain and everything connected to it.

I haven't bobbed and weaved anywhere. I've stated clearly the fallacies of statistics. Because there is the angle within the sphere of it's use of 'it's only probabilities' as if probabilities are in any way valid. They are nothing more than guesses. But these guesses are used everywhere in real situations where facts are needed. Thus their inclusion at any time, is asinine. We've guaranteed failure above and beyond what might otherwise occur. We may say 'they are just probabilities', but by any way we use them, that disclaimer is thrown right out the window.

Yes, it should be obvious that this isn't personal. I'm just trying to get a clearer handle on why your thinking is so muddled. When did they get to you?

My thinking isn't muddled. I don't lump together man made models with scientific processes that can be described with numbers. I'm also not the person that implements it in everything (obviously not saying you), and considers it a good idea to create them for every situation rather than finding an actual scientific way to describe something, if even ever possible. I see it as a futile exercise, because it will never be true, and one will never really know how untrue it is, just that it will never BE true. Thus the entire point to attribute something so complex based on a model is one that is folly. Why even engage in it? Besides maybe fun. But that wasn't the goal. It was to determine something. You can't determine anything based on a model. Finding conclusion on a bedrock of guesses just doesn't make sense.

Again, I believe in factual ideas. Not a particular person. You could say I subscribe to ideas that are subscribed by many people and groups, even if on the surface they are very different. That's why I can visit some sites that one might see as anti the group you mention, yet in reality have certain commonalities with specific ideas. Ideas aren't always fostered by one group. You can find many of the same ideas fostered by many different groups. It isn't like the group you mention has created many of these ideas. Most of what they and others support, originally come from a different wellspring. But I also tend to gravitate around certain groups which have a boatload of ideas and/or valid pertinent questions.

I tended to think you weren't really getting personal, it's just when people get the impression the 'you people' type expression is being thrown at them, one does start to question it. But it happens, so you know whatever. It's all good. I guess you have to be on the receiving end, you were instead on the writing end. No problemo.

Just take one thing. Scientific processes described numerically or as a numeric process are not the same as a man made one. One is real. The other cannot be. Because if man discovers a pattern that is scientific, it is real process that has always been there....a person creating a model, or gauging probability, is just guessing. The two are not the same. Model makers never guess right. They never will.

I hope you see what I at least see as an easy distinction between the two. Not inferring anything if not, just to me, it's plain as day what I'm trying to describe. Because the profession is not being used to discover decaying nuclei type processes, it's being used to guess. (obviously not absolute, but I consider the two separate)

Worse still, is just about everyone takes it as fact. The proprietors of the guesswork, aren't doing their job at explaining the disclaimer....and to bring it back why I started this whole thing is that whenever someone tries to create some validity from numbers like this out of thin air, it's going to be wrong, and can be another instance of fooling people into forgetting the disclaimer. Even the author. If we all go around practicing it regarding everything, what good is the disclaimer, since it's tenants are being disregarded.
 
Last edited:

elindholm

edited for content
Joined
Sep 14, 2002
Posts
27,447
Reaction score
9,604
Location
L.A. area
Actually I made the point quite clear. 0 or 1. Probability, confidence intervals, correlation, never actually hit 1, and no matter how high they go up it's still a .00000000001 or a .999999999 or somewhere in between or even further out in the fractions of a number. They never actually hit 0 or 1. This is reality.

So far, so good.

Which means in reality, it never hits anything.

Nope, you lost me.

This is quite simple and easy. They never tell you 0 or 1. Thus they never tell you anything real.

I guess your point is that no information about the future is useful unless it offers an iron-clad prediction. I disagree. For example, you say you went to college; presumably one reason was that you hoped to learn something. And probably you did, so it worked out, but there was no guarantee that you would learn anything in college. Only a likelihood.

I could go on and on, but I suspect that we can't bridge this fundamental disagreement. I believe that one can make good decisions based, in part, on assessing the likelihood of possible outcomes. You don't share that belief, I guess, preferring instead to dwell on the (accurate, but uninteresting) fact that nothing in the future can be known for certain.

Guesses can be right, but the process of guessing isn't science, even if it sometimes it first involves a guess from an observation. The science isn't the guess. It's the result of whether the process observation cannot be disproved. Don't work it backwards. Science isn't guessing. Science is the end result process irrespective of whatever guesses came before it.

