Is Anything in the Universe Truly Random?

[Avichal]

Well I just read about whether mind can produce random numbers or not here.
But that makes me wonder whether anything in this universe is random or not? I haven't yet studied quantum physics but I know that a lot in it is based on probability.

So is anything truly random in this universe? Also lately I am wondering about whether we live in a deterministic world or not? Whether something is truly random or not seems related to it.
Do we have answers to such questions or are we still searching?

 

[Chi Meson]

Even if we do live in a deterministic world, a principle of quantum physics shows us that we cannot know all information about our world to perfect precision. Basic chaos theory will then take over to show that without perfect knowledge of our present "initial conditions," the future will be utterly unpredictable.

Add to the top of that indeterminable pile the probabilistic nature of fundamental particles, where the outcome of any atomic decay, etc, is truly unknowable, and you have, essentially a future that cannot be written, plotted or calculated.

Any speculation on whether we could predetermine random occurrences are just that, speculation. Basically a game of "OK, so we can't; but what if we could?"

If atomic decays, etc, can be shown to follow a predetermined outcome (which we currently believe is not the case, and I have no idea how we could prove that, but I'm not the PhD here) then I'd say there was a case for it.

 

[Maui]

If you toss a coin 1000 times, you'd get an approximately 50/50 distribution of heads/tails, so yes, it seems the randomness is genuine.

 

[Jimmy Snyder]

Half of the pages in a book are odd numbered, and the other half are even numbered. So the pages of the book are randomly numbered.

 

[Maui]

The pages are exactly 50/50 distribution, the coin toss is approximately 50/50 and involves radondmness(or unpredictability). Only a conspiracy can 'explain' the approximately 50/50 distribution of thousands of trials(if true randomness were false) and we know conspiracies are very often wrong and not quite inline with known science.

 

[Avichal]

If you toss a coin 1000 times, you'd get an approximately 50/50 distribution of heads/tails, so yes, it seems the randomness is genuine.

Yeah but doesn't the head or tale depend on the force with which we toss. So if we toss with the same force each time and we know all the factors that are acting upon it we can calculate whether it will be head or tale at each toss. So not really random

 

[Maui]

Yeah but doesn't the head or tale depend on the force with which we toss. So if we toss with the same force each time and we know all the factors that are acting upon it we can calculate whether it will be head or tale at each toss. So not really random

No, i meant something different. Throw the coin against the wall so it hits it and lands afterwards. You can't beat the odds, there is no way around this. There is a whole multi billion dollar entertainment industry exploiting this basic physics phenomenon. The longer you play at a casino the less your odds of winning(in the long run you are always broke).

As for the conspiracy, Occam's razor is usually enough to dispell it.

 

[Jimmy Snyder]

The pages are exactly 50/50 distribution, the coin toss is approximately 50/50 and involves radondmness(or unpredictability).

Sometimes the pages are exactly 50/50, and sometimes not. It is approximately 50/50. Sometimes the coin toss comes out 50/50 and sometimes not. It is approximately 50/50.

 

[Maui]

Sometimes the pages are exactly 50/50, and sometimes not. It is approximately 50/50. Sometimes the coin toss comes out 50/50 and sometimes not. It is approximately 50/50.

I don't understand what you are saying, but in true randomness(as is always assumed in the sciences, also called 'naturalism'), the odds of heads or tails is approximately 50/50 and with the increase in number of trials the odds converge to 50/50.

 

[Jimmy Snyder]

I don't understand what you are saying, but in true randomness(as is always assumed in the sciences, also called 'naturalism'), the odds of heads or tails is approximately 50/50 and with the increase in number of trials the odds converge to 50/50.

