• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Coin flip not random

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Originally posted by: Arcadio
I knew that. That's why I use random.org every time I need a random event instead of a coin flip.
Click to expand...

What if you're not near a computer? Oh wait, this is ATOT...nevermind.
 
That was a rather interesting research paper. I think the 51% claim is taken a little out of context.
Conclusion:
Despite these important caveats we consider the bias we have
found fascinating. The discussion also highlights the true difficulty of carefully studying
random phenomena. If we have this much trouble analyzing a common coin
toss, the reader can imagine the difficulty we have with interpreting typical stochastic
assumptions in an econometric analysis.
The caveats and analysis also point to the following conclusion: Keller?s analysis
gives a good approximation for tossed coins. To detect the departures of the order of
magnitude we have found would require 250,000 tosses. The classical assumptions of
independence with probability 1/2 are pretty solid.
Click to expand...

You can read the entire paper here:
http://www.stat.berkeley.edu/u.../diaconis_coinbias.pdf
(You're welcome, for the 5 of you who may actually read it.)
 
Also, for what it's worth, their analysis did not include for the coin to bounce on the table or on a floor. The coin has to be caught.

That's pretty amusing, because I can flip a coin and have it come up heads 100% of the time, or tails 100% of the time, if you're going to let it land in my hand (sucker!)
 
I've come to a science myself flipping coins. Using my technique, I can influence the outcome to about a 75% accuracy.
 
Originally posted by: Fox5
Originally posted by: her209
Originally posted by: mundane
Originally posted by: Arcadio
I knew that. That's why I use random.org every time I need a random event instead of a coin flip.
Click to expand...
Okay, I'll bite. Why would you need a random number in your daily life?
Click to expand...
Fake phone number
Click to expand...

I've been using python's random class to generate a random number whenever I needed one. Generally for choosing maps in multiplayer games.

I use an electronic dice roller (LEDs that light up) that I made for general random decisions throughout my life. It doesn't really give random output though.
Click to expand...
(minor thread-jack)
Any thoughts on this?
Its claim:
This may well be the world's only discreet logic circuit to generate truly random output.
Click to expand...
The 200Hz oscillator samples the phase noise of the 2MHz oscillator yielding a truly random coin-toss output.
Click to expand...

It might become a side project eventually.
I've tried a few commercially-available LED candles, and I've found this:
Philips LED candles don't flicker. They blink. It's very obvious that it's nothing close to a candle.
Various other candles that come close to a flickering behavior seem to have a repeating pattern after a minute or so, similar to cheap white noise generators. Two of them side by side will eventually sync up, much like turn signals which are timed with different frequencies; at certain points in the cycle, they appear to nearly sync up.


I've also heard that a cheap way to create random noise is to take a CCD, enclose it in a completely dark, cold container, and boost its gain. The static generated would then be random, the product of electron vibration or motion, or some such thing.



Originally posted by: Regs
This is like...multiplying any number by 0.
Click to expand...
Uh oh


 
Originally posted by: GodlessAstronomer
Originally posted by: SunSamurai
What the hell? Coins arnt separate from the laws of nature?

But... they are COINS.
Click to expand...

Read the full article, that's not the point. The main conclusion here is that when human beings flip coins, the outcome isn't random. Traditionally we have assumed that the outcome should be essentially random due to chaos theory - small differences in the initial conditions lead to large variations in how the coin behaves. This article shows that this doesn't lead to true randomness.
Click to expand...

Yeah no shit, that's what I was implying. Nothing is truely random. Random cannot exist. Even things like the uncertainty principle only apply to human capability to measure, they do not extend to what actually is.

A toss is only random enough because we have no practical way to determine it before hand. That doesn't mean its not governed by strict laws
 
Funny this topic should come up this week.

I was at the drag races Saturday night and when you and your opponent want the same lane you have to flip a coin for lane choice. I never really give much thought to the whole coin flip process and while he was looking for a quarter I decided since I had called heads the last 3 times and won, I would call tails this time....

He flipped it, it came up heads, and I ended up losing the round.......

randomness fail
 
Originally posted by: moparacer
Funny this topic should come up this week.

I was at the drag races Saturday night and when you and your opponent want the same lane you have to flip a coin for lane choice. I never really give much thought to the whole coin flip process and while he was looking for a quarter I decided since I had called heads the last 3 times and won, I would call tails this time....

He flipped it, it came up heads, and I ended up losing the round.......

randomness fail
Click to expand...

Another victim of the gambler's fallacy.
 
Originally posted by: SunnyD
I've come to a science myself flipping coins. Using my technique, I can influence the outcome to about a 75% accuracy.
Click to expand...

