The Dink Network

Random number generators

Why waste time and money using a random number generator? Instead you can simply use anything from dice to tossing a coin. While yes, the results that they produce are not completly random, it doesn't matter as the influences that make the outcome of a coin toss or throw of a die might as well be. The popular argument is that If a die, for example, has a some dirt on one of its sides that will increase the chances of it landing on a particular side by 0.000001% then it is therefore not random. While this statement is true this is not the whole picture. How did the dirt get there? Perhaps a group of dust particles stuck to it at some point because of static attractions. What are the odds of this piece of dust having the exact weight that it did? Landing in the exact place that it did on the die? Odds of the dust staying stuck to the die? Each of these questions has almost an infinite number of possible results, and there are thousands, perhaps millions more influences that would will distort the odds of the end result. Once you multiply the collective odds of this particular piece of dust being where it is, it turns out that the odds of it being exactly where it is are so slim that it is trillions and trillions to one. However, even this does not paint the whole picture. How did the dust even get near the die in the first place? If it weren't for a sudden gust of wind ten years ago, that single peice of dust could be on the other side of the world. What caused the gust? Would that not have happend if there was no city near by producing heat energy and causing an area of lower atmospheric pressure? If one person in particular had decided to slightly turn their heating up, how many grams of insulating synthetic wool were in a particular cavity in one of the walls? What if the cavity wall insulation was never invented? The list is unfathomably huge, perhaps infinite. What is the die was made differently, what if dice were never invented at all? What if the human race died out thousands of years ago?

My point is, that while yes, rolling dice to try to achieve a random result can be biased, there are so many variables that influenced the final outcome of that one roll that we have no control over could therefore might as well be random. Why spend £100 on a random number generator when the end result of rolling a die is SO close to being random.

Okay, so my point is a little half baked and mabye isn't valid - I haven't made my mind up yet, just something kind of intresting disguised as a rant about random number generators. What do you think?

A random number generator is truly random and therefore stops all complaints about it being biased, in a lottery of some sort for example, whereas with a die you'd always get someone whining that it wasn't fair.

About the die, you could argue that as you know that a bit of dirt could be there, that's what made it not random. If I stuck something to a die and then rolled it, would that be random because there is a possibility I wouldn't have done it?

Not random - I never said that rolling the die it would be random, however it is just one of those trillions of thing that make the outcome of the eventual roll more random. Even if you hadn't stuck the dirt on, there was always a possibility that you would thus making the outcome more random. If an infinite number of things were possible in this universe, then the outcome would be infinitly close to random... but not quite random... I think. All I ever suggested was that rolling a die was so close to being random, as it is influenced by so many variables outside off our control, that there is only a slight difference, if any in using a die and a random number generator.

Well, a dice being slightly biased isn't bad for most purposes as long as you don't know how it is biased. If your dice is slightly biased then taking a certain dice and not the other is random too, so I don't see the problem really...

I may be wrong about this, but I don't think random number generators are truly random. They use the computer's internal clock to derive numbers which appear to be random.

Those are software random number generators, which are (as you said) pseudo-random number generators. There are also hardware random number generators which are dependent on a random physical process, often involving radioactive decay or quantummechanics.

I thought it was algorythms or something along those lines - still not random. Is there such a thing as a truely random event? Is the universe not all just an endless cycle of cause and effect?

*I'm actually asking a question, not hypothetical .

If it's all cause and effect, then I suppose it depends on whether the beginning of the universe was random or not. If you go with the theories that there was no beginning, then, well, things could get complicated. And what about chaos theory?

This discussion could get theological...

I think... That in our universe - what we perceive as reality, there is no such thing as random. It is hard for us to even grasp the concept of it. Something happening - but for no reason? How is that possible? If a coin lands heads or tails, we can basically say that that happend because of how high or fast you tossed it, but if the outcome was random... Not influenced by anything, it just, happend then that is strange. That is why I was interested in radioactive decay, how can it be random, when an atom expells a neutron (or whatever other crazy shoot that comes out of those things) it must be for some reason - why it was particularly at that time when it gave off radiation. Perhaps there is such a thing as random - however it seems to me that our understanding of the universe we live in is very limited and we know nothing of the concepts that could be outside of our world. When I try to think of it deeper, my head hurts .

Correct me if I'm wrong but chaos theory is pretty much what I've been talking about? How the apparently smallest, most insignificant event years in the past, hundreds of miles away can cause massive dissasters or whatever.

To think... This whole discussion would never have taken place if it wern't for me taking a few to many hits from my pipe .

Yes - I guess you're right about chaos theory. I was thinking it was random happenings that threw off calculations (e.g. weather forecasting) but it's cause and effect really, just something so tiny or far away or whatever that it's impossible to take into account. Oh well.

About tossing a coin, it could be that as it is utterly impossible for us humans to measure beforehand exactly how hard we will flick it, then for us that is random, even if technically it isn't. It's all how we perceive it, like you're saying I think. I'm sure a robot could be built to toss coins always heads.

Good conclusion. I'm off for some more weed - it makes me smarter.

Okay, now that I'm smarter, I have realised how this discussion can go deeper. One of the biggest philisophical debates for hundreds of years has been the aargument of free choice vs. fate. If everything is cause and effect, has everything you will ever do already been set in stone from the very first "cause"? Well.. If there truly is nothing random, then, well yeah. Nothing I can do to sttop myself writing this, unless it was already made that I would. Every little detail of your life, how hard you press each key of your keyboard when you type.. Everything.

Theoretically, this means that it would be possible to predict the future. If you could enter into a computer, every possile variable about everything into our universe and how they effected each other... Of cource this is pretty much impossible, but kinda intresting all the same.

