<< Back to Off-topic Forum   Search

Posts 31 - 50 of 63   <<Prev   1  2  3  4  Next >>   
a humble theory: 12/10/2016 14:56:49

Japanball
Level 56
Report
HTF is that not racist?
a humble theory: 12/10/2016 15:10:40


God Emperor 
Level 57
Report
a humble theory: 12/10/2016 15:13:20

Japanball
Level 56
Report
Why not start using indigenous peoples as bullet fodder?
a humble theory: 12/10/2016 16:50:00

Japanball
Level 56
Report
Why indigenous people failed:
1.Bullet fodder
2.Bullet fodder
3.Bullet fodder
a humble theory: 12/10/2016 18:00:43


OnlyThePie
Level 54
Report
I mean, I'm awful at this game, but I hate Trump. So your position is fatally flawed.
a humble theory: 12/10/2016 18:23:07


Huitzilopochtli 
Level 57
Report
Tabby, I reported you for being so racist. Especially the "Failure to have a sufficiently advanced civilization" comment which is plain ignorant.
a humble theory: 12/10/2016 20:40:04


Math Wolf 
Level 64
Report
There we go again, Wally Balls abusing statistics and thinking he understands something.

Let's first get this out of the way:
Basic statistics will tell you if A is correlated with B and B is correlated with C, then A DOES NOT NEED TO BE CORRELATED WITH C. Shall I put it in bold next time to get the message across?

Next, in your example:
(1) Being good at Warlight is correlated with intelligence (Probably correct with assumingly a causal link from intelligence to being good at Warlight)
(2) Intelligence is correlated with education level (Proven to be correct however more due to other causal sources like wealth of the parents which is correlated with both)
(3) Education level is correlated with Clinton support (Assumingly correct based on both pre-election polls as well as exit polls, but unclear at this point what is the causal link)

What does this prove? Nothing. There is no direct evidence that IQ is correlated (negatively) with Trump support, let alone that being good at Warlight is correlated with Trump support, another step down the line.

Additionally, it is incredibly funny (not) that you post a link to the correlation implies causation wikipedia page as you yourself are blabbering about correlation while making causal statements.

Wally, deal with it: there is no proof. While what you say may be true, there is absolutely and utterly no proof for it and there won't be for years to come. Take my advice and walk away from this and your other ******* threads. By showing how badly you understand the statistics yourself, you are indeed a counterexample to each of the claims you are making as awesomeguy points out.

(And in case you weren't paying attention, I'm neither a Trump nor a Clinton supporter.)

Edited 12/10/2016 20:40:40
a humble theory: 12/10/2016 21:02:06


Leibstandarte (Vengeance)
Level 45
Report
You really are a math wolf.
a humble theory: 12/10/2016 22:35:38


Wally Balls 
Level 59
Report
When you say there is absolutly a correlation between IQ and education, you lost any crediblity you might have had with me.


There is obviously a correlation between IQ and education.

I stand by my theory that there is a correlation between intelligence and how good one is at Warlight, despite you clearly being an exception.
a humble theory: 12/10/2016 22:39:48


Belgian Gentleman
Level 57
Report
Guys we need to make more productive content. Meow!
a humble theory: 12/10/2016 22:40:14


Wally Balls 
Level 59
Report
Wally, deal with it: there is no proof.


I never said there is any scientific method compliant proof. My pointing out a series of strong correlations is not my attempt at scientific method compliant proof of anything nor claiming I have scientific method compliant proof of anything.

you yourself are blabbering about correlation while making causal statements.


I never made any causal statements. I'm well aware of the fact that correlation isn't scientific method compliant proof. What you're seemingly unaware of is that a lack of scientific method compliant proof of something does not prove it wrong nor prove it's opposite.

I have no scientific method compliant proof that you are a human. Just a series of strong correlations. Does that mean you are not a human?

Edited 12/10/2016 23:04:26
a humble theory: 12/10/2016 22:42:59


Wally Balls 
Level 59
Report
*7/7

Edited 12/10/2016 22:51:16
a humble theory: 12/10/2016 22:59:04


Belgian Gentleman
Level 57
Report
7/8 if Wally Balls were included

Edited 12/10/2016 22:59:40
a humble theory: 12/10/2016 23:00:44


Wally Balls 
Level 59
Report
i am not a trump supporter
a humble theory: 12/10/2016 23:01:35


Belgian Gentleman
Level 57
Report
I know that's why it isn't 8/8
a humble theory: 12/10/2016 23:33:30


Swisster 
Level 64
Report
So just so I am clear to what you are saying; a persons high ability to problem solve and learn (IQ) will absolulty lead to an education?
a humble theory: 12/10/2016 23:38:20


Math Wolf 
Level 64
Report
Wally, deal with it: there is no proof.