January 1st lands on a Monday 1/7 of the time, on Tuesday 1/7, etc. Does this mean that the day January 1st lands on is random? Your description of the coin toss is the same. When you say it comes up heads 1/2 of the time and tails 1/2 of the time, you are merely stating that the distribution is uniform. Randomness is something more. For example, you would also have to prove that the four outcomes hh, ht, th, tt each come out 1/4 of the time and that hhh, hht ... ad infinitum. You could never satisfy yourself that the outcomes were random with only a finite number of tosses. It's even in your own statement about convergence. How can you be sure of convergence with only finitely many tosses?

 

[Maui]

January 1st lands on a Monday 1/7 of the time, on Tuesday 1/7, etc. Does this mean that the day January 1st lands on is random?

No because there is nothing unpredictable in the sequence of days. The sequence of days is orderly, the heads/tails distribution is not(only as a large ensemble of trials). This is exactly what you would expect of a truly random(unmediated) event.

Your description of the coin toss is the same. When you say it comes up heads 1/2 of the time and tails 1/2 of the time, you are merely stating that the distribution is uniform.

Yes, it's uniform IN THE LONG RUN(!), but not in the short(!). This is a cruicial point that you seem to be missing.

Randomness is something more. For example, you would also have to prove that the four outcomes hh, ht, th, tt each come out 1/4 of the time and that hhh, hht ... ad infinitum. You could never satisfy yourself that the outcomes were random with only a finite number of tosses. It's even in your own statement about convergence. How can you be sure of convergence with only finitely many tosses?

Because we've been using this assumption all the time and the observations have been confirming it day after day over thousands of years. Do you believe that casinos could lose large sums of money because they have only been able to test randomness with only a finite number of roulette spins? And say for 1 full year the ball could land on red 75% of the time?

 

[Jimmy Snyder]

I googled
is a coin toss truly random
without quotes. I suggest you try it.

 

[marty1]

Listen to Maui. The short run is the important factor. You flip a coin and know much less about what will happen next than what will happen 1,000 times. Randomness is about not being able to predict the elements of the sequence not long term trends. You can easily guess 50% are heads but will it be hhttttthhhhhththhhtttt?

 

[Jimmy Snyder]

So is anything truly random in this universe?

Because we've been using this assumption all the time and the observations have been confirming it day after day over thousands of years.

Yes Avichal, by assumption.

 

[Chi Meson]

The randomness of the coin toss comes from the impossibility to perfectly replicate the initial conditions of the toss. If a machine, carefully calibrated, tossed the coin with an identical initial force, in precisely the same location, causing exactly the same rotation, and landing spot, then the toss could be repeated with predictable results.

But humans are not capable of repeating such precise initial conditions, so the variation of rotation, trajectory, position of edge when it hits etc are what make the outcome unpredictable, or "random."

The machine I mentioned would need to be of ultra high precision and ultra-slim tolerance just to predict coin tosses, and even then would be unlikely to achieve 100 %.

The rest of the universe is much more intricately interdependent on "initial conditions."

Correct me if I am wrong, but Jimmy's point may be that "random" requires a definition. Random is as random does. Take care how we define it.

 

[SHISHKABOB]

I think that a lot of times something can appear to be random from a certain perspective, even though it is probably not random.

For instance, when I am walking down the street and I look at people, the directions in which they are walking are pretty much random since I don't know anything about them or where they are going.

Of course, it's not random at all: each person (probably) has a very specific place to which they are headed. But from my perspective they might as well be a bunch of molecules in a MB distribution.

 

[Jimmy Snyder]

Correct me if I am wrong, but Jimmy's point may be that "random" requires a definition. Random is as random does. Take care how we define it.

At any rate, randomness is not simply a matter of uniform distribution as indicated in post #3.

 

[marty1]

Is "randomness" a measure of your success/failure rate in predicting what the state of a system will be at a specific future time?

Is a coin toss only 50% random if you always guess heads next?

Random is the deviation from expectation.

The trend of 50% heads is not random. It is what you expect and what you can predict for many coin tosses. The next state, H or T, is random.

 

[Pythagorean]

There are three definitions of random that we encounter.

