Moderated Coin Flipper

[Leumas]

Is it random or does it become deterministic?

Not the question from the opening post. Please leave the goal posts in place.

Thank you so very much for the above demonstration

It is an illustration of the running theme in all your posts in this thread... namely being arrantly and irrefragably wrong.

... A single coin toss produces an unpredictable result. But we can predict the approximate results of ten thousand coin tosses. Now, is this random?

The above post led me to write a little WebApp to play with to see the result of coin tosses varying from 10 at a time to 10,000,000 at a time...

 

[jsfisher]

Is it random or does it become deterministic?

Not the question from the opening post. Please leave the goal posts in place.

Yep, here is the relevant part of the opening post: "we can predict the approximate results of ten thousand coin tosses."

Thank you so very much for the above demonstration

It is an illustration of the running theme in all your posts in this thread... namely being arrantly and irrefragably wrong.

... A single coin toss produces an unpredictable result. But we can predict the approximate results of ten thousand coin tosses. Now, is this random?

That would be an attempt to divert attention, there, Leumas. This thread was all about your attempt to construct a computer program to demonstrate "we can predict the approximate results of ten thousand coin tosses".

This is the Computers and the Internet sub-forum, right? So, the topic relates to your program (and others like it) and whether the base statement is supported or not.

Whether "is this random?" is not a question for this sub-forum. It would be best discussed in the original thread in which acbytesla posted..

Leumas, you aren't trying to post off-topic in this thread, are you?

 

[jsfisher]

By the way, Leumas, your attempts to shift the discussion to the "is it random?" diversion makes me confident that you are without competent objection to the affirmative answer to the original question, "we can predict the approximate results of ten thousand coin tosses".

 

[psionl0]

Is it random or does it become deterministic?

For all practical purposes, a coin toss is always a random event so you will never get a deterministic result no matter how many times you toss.

You are demanding proof for something that nobody is claiming.

 

[jt512]

Is it random or does it become deterministic?

The only thing that determines whether the next outcome in a sequence is random or deterministic is whether it was randomly or deterministically generated. If it was truly randomly generated, then it will be random; otherwise, it will be deterministic. Every version of your silly coin-flipper, including the so-called "true" one, generates its output deterministically, because every version, including the "true" one, employes a PRNG.

The probability of the nth outcome being a "head" of a truly randomly generated of sequence coin-flip-like events is 0.5; the probability of the nth outcome being a "head" for a pseudo-randomly generated such sequence is either exactly 0 or 1, because the outcome has been determined in advance with certainty by the algorithm.

 

[Leumas]

For all practical purposes, a coin toss is always a random event so you will never get a deterministic result no matter how many times you toss.

Yet again you demonstrate that you are an upstanding chap.

You are demanding proof for something that nobody is claiming.

I did not demand it... it was asked and this is why this whole thread and Coin Flipper V4.1 are here because of this...

...A single coin toss produces an unpredictable result. But we can predict the approximate results of ten thousand coin tosses. Now, is this random?

 

[psionl0]

I did not demand it... it was asked and this is why this whole thread and Coin Flipper V4.1 are here because of this...

A single coin toss produces an unpredictable result. But we can predict the approximate results of ten thousand coin tosses. Now, is this random?

I guess it depends on how you are interpreting "approximately. If by that he means that you can predict the relative frequency of heads to within a couple of standard deviations of the mean 96% of the time then acbytesla is spot on.

Even if he means something different, I still don't see how that calls for a proof that a large number of coin tosses can yield a totally deterministic result.

 

[Wudang]

Not so much a gish gallop as a gigo gallop.

 

[Leumas]

Not so much a gish gallop as a gigo gallop.

Have you finished reading the book you linked to... or was it "gigo gallop"?

..
For a useful javascript book (free as in beer) try https://books.goalkicker.com/JavaScriptBook/ which is built from highest-rated items on Stackoverflow. I couldn't be bothered looking any deeper.

When are you going to

And now that you have hacked my code and can see what it can do... why don't you make an app that can prove mine inferior or ... better still... that can demonstrate how your superior RNG can prove the mythical wished upon DETERMINISTIC "convergence" to 50%??? Come on... you read the book... no??

So you should now be able to demonstrate your superior knowledge and superior RNG and prove once and for all that my app is claptrap and DEBUNK its EMPIRICAL DATA that rives asunder the mythical deterministic "convergence" wished upon.