I never said there is any scientific method compliant proof. My pointing out a series of strong correlations is not my attempt at scientific method compliant proof of anything nor claiming I have scientific method compliant proof of anything.
you yourself are blabbering about correlation while making causal statements.


I never made any causal statements. I'm well aware of the fact that correlation isn't scientific method compliant proof. What you're seemingly unaware of is that a lack of scientific method compliant proof of something does not prove it wrong nor prove it's opposite.

I have no scientific method compliant proof that you are a human. Just a series of strong correlations. Does that mean you are not a human?

I actually agree with your first point and I'll even give you that you are correct that you didn't make any causal statements in this topic. If you want to keep arguing about this point though, I'd gladly dig up some statements from previous topics.

You may want to think the rest of your statements through though. Do you really think that of all people I don't understand that "lack of scientific method compliant proof of something does not prove it wrong nor prove it's opposite."
To quote from my original post: "While what you say may be true, there is absolutely and utterly no proof for it". I'm teaching this kind of material and more advanced versions to both undergrad students and grad students about ten to fifteen hours each week, you may want to think about that before assuming that I don't know what I'm talking about or don't understand a basic statistical concept.
Note that I also said that such proof won't be there for years. I do believe, that if your statements are true, then extensive mediation models may (and should) be able to actually quantify it given a large enough dataset of sufficient quality. But such research takes time. And you'd do well to wonder how come that I know that this kind of research takes time.
a humble theory: 12/10/2016 23:41:23


Wally Balls 
Level 59
Report
"While what you say may be true, there is absolutely and utterly no proof for it".


We agree.

You're yet another of those people who know I'm right but want to come in and argue with me anyway because you don't like the way that I communicate that I'm right. I don't care if you don't like the cut of my jib so long as we agree that I'm right. And we do.
a humble theory: 12/11/2016 01:13:09


Math Wolf 
Level 64
Report
"may be true"

My whole point is about that it is not known. It could be correct, it could be wrong. You believe it is correct, while I say that it's too early to conclude anything.

Early evidence seems to support your claim. However, early evidence also predicted a Clinton win. Limited data could be fool's gold. It may be that Trump supporters actually had a higher IQ. How could this be, given that Clinton won the more educated voters one might ask?

As an example, what happens to this interpretation if we add age as an additional confounder? It may turn out that more intelligent older voters chose Trump, while less intelligent older voters chose Clinton (and there are actually some indications that this is the case!) However, older people also tend to be a lot less educated than younger people, so the more intelligent older people who voted for Trump in this hypothetical example, wouldn't show up when we only consider education. Now, on top of age, we could also add rural voters versus urban voters. Urban voters are more educated and went more for Clinton while rural voters are less educated and went more for Trump. Yet, it may be that the more intelligent rural voters (who aren't very educated either) went more for Trump while the less than averagely intelligent urban voters (who are still largely educated) went for Clinton. Again, you'd get a result where educated voters split for Clinton while intelligent voters actually split for Trump.

What should this teach us (and I mean everyone, not just you): it means that at this point we're only seeing the tip of the iceberg. Demographics are incredibly complicated, polls are flawed and data incomplete. On top of that, the statistical models that are able to properly capture these trends are way too often ignored for easier, biased comparisons. The result is that the real association is very difficult to find, let alone prove. Sadly enough the field of statistics, when used well, is especially very good at telling people why claims CANNOT be made.

That doesn't mean you are wrong of course, it simply means that we'll have to wait at least a few months, but most likely many years before we can conclude that you were right, or wrong.

And the only reason why I put in the effort to post these kind of things, is because you (ab)use my field of study to make claims while waiving an air of knowledge around you. I don't bother to respond to most of the pro-Trump posts here because they are mostly just parroting stuff and not making claims using proper statistical or scientific language.
At the same time, I feel that you may actually have the capability of understanding these concepts and that makes it extra sad that you abuse them.

Edited 12/11/2016 01:13:26
a humble theory: 12/11/2016 01:31:29


Lordi
Level 59
Report
Wally's method: count all cases pointing into the right direction and ignore the rest. Let's just agree to 100/100 and save him the trouble.
Posts 31 - 50 of 63   <<Prev   1  2  3  4  Next >>