1) The "laymen"/social version: "Well, that was random!"
This is basically equivalent to unexpected or out of the ordinary

2) There's a vague notion of random being used as an opposite to deterministic. If a process is truly "random" in this usage, it seems to indicate a lack of deterministic causality in a system. I.e. probabilistic (like quantum wave collapse when measuring).

3) Mathematical definition from Wolfram Alpha:

A random number is a number chosen [...] from some specified distribution such that selection of a large set of these numbers reproduces the underlying distribution.

http://mathworld.wolfram.com/RandomNumber.html

----

But your question casts doubt on "true" randomness (in the case of 2), I think, is what you mean). Which, I think, is founded in many cases. Sometimes we have deterministic systems, but they're chaotic and complex, so we just take them to be random when we model them. For instance, noise in a signal we may model with a random function on the computer, but if we were studying noise we may also have a deterministic way of modeling it. If we're measuring signal, we don't care how the noise came about, so the random model is sufficient.

 

[marty1]

3) Mathematical definition from Wolfram Alpha:

http://mathworld.wolfram.com/RandomNumber.html

Wolfram defines random using the term "by chance". Isn't that kind of a self-referencing definition.

 

[arupel]

In the classical sense, nothing is random. Everything has a cause and an effect. Even in the case of the supposedly random roulette wheel, some students and their instructors have determined methods of winning. My belief is that the only randomness occurs is in how we define randomness.

The only seeming major exception to this rule is Chaos Theory, where immeasureably small errors will result in large divergent results at a later time. But even in this case the word is immeasureably small errors. Not perfect knowlege.

Quantum mechanics is a different matter. Basically we can know all the causes perfectly and how the equations work perfectly (perfect knowledge in all senses) and the best that we can get is a probability distribution of the results. Granted that the probability distribution is determined by the above, but it can only be a probabily distibution, not
a determined result.

 

[marty1]

I'll probably get banned for expressing a thought...but...

Regarding cause and effect, an ordered set of effects can be caused by a random mechanism. This is counterintutive to me, but just look at the Universe as proof. A random mechanism could also cause random effects, but in the end it seems that order does prevail. Survival of the stable sequence, I suppose, which is then embedded in the mechanism that produces future effects. It is much more difficult for an ordered mechanism to cause a random result. The best we can do is hide the order.

 

[Jobrag]

Have a look at this http://www.bbc.co.uk/podcasts/series/iots click on Random and Pseudorandom

 

[wuliheron]

Randomness merely refers to something that has no perceivable order so when you ask if something is "truly" random I'd have to say yes, there are plenty of things we don't perceive any order in. I'd go even further and suggest order and chaos are merely terms we use that define each other and the question is a bit like asking if anything has a true shadow. Like the random a shadow in and of itself conveys no energy or information and obeys no known laws of physics. It is like the spaces between the notes in music that gives their existence meaning and definition.

Now if you are asking if randomness is a metaphysical reality or somesuch all I can say is metaphysics can't be proved or disproved by definition and you might as well ask if God exists.

 

[Jimmy]

Have a look at this http://www.bbc.co.uk/podcasts/series/iots click on Random and Pseudorandom

Thanks for the link.

I find it amusing that the BBC Radio player volume goes to 11.

 

[Pythagorean]

I'll probably get banned for expressing a thought...but...

Regarding cause and effect, an ordered set of effects can be caused by a random mechanism. This is counterintutive to me, but just look at the Universe as proof. A random mechanism could also cause random effects, but in the end it seems that order does prevail. Survival of the stable sequence, I suppose, which is then embedded in the mechanism that produces future effects. It is much more difficult for an ordered mechanism to cause a random result. The best we can do is hide the order.

I agree with the first statement, that an ordered set of effects can look like cause and effect and be random, but I disagree with the other statement: "It is much more difficult for an ordered mechanism to cause a random result" (assuming you're still referring to cause and effect by 'ordered mechanism').