Or better still ... can you specify the deterministic equation n=f(p,ε) mentioned in this post and use f(p,ε) to calculate n for p=1 and ε=0??

Myriad could not do it... can you?

 

[Leumas]

I guess it depends on how you are interpreting "approximately.

Yes... but... regardless... it is still random whatever "approximately" means.

Even if he means something different, I still don't see how that calls for a proof that a large number of coin tosses can yield a totally deterministic result.

Exactly...

See the post below about Coin Flipper V5 and the nifty little Coin Flipper Game and the snazzy new graphing feature with zooming and panning and placing the mouse on a graph point to read its values.

 

[Leumas]

The Continuing Quest For Empirical Data

Here is the Coin Flipper Game

It is a game that follows the outline the Empirical Experiment described in this post.

But... it is made as a gambling game. Once in the game page... press on the tools icon on the top left of the screen to see brief instructions.

You can play it from within new version of Coin Flipper V5

[IMGW=378]http://godisadeadbeatdad.com/CoinFlipperImages/CoinFlipper5.png[/IMGW] [IMGW=350]http://godisadeadbeatdad.com/CoinFlipperImages/CoinFlipper5_Game.png[/IMGW]​

Coin Flipper V5 also has a new feature where the graphs are generated and drawn in the app instead of having to take the data to another app to draw the graphs. The facility to copy the data to the clipboard and take it to another app is still available too.

The graphs can be zoomed and panned and even saved as an image to the downloads folder, using the icons above each graph.

[IMGW=350]http://godisadeadbeatdad.com/CoinFlipperImages/CoinFlipper5_Graphing.png[/IMGW]

[IMGW=450]http://godisadeadbeatdad.com/CoinFlipperImages/CoinFlipper5_Graphs.png[/IMGW]​

 

[jsfisher]

I have stumbled across this article. I had been giving Math.random() too much credit, it seems. From what I have now read, it is inappropriate to use Math.random() as it is being used for sequences longer than about 16,000 values.

The coin flip application examples with sequences of a billion or more values is far beyond appropriate for Math.random().

The CSPRNG method may be appropriate (don't know) while the claimed "true" random number generator implementation is complete garbage.

All this just goes to reinforce the point raised much earlier that Leumas' application demonstrates the quality of the random number generator he's using and not the behavior of an ideal coin being repeatedly tossed.

 

[Leumas]

I have stumbled across <snip an article from 2015 CE>

[imgw=400]http://www.internationalskeptics.com/forums/imagehosting/thum_5128262cfa996b2c50.jpg[/imgw]​

In my reckoning Empirical Data will always trump any hand waving bare assertions.

Having faith in mathematical 300 years old theorems being what one CONSTRUES them to mean when they do not mean anything of the sort is fine and dandy.
....

So

come on... get the 10,000 coins and start tossing.​

Or

get cracking on calculating n= f(p=1,ε=0)... or get a math priest to divine it for you.​

Verify for yourself using 300 years old math that no one here can specify so far (let's hope one day soon)... or use EMPIRICAL EXPERIMENTATION to show that it is indeed random (i.e. probability < 100%) that guessing the H_g for 10,000 coin tosses will be determinable with no random error.

What is that you say... doing all that is prohibitive in time and cost and physical efforts?

Ok... you can use Coin Flipper V5 instead to do all that tossing and tallying and displaying of the results for you... you can now easily obtain EMPIRICAL DATA for yourself.

What is that you say.... you are concerned about PRNG and TRNG and whatnot and whatchamacallit... then go do it physically or figure out the 300 years old math formula above....
Failing to do any of the above 3 options but then handwaving this or that is nothing but an Ipse dixit fallacy.

...

 

[jsfisher]

[ imgw=400]http://www.internationalskeptics.com/forums/imagehosting/thum_5128262cfa996b2c50.jpg[ /imgw]​

Was there a point you wanted to make, or was pearl-clutching the best you have?

Your app is at least two-thirds broken. There has been evidence to support that statement. Do you have anything like that? You know, evidence?

 

[jsfisher]

I have stumbled across <snip an article from 2015 CE>

Yes? Is there some point to your observation the article I cited was from 2015? Did the implementation of Math.random() change since then in some significant way? You highlighted your edit as if were somehow significant, but provided no reasoning.

The more likely reason you lead with a vacuous assault was that I provided evidence and you have nothing but hollow assertion.