How many science courses did you take at ASU? Very few scientific results can specify precise future outcomes.

I'm not using any tactic. No one has ever taught me any tactic. I never went to some spin school or something lol.

I don't believe you. How was I able to peg you right away as -- hang on, let me get this right -- someone who is not a LaRouche supporter or follower, but who supports ideas shared by LaRouche and his followers? Your allegiance just leapt off the screen.

Let's see, actually the description of the rate of nuclei isn't statistics. You can describe it as a process that can be described numerically. You are confusing the two.

No I'm not. You are invoking a specialized definition of "statistics" that I'm not familiar with.

You see man made number manipulation of any sort, isn't a natural process of science.

I think your point -- and, by the way, it should bother you that I'm better at explaining what you mean than you are -- is that statistical models do a better job of representing the physical world than they do with the human one. I would agree with that.

You're confusing natural patterns with man made ones, and equating the two as equal.

I'm not doing any such thing. Why don't you slow down and try to identify what, exactly, it is about anything I've written that you think is incorrect.

It's just the real world disasters that arise from it are more widely felt since it happens in integral processes to humanity. The point is, whenever it is used (man made), is as valid as any other man made use. They are pure guesses, that aren't science. Just certain places we feel the impact more than others.

I think your point -- here we go again -- is that, when trying to enact policies that have potential world consequences, one must recognize the danger of relying on mathematical models. I would definitely agree with that. On the other hand, what is the alternative? Even a decision to do nothing is still a decision. What better way is there to proceed other than making the best guess possible based on the available data, as imperfect as that process might be?

I'm describing the limitations of the entire field. It's 0 or 1. 800 out of 1000 isn't either and never will be. Thus you cannot ever actually use the knowledge.

Why not? Do you smoke? If you don't, is it because you believe that smoking is potentially harmful to your health? Some people smoke five packs a day and live until 95, so there are no guarantees, and yet nearly everyone would agree that, if you are interested in long-term health, it makes sense not to smoke. Do you reject that analysis?

The fact remains that six sigma events are breached with regularity that makes the whole idea of six sigma moot. It happens WAY more often.

That's just completely incorrect. Six-sigma events appear to happen frequently only because we don't notice all of the times that they don't happen.

Well it met Six sigma standards. This is what we base our world off of, is crap like this. It's failure built in at every level.

Again, what's the alternative? I read somewhere that the space shuttle engineers estimated that the probability of any mission ending in catastrophic failure was about 1 in 50. Well, sure enough, we had two catastrophic failures, out of ballpark 100 missions. Now frankly, if I were going to be one of the astronauts, and someone told me those odds, I'd probably decline. But the men and women who went into space knew the risks and took them on. Since you can't even guarantee that your heart will continue to function for the next 24 hours, how do you decide where to set your tolerance for risk? Given that zero risk is an unattainable phenomenon, what do you do next?

Using models to find the probability of rain is inferior to actually scientifically determining the processes of rain and everything connected to it.

Well, in case of rain, the models are based on the science, obviously. The reason that they aren't more successful is that weather patterns are impossibly complicated.

I've stated clearly the fallacies of statistics.

No, you've articulated some of their weaknesses.

But I also tend to gravitate around certain groups which have a boatload of ideas and/or valid pertinent questions.

Except that the questions are fundamentally unanswerable. Given that the future can't be predicted with certainty, how do you make strategic decisions about anything?

Scientific processes described numerically or as a numeric process are not the same as a man made one. One is real. The other cannot be. Because if man discovers a pattern that is scientific, it is real process that has always been there....a person creating a model, or gauging probability, is just guessing. The two are not the same. Model makers never guess right. They never will.

Actually you are wrong here. I can provide an example. In March 2011, a major earthquake occurred off the coast of Japan. Scientific models had predicted that, after an earthquake of that magnitude, a tsunami was quite likely. So officials in Japan sounded alarms, and the human losses, although devastating, were less than they could have been.

Now yes, you can say, that was a physical process, and so it falls into the realm of science. But in fact there had been no previous record of an earthquake of that size in that location, so the only reason to suspect a tsunami was that models, based on past observations of similar situations as well as the science involved, predicted one.