For instance, statistical mechanics and thermodynamics is a large ensemble simplification. You could write out the differential equations for the system of particles if you really wanted... but it would be messy.

This is typical in the nonlinear sciences (particularly with respect to chaos). An ensemble of deterministic particles that are all basically the same (in terms of mass, charge, etc) will look random even though they're not. We often choose to model them as random in physics/biology because it would be too much work for something we're not interested into model all the motion. We can just use a random function to loosely model it and then focus on the signal of interest for detailed modeling. Some might even argue that there is no real random in classical physics; that 'random' is the easiest way to describe dynamics we don't understand.

 

[marty1]

This is typical in the nonlinear sciences (particularly with respect to chaos). An ensemble of deterministic particles that are all basically the same (in terms of mass, charge, etc) will look random even though they're not.

That was my point. Those are not truly random things even though they look random or we imagine they are for the simplicity of calculation.

"True" randomness cannot be generated from a system that has at its core an ordered mechanism for causing events. Get down to the quantum scale, however, and "true" random events can occur. I think there have been enough years of quantum experiments to confirm Feynman's statement “There are no 'wheels and gears' beneath this analysis of Nature.”

At the macro scale the best we can do is hide the order. At the macro scale there is an extremely orderly mechanism underlying cause and effect.

Even though all of this order is the product of a truly random (perfectly random) mechanism, the opposite is impossible.

 

[wuliheron]

That was my point. Those are not truly random things even though they look random or we imagine they are for the simplicity of calculation.

"True" randomness cannot be generated from a system that has at its core an ordered mechanism for causing events. Get down to the quantum scale, however, and "true" random events can occur. I think there have been enough years of quantum experiments to confirm Feynman's statement “There are no 'wheels and gears' beneath this analysis of Nature.”

At the macro scale the best we can do is hide the order. At the macro scale there is an extremely orderly mechanism underlying cause and effect.

Even though all of this order is the product of a truly random (perfectly random) mechanism, the opposite is impossible.

Which again is just another denial of the fact all that random refers to anything has no perceivable order.

Random
proceeding, made, or occurring without definite aim, reason, or pattern: the random selection of numbers. Dictionary.com

lacking a definite plan, purpose, or pattern Mirriam-Webster

Having no definite aim or purpose Oxford

 

[marty1]

Which again is just another denial of the fact all that random refers to anything has no perceivable order.

Random
proceeding, made, or occurring without definite aim, reason, or pattern: the random selection of numbers. Dictionary.com

lacking a definite plan, purpose, or pattern Mirriam-Webster

Having no definite aim or purpose Oxford

An unsolved differential equation.

 

[Pythagorean]

That was my point. Those are not truly random things even though they look random or we imagine they are for the simplicity of calculation.

"True" randomness cannot be generated from a system that has at its core an ordered mechanism for causing events. Get down to the quantum scale, however, and "true" random events can occur.

Well, first it depends on your definition of "True" randomness (I agree that Wolfram's definition is not sufficient, though I think part of it is necessary).

But, more importantly, I'm pretty sure quantum chaos is still a very open subject. We have very little knowledge about open quantum systems our large-scale coherence. Deterministic chaos and quantum probability must be reconciled somehow (some might point to the deficiency in the definitions of space and time, as they manifest at the Planck scale, but others point out the significance of so-called "subplanck" structures at these scales:

http://www.ncbi.nlm.nih.gov/pubmed/11507634

You seem to imply that:

quantum probability --> deterministic chaos

But I don't think there's any evidence for that at all. Of course, I could be completely wrong, but I'd be interested to see the evidence.

In my view, these are two huge, complex concepts based off a myriad of human assumptions and in different aspects, both:

quantum probability --> deterministic chaos
deterministic chaos --> quantum probability

are true (or our understanding and language for cause and effect is currently inept)

 

[wuliheron]

An unsolved differential equation.