Yeah, I flipped lots of coins in high school and was able to influence the result of the flip significantly by practice. The key is to keep the spinning controlled and precise. I'd say that I could get a coin to land on the same side as the initial side nearly 70% of the time.
 
Originally posted by: SunSamurai
Originally posted by: GodlessAstronomer
Originally posted by: SunSamurai
What the hell? Coins arnt separate from the laws of nature?

But... they are COINS.
Click to expand...

Read the full article, that's not the point. The main conclusion here is that when human beings flip coins, the outcome isn't random. Traditionally we have assumed that the outcome should be essentially random due to chaos theory - small differences in the initial conditions lead to large variations in how the coin behaves. This article shows that this doesn't lead to true randomness.
Click to expand...

Yeah no shit, that's what I was implying. Nothing is truely random. Random cannot exist. Even things like the uncertainty principle only apply to human capability to measure, they do not extend to what actually is.

A toss is only random enough because we have no practical way to determine it before hand. That doesn't mean its not governed by strict laws
Click to expand...

The bolded above depends on how you interpret quantum mechanics. The most widely accepted interpretation (the Copenhagen Interpretation) says that certain events in nature are intrinsically random and the universe is essentially non-deterministic. So far all of the evidence points this way (as opposed to a so-called Hidden Variable Interpretation).
 
(minor thread-jack)
Any thoughts on this?
Its claim:
This may well be the world's only discreet logic circuit to generate truly random output.

The 200Hz oscillator samples the phase noise of the 2MHz oscillator yielding a truly random coin-toss output.


It might become a side project eventually.
I've tried a few commercially-available LED candles, and I've found this:
Philips LED candles don't flicker. They blink. It's very obvious that it's nothing close to a candle.
Various other candles that come close to a flickering behavior seem to have a repeating pattern after a minute or so, similar to cheap white noise generators. Two of them side by side will eventually sync up, much like turn signals which are timed with different frequencies; at certain points in the cycle, they appear to nearly sync up.
Click to expand...

I'm not sure exactly what's going on in that circuit, but I would think what'd you get is a continuously varying waveform that would eventually repeat. But I'm also assuming the oscillators reach steady state, if they are just sampling each other's noise before a stable waveform is created, then it could possibly be truly random (ish).
 
Originally posted by: moparacer
Funny this topic should come up this week.

I was at the drag races Saturday night and when you and your opponent want the same lane you have to flip a coin for lane choice. I never really give much thought to the whole coin flip process and while he was looking for a quarter I decided since I had called heads the last 3 times and won, I would call tails this time....

He flipped it, it came up heads, and I ended up losing the round.......

randomness fail
Click to expand...

Did you think that the 4th flip would somehow be influenced by the first 3 flips?
 
Originally posted by: SunSamurai
Yeah no shit, that's what I was implying. Nothing is truely random. Random cannot exist. Even things like the uncertainty principle only apply to human capability to measure, they do not extend to what actually is.

A toss is only random enough because we have no practical way to determine it before hand. That doesn't mean its not governed by strict laws
Click to expand...

What I find interesting is that the author dragged Taleb into this. If he had actually read Taleb he'd know that randomness is actually a matter of perspective. In a coin flip situation involving a flipper and an observer, the outcome may be predetermined from the flipper's view if he can always flip tails yet purely random from the observer's point of view if he has no method of knowing or determining the flipper's skill.

Even results such as this fall into an attribution error: an outcome not predicted occurred (51% result) and we, as humans, try to find an answer for it, like "skill", or "Human error", or "the side showing up has some advantage" when the real explanation is that the observed result could not have been predicted beforehand and any explanation after the fact is baseless and wrong.
 
All I know is, when we flip a coin over something, it has to hit the ground and bounce around. 😉
 
Originally posted by: Jeff7

I've also heard that a cheap way to create random noise is to take a CCD, enclose it in a completely dark, cold container, and boost its gain. The static generated would then be random, the product of electron vibration or motion, or some such thing.
Click to expand...

That's called thermal noise.
 
Originally posted by: mundane
Originally posted by: Arcadio
I knew that. That's why I use random.org every time I need a random event instead of a coin flip.
Click to expand...

Okay, I'll bite. Why would you need a random number in your daily life?
Click to expand...

Magic 8 Ball says..........Your Mom
 
Originally posted by: SunSamurai
What the hell? Coins arnt separate from the laws of nature?

But... they are COINS.
Click to expand...

My money violates the laws of physics. Outflow >> Inflow 🙁

Always knew about the 51%, but the fact that it's not the weight/density that affects the outcome is new to me.
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top