You app is broken, but even unbroken is could not provide empirical data of anything. It attempts to (poorly) implement a coin-flip model, so all it does is report data from that model.

If you can address the substance of my post, then please do so. Otherwise, just give it a rest, please.

 

[Leumas]

Very early in the text, The Art of Programming, Volume 2, < published 1965 CE>

I have stumbled across <an article from 2015CE>....

Seriously... you were just walking along and it just popped up?

Are you sure you did not look high and low for something that you can get CONCERNED about?

Why did you not "stumble" upon this 2018 article instead of one from 2015... but at least it is better than the book from 1965 CE that you tried to use earlier to denounce PRNGs altogether.

Yes? Is there some point to your observation the article I cited was from 2015? Did the implementation of Math.random() change since then in some significant way? You highlighted your edit as if were somehow significant, but provided no reasoning.

Your app is at least two-thirds broken. There has been evidence to support that statement. Do you have anything like that? You know, evidence?

Thanks yet again for the above that demonstrates the running theme in all your posts in this thread... namely being arrantly wrong...

You might notice that this article shows how the article you happened to stumble upon is OUTDATED and obsolete.... QED!!!

...The important thing to know about all this is that (1) browsers decide which algorithm they want to use to calculate Math.random() and (2) in 2015 pretty much every browser (the major ones at least) ditched their old PRNG algorithms and now they all use the same one: called xorshift128+.

As it turns out, xorshift128+ does a significantly better job at being pretend-random than the older algorithms; plus it’s extremely light weight and computationally fast. So it was adopted pretty much across the board, which speaks volumes to its effectiveness when you consider that there had previously been a lot of differing opinions on the matter.
...
Conclusion (tl;dr)
To package everything up neatly, here’s an overview.

Question: how does JavaScript’s Math.random() generate random numbers?

Answer:

JS doesn’t do anything, it’s up to the browser

As of 2015, most browsers use an algorithm called xorshift128+

The numbers generated by xorshift128+ aren’t really random, the sequence just take a long time to repeat and they’re relatively evenly distributed over the expected range of values.

So, as it turns out, all we are really doing here is taking some input, swirling it around with some math, and spitting out a result. A completely predictable, nonrandom process. But one that feels random enough to us that it serves its purposes as our casual source of chaos in JavaScript.

 

[jt512]

Coin Flipper V5 also has a new feature where the graphs are generated and drawn in the app instead of having to take the data to another app to draw the graphs. The facility to copy the data to the clipboard and take it to another app is still available too.

The graphs can be zoomed and panned and even saved as an image to the downloads folder, using the icons above each graph.

Does it automatically zoom in on short-term random diverging subsequences, or does that still have to be done manually?

 

[jt512]

Why did you not "stumble" upon this 2018 article instead of one from 2015... but at least it is better than the book from 1965 CE that you tried to use earlier to denounce PRNGs altogether.

The question you've been asking about the coin flips your app generates is answered in that article in one short sentence:

"For starters, it’s not really random."

 

[jsfisher]

Seriously... you were just walking along and it just popped up?

It just popped up.

Are you sure you did not look high and low for something that you can get CONCERNED about?

Random number generation is an interesting topic, so I pursue it. You are beneath any CONCERNs I have despite your concern it be all about you.

You also seem to want to simply assume everything is fine with your app when close inspection says it isn't.

Why did you not "stumble" upon this 2018 article instead of one from 2015... but at least it is better than the book from 1965 CE that you tried to use earlier to denounce PRNGs altogether.

Well, for one, that particular article is not very informative. Are you seeing something there to contradict that, or do you just have a hangup on dates?

If dates matter so much, should we be concerned you are using a programming language dating from 1996??

Thanks yet again for the above that demonstrates the running theme in all your posts in this thread... namely being arrantly wrong...

You might notice that this article shows how the article you happened to stumble upon is OUTDATED and obsolete.... QED!!!

As you might notice if you actually read these articles that the article I cited does in fact refer to the current implementation of Math.random() found it all the mainstream implementations of javascript.

My article is neither outdated nor obsolete. Besides, your app is still not providing any empirical evidence of anything. That you fail to comprehend that is disturbing.

 

[Leumas]

...
My article is neither outdated nor obsolete. Besides, your app is still not providing any empirical evidence of anything. That you fail to comprehend that is disturbing.

The question you've been asking about the coin flips your app generates is answered in that article in one short sentence:

"For starters, it’s not really random."