Message from @MOS MAIORVM (AKA Sola)

Discord ID: 311376152759369728


2017-05-09 05:33:56 UTC  

That are not true. You see this a lot in recent high energy physics

2017-05-09 05:34:00 UTC  

And astronomy

2017-05-09 05:34:14 UTC  

Where there is a surfiet of data

2017-05-09 05:34:37 UTC  

And papers being written every day based on it that will turn out to be spurious

2017-05-09 05:34:43 UTC  

The vast majority

2017-05-09 05:34:48 UTC  

Actually it would give you less ability to construct viable statisticial models because it allots for more variation

2017-05-09 05:35:11 UTC  

Viable as in viable seeming

2017-05-09 05:35:30 UTC  

I don't see how that changes anything

2017-05-09 05:35:44 UTC  

More data = more possibility for deviation of an imagined pattern

2017-05-09 05:35:54 UTC  

My point isnt the things you said dont matter

2017-05-09 05:36:07 UTC  

This helps remove human bias when theories don't match up

2017-05-09 05:36:31 UTC  

I respectfully think most of the scientific community disagrees

2017-05-09 05:36:38 UTC  

But you're free to your opinion

2017-05-09 05:36:44 UTC  

More data is not necessarily more possibility for deviation

2017-05-09 05:36:54 UTC  

But it is

2017-05-09 05:36:58 UTC  

In an assumption

2017-05-09 05:37:02 UTC  

That's simply fact

2017-05-09 05:37:04 UTC  

Not at all

2017-05-09 05:37:20 UTC  

Maybe you get lucky and what you study has a narrow spread

2017-05-09 05:37:36 UTC  

That's not about possibility

2017-05-09 05:37:39 UTC  

you're giving a specific example

2017-05-09 05:37:43 UTC  

More 'resolution' in a wide field

2017-05-09 05:37:50 UTC  

Doesnt help you

2017-05-09 05:37:59 UTC  

When you have no framework for the field

2017-05-09 05:38:07 UTC  

The framework comes first

2017-05-09 05:38:14 UTC  

That's very generalized

2017-05-09 05:38:18 UTC  

my point still stands though

2017-05-09 05:38:23 UTC  

Its a general discussion no

2017-05-09 05:38:48 UTC  

The more data you have, the more possibility for variation, compared to less data.

2017-05-09 05:39:14 UTC  

that's a very basic concept

2017-05-09 05:39:25 UTC  

Huh?

2017-05-09 05:39:43 UTC  

I thought you said more deviation from an imagined pattern

2017-05-09 05:39:47 UTC  

Not more variation

2017-05-09 05:39:59 UTC  

More variation helps my argument surely

2017-05-09 05:40:12 UTC  

If I flip a coin one time, I will likely never have the variation that I will find if I flip it 10 times.

2017-05-09 05:40:47 UTC  

But if you toss it 100,000 times

2017-05-09 05:40:57 UTC  

You wont be better off if you do it 10 mil

2017-05-09 05:41:08 UTC  

"Actually it would give you less ability to construct viable statisticial models because it allots for more variation"

2017-05-09 05:41:14 UTC  

is what I said

2017-05-09 05:41:33 UTC  

No?

2017-05-09 05:41:43 UTC  

No?