You want to draw the line at making predictions regarding human behavior. I'm okay with that. But then why can't you just say so? Why do you make me go through all of the work decoding your prose?

Because the profession is not being used to discover decaying nuclei type processes, it's being used to guess.

Worse still, is just about everyone takes it as fact. The proprietors of the guesswork, aren't doing their job at explaining the disclaimer

I mainly agree with that. Unfortunately, we are dealing with a population that can't even accept the overwhelming scientific evidence for things like evolution and global warming, so you and I both know that they can't possibly appreciate the minutiae of confidence intervals in statistical models that aspire to predict human behavior.
 

Bufalay

ASFN Lifer
Joined
Jul 21, 2006
Posts
4,679
Reaction score
786
Actually I made the point quite clear. 0 or 1. Probability, confidence intervals, correlation, never actually hit 1, and no matter how high they go up it's still a .00000000001 or a .999999999 or somewhere in between or even further out in the fractions of a number. They never actually hit 0 or 1. This is reality. Which means in reality, it never hits anything. This is quite simple and easy. They never tell you 0 or 1. Thus they never tell you anything real. If they could tell you 0 or 1, and are as constant and reliable as decaying nuclei, I'd say, hey statistics tell you something. But they don't.

The probability that flipping a fair coin turns up either heads or tails is 1.
 

Mainstreet

Cruisin' Mainstreet
Supporting Member
Joined
Oct 19, 2003
Posts
117,418
Reaction score
57,615
I think that maybe they're both two of a kind,and they don't like looking at themselves.

Funny. :lol:

This discussion reminds me of why I hated my statistic courses in college.
 
OP
OP
3rdside

3rdside

Hall of Famer
Joined
Nov 4, 2002
Posts
1,531
Reaction score
202
Location
London, UK
I would quote you again about writing off Dragic but I can't actually find what you said as my original post got deleted by the Mods. A way too heavy handed approach imo as i wasn't told where it was put (in the admin trash can apparently) and they were wrong about its content. Nowhere did I suggest to rate posters, hence the name of this new thread.

I assumed that was obvious from the context. I'm not sure what probability I'd put on it, maybe 85-90%.

Not quite obvious but assuming that's what you meant, what's a 4yr / $34m player worth in PER terms; 21 points maybe?

PER Definitions
A Year For the Ages: ---------------35.0
Runaway MVP Candidate: -----------30.0
Strong MVP Candidate: -------------27.5
Weak MVP Candidate: --------------25.0
Bona fide All-Star: -----------------22.5
Borderline All-Star: -----------------20.0
Solid 2nd option: -------------------18.0
3rd Banana: ------------------------16.5
Pretty good player: -----------------15.0
In the rotation: ---------------------13.0
Scrounging for minutes: -------------11.0
Definitely renting: --------------------9.0
The Next Stop: DLeague -------------5.0

I don't really know as I'm not that down with salaries but Dragic's PER is already 18. Give him 30+ minutes and is it possible i.e. greater than a 10-15% chance, that he could get to a 21 PER in his first season?

I'd say there is...

Nash as a precedent does not support any argument for any other player.

...and I think this statement could be wrong (give it a year) if you consider the following:

This is going to be Dragic's first season as a starter. Nash didn't hit big time numbers till his first full season as starter in his 5th season when he started 70 games; prior to that he hadn't scored higher than 8.6 points a game, in his fourth season, when he started 27 games.

Dragic hit 11.7 points last season, his 5th season, and he also started 27 games. In his four seasons before that he started 8 games combined whereas Nash in his first four seasons started 78.

I am definitely not saying "Nash did this in his first five seasons therefore Dragic will do the same" but I am saying "based on Nash's first five seasons, if you give Dragic some time in the starting role while also taking into consideration his growth as a player over his first five seasons he is > 10-15% likely to hit a 21 PER". Maybe 15-20%.

But it's just guess work as you say (in the absence of going through all the point guards in the history of the NBA and working out what the probability might actually be of course).

And you're more likely to be perceived as right on the basis that you post more, talk more confidently and have earned a reputation (the latter two points no doubt based on the time you’ve spent watching / reading / understanding the game)...