An unsolved whatever. The idea that nature must obey the preconceptions of scientists is so old and has been disproved so many times you'd think academia would insist by now on everyone receiving a doctorate getting a tattoo that says it just ain't so. A new law of nature, science is invariably proven wrong.

 

[marty1]

Well, first it depends on your definition of "True" randomness (I agree that Wolfram's definition is not sufficient, though I think part of it is necessary).

But, more importantly, I'm pretty sure quantum chaos is still a very open subject. We have very little knowledge about open quantum systems our large-scale coherence. Deterministic chaos and quantum probability must be reconciled somehow (some might point to the deficiency in the definitions of space and time, as they manifest at the Planck scale, but others point out the significance of so-called "subplanck" structures at these scales:

http://www.ncbi.nlm.nih.gov/pubmed/11507634

You seem to imply that:

quantum probability --> deterministic chaos

But I don't think there's any evidence for that at all. Of course, I could be completely wrong, but I'd be interested to see the evidence.

In my view, these are two huge, complex concepts based off a myriad of human assumptions and in different aspects, both:

quantum probability --> deterministic chaos
deterministic chaos --> quantum probability

are true (or our understanding and language for cause and effect is currently inept)

Maybe I'll try to simplify my platform, maybe not...

A mechanism that randomly causes events can at times produce order (by chance). Some of this order may result in other mechanisms that can themselves cause events but the important ones are those that have lasting consequences that reinforce their own persistence. My position is that those larger mechanisms have an abillity to produce random results that diminishes with scope.

 

[marty1]

But, more importantly, I'm pretty sure quantum chaos is still a very open subject. We have very little knowledge about open quantum systems our large-scale coherence. Deterministic chaos and quantum probability must be reconciled somehow (some might point to the deficiency in the definitions of space and time, as they manifest at the Planck scale, but others point out the significance of so-called "subplanck" structures at these scales:

http://www.ncbi.nlm.nih.gov/pubmed/11507634

I suppose I can agree. The fact that a quark, for example, seems to be a persistent fixure in our universe would imply to me that there is at least some structure smaller than it keeping it here although we might not be able to ever measure it because its reach ends there (the machinery inside does not cause events independently "outside" the quark).

 

[chiro]

Mathematically, a purely random process is one that has maximum entropy. You can use this definition to really study random-ness in a well defined mathematical context.

As has been pointed out, one thing relates to whether you have all the information in order to describe an assess the process and one thing to be aware of is that most things that are analyzed in a systematic way are inductive and extrapolatory.

By inductive and extrapolatory I mean that we start off with a very narrow spectrum of observations and then try and explain the entirety of the system given this information.

This is the opposite to inductive where you start off with a complete description of something and then use this to get attributes for some particular subset of the system.

The inductive approach is what scientists do and the deductive approach is what mathematicians do and somewhere in between is what happens when you look at everything being put together.

Now in statistics (which is what science is using as its language ever more slowly), we can make inferential statements about things but we do it with a machinery of mathematics behind it (which is bound by its own assumptions).

But statistics and its use is based on having incomplete information (a sample, is usually always considered to be a narrow subset of the entire population) and the best we can do is make an inference under relative uncertainty and this is what statisticians and people that use statistics actually do.

One final thing about randomness is so far, we can't really describe or easily deal with models with severe complexity both in an algebraic and also in a geometric form. Most people can't deal with more than 5-10 variables and beyond that it becomes too complex.

This relates to randomness in a way that a lot of things are called "random" if we can't see an immediate pattern even though the pattern may exist and may not really be hard to describe algebraically or geometrically even though its hard for "us" to do so.

 

[wuliheron]

You could say the opposite for "ordered" and claim mathematically it has minimum entropy and we have incomplete information on whether anything is "truly" ordered. Furthermore, we can't fully model orderly systems and even the 3 body problem continues to frustrate.