....Which is why I proposed the spreadsheet:

If I state that Dragic will hit a 20 PER in his first year and note his score at the end of the season there will be a + / - on it. And if we do this for any player (the more contentious ones I guess) or the team - number of team wins / finishing position in the conference / number of games in a playoff series etc, then in 10 years time we're going to have a pretty good idea at who really is good at predicting stuff and who isn't; which should give weight (or take it away) to people who make bold statements (or don't speak up enough).

It may just be that, in fact, guys like yourself and other knowledgeable posters are the best at predicting as you would appear to a) post the most and / or b) post the best but it would be cool to see if there’s a correlation.

And the real skill will come (and subsequent massive mis-predictions) in predicting a year like this year or when we went 62-20 in Nash’s first year.

The stats will probably (ha) be interesting.
 
Last edited:

elindholm

edited for content
Joined
Sep 14, 2002
Posts
27,447
Reaction score
9,604
Location
L.A. area
Just for the record, I don't endorse the whole "post count equals clout" thing. Everyone was new to the board once, and having accumulated a lot of posts doesn't, by itself, mean anything.

If you want to keep a spreadsheet of predictions and their accuracy, go for it. I agree that it might be interesting, but it sounds like a lot of work for (potentially) not much payoff.

As for Dragic, though, a PER of 21 would put him (to look at last year's stats) in the class of Josh Smith, Harden, Pau Gasol, and Deron Williams. I think that's setting the bar very high, probably unfairly so -- if he does get to 21, and doesn't have some glaring disqualifier (e.g. played less than 1500 minutes, or was an abomination defensively, or what have you), then that will be quite an achievement. If Dragic posts a 21 next year, I will happily say that I was wrong about the value of his contract.

Let's go with 19 as a benchmark, which is the territory of Hibbert, Ibaka, Bosh, and Thaddeus Young. 18 or less and he's a disappointment, 20 or more and he overachieved, ballpark 19 and we're getting about what we're paying for. Does that seem fair?
 
OP
OP
3rdside

3rdside

Hall of Famer
Joined
Nov 4, 2002
Posts
1,531
Reaction score
202
Location
London, UK
i'd be up for doing it if others were interested even though it is too much work to do it for everyone. If we could get the main posters to agree to providing predictions - maybe 10 or so - then i'd manage the spreadsheet side of things.

Good win for the USA tonight btw - gold surely on its way on the back of that.
 

elindholm

edited for content
Joined
Sep 14, 2002
Posts
27,447
Reaction score
9,604
Location
L.A. area
Bump: As of Mo. 12/10, Dragic's PER is 20, which means that he has (modestly) beaten my expectations so far.
 
Last edited:
OP
OP
3rdside

3rdside

Hall of Famer
Joined
Nov 4, 2002
Posts
1,531
Reaction score
202
Location
London, UK
Saw that, it was nearer 21 before the 7 game skid. Good to see that my interpretation of statistics lead to me making a decent judgement call - seeing as I haven't seen Dragic play ever (live in the UK - no NBA over here) does that put paid to cardsfan's view that statistics can't tell you anything meaningful?! (cf: it's freaking scary listening to what you have to say and how you say it - it reminds me of listening to Creationists - if your lecturer at ASU was any good (assuming he's the guy that led you into this strange way of thinking) he wouldn't be at ASU, he'd be at Harvard; he's missing credibility in other words.)
 

SunsTzu

ASFN Lifer
Joined
Aug 28, 2003
Posts
4,866
Reaction score
1,674
Let's go with 19 as a benchmark, which is the territory of Hibbert, Ibaka, Bosh, and Thaddeus Young. 18 or less and he's a disappointment, 20 or more and he overachieved, ballpark 19 and we're getting about what we're paying for. Does that seem fair?

I don't really like using PER much, but if you're going to use it to determine value then it should be pointed out that of those 4 player 3 are making near max to max contracts and even Young is making significantly more than Dragic.
 

elindholm

edited for content
Joined
Sep 14, 2002
Posts
27,447
Reaction score
9,604
Location
L.A. area
I don't really like using PER much, but if you're going to use it to determine value then it should be pointed out that of those 4 player 3 are making near max to max contracts and even Young is making significantly more than Dragic.

I think that, if you read the thread, you'll better understand what the discussion was about.
 

Forum statistics

Threads
552,766
Posts
5,402,810
Members
6,313
Latest member
50 year card fan
Top