One more thing... Radioactive decay? Is that definately random? If it is, then everything I have just said is completly void as something like quantum mechanics or whatever could massivly effect the outcome of the universe. Is it possible that we just dont know enough about radioactive decay to see a pattern? That seems likely to me. Anyways, goodbye world, I'm off to play some computer games.

Well there is the Heisenberg uncertainty principle, which says that you can't know the exact position AND the exact momentum of a particle. So you can't know "every possible variable about everything in our universe".

It's all very simple really - random doesn't exist logically. And computers can't get beyond their logic, so they need outside information to generate random numbers.

There's also tunneling, where if you have an energy barrier, waveforms have a probability of 'tunneling' through it, even though they do not have nearly enough energy to go over it. Don't forget vacuum energy, and the particles that appear/disappear with it either.

Chrispy, does this mean that that makes it random, or just far too complicated to calculate, and therefore seemingly random?
(This article involves this discussion, if anyone's interested enough to read it (or skim it like I did - ignore the middle part!))

Tunafish, if everything you do is settled already from the first "cause", how would choices work? When I think or make a conscious decision, is that only a result of a certain way the neurons (or whatever) in my brain have been bouncing since yesterday?

Well, for the waveforms, I've only ever studied them (briefly!) as a probability function. If this means that they are random, I'm hesitant to give an emphatic yes, as larger systems tend to become seemingly ordered. Since the probabilities can be know, limited by Heisenberg, you can slightly narrow down the possibilities. How much is beyond the scope of my studies.

Another problem for trying to make predictions using cause-effect simulations are vacuum particles, ie, virtual particles, as in uncommon situations they have a tendency to interact with regular matter, and we don't really understand vacuum energy yet. Sure, you could exclude those theorized interactions, but you couldn't be certain. Basically the pop into existence, then annihilate each other completely without releasing energy. Unless they don't.

Complexity, and time are also huge problems. First off, we don't know why things move. We also don't know why there is time. We have evidence that we experience time in discrete quanta, but we have no idea how to merge this with the idea of momentum. Complexity comes into play with the problem that inter-particle interactions increase much faster than the factorial of the number of things you are simulating. factorial(x) is 1 force. We have more than that. You can always find ways to approximate these calculations, but the defining characteristic of entropy is the idea that small changes have large consequences. These small changes are lost in approximations.

Randomness in itself is a utterly fascinating subject, but it's too large for this margin to contain.

Oh, and just as an interesting side fact, did you know that temperature is only defined at an equilibrium. As in, inside a fire, or over *any* heat gradient, there is no temperature, per say? There are particle velocities. It's an interesting disticntion, and if I was more awake, I'd tie it into the above, but I am sleepy.

When I think or make a conscious decision, is that only a result of a certain way the neurons (or whatever) in my brain have been bouncing since yesterday?

It seems likely to me, but I really have no idea - it's just a theory I had (probably a lot of others thought of it too).

dang I hate being in high school, we learn shoot all. We never learn anything remotely intresting in physics or chemistry and as for philosophy - there is one school in my whole country that teaches it. ONE. It took my class a week to grasp that the angle of reflection = angle of refraction. And as for three states of matter which has been drilled into us for 4 years, I then find out it's not even true! Man, what other crap have they taught us.

Undergrad level physics is like that too. First they teach us a simple method for energies and velocities, then a harder method, then a yet harder method, each of which is an expansion of the last for the most part.
If you really want to learn it, pick up/borrow a first year textbook from whatever source you can find, and start using it. Learn trigonometry well. Learn integration well. The biggest problem one has in self-learning physics, is that you have to learn how to set-up a problem correctly.

Argh, I would, but i'm already teaching myself additional mathmatics and it's boring the hell out of me. I do occasionally look at some more advanced stuff on the internet, but I have no use for it at all, it just intrests me. I'm not sure what I want to do at university anymore though.. I was thinking medicine but now mabye I'm thinking something more intresting. The thing is, all my brothers decided to do intresting things and, well... One of 'em has a phd philosophy and works in a crappy customer services job for paypal. An other one has a first in phycology from cambridge (never wanted to be phycologist) and now plays online poker for a living, making a very average wage for massive hours. And my last brother got a 2-1 maths and now signs on for a living. Oh and my sister did drama and media studies at uni and currently works in a supermarket for £5 an hour. I guess they are all relatively happy, but I mean If i'm gonna get a degree I might as well do something I have a vauge intrest in making a career out of, right?

Engineering, computer science, journalism, business... anything which has a non-academic profession associated with it is probably good.

Or, you know, real science.

Math can be boring sure, but, well, Medicine for the first few years is rote memorization. Economics can be good if you are a people person, and got the right sort of personality. Just so you know, physics is mostly dinkized math and making formulas, chemistry is logic and using formulas, and biology is memorization. Sure, it's a *gross* simplification, but that's what a lot of it is. Engineering has many many subfacets as well, ie, chemical, electrical, a fair sort. Go into something that you can find a lot of co-op jobs for, and you'll likely have a good deal of options post graduation.

Any degree can get you a good job. It's just that some degrees you need to be *exceptional* to get a good job, and mediocre to get the degree, while in others, only people who work really hard or are that dang good get degrees.

Well if you really want a job you're probably best of doing something in business or computer science. Though it might be different in other countries. And science, I don't know. It might differ greatly between studies. (And to what you define as real science probably...)

If it's got science in the name, it's not science.

Business is also a little tricky, as, like I said, you need to have the right sort of personality. You can't be introverted, need to look good in a suit, need to be highly sociable, need to pull of high-class and that sort of thing. If I tried it, I'd fail horribly, as I just couldn't take the stuffiness seriously, and I'd quickly get bored out of my mind. One of my friends on the other hand, he's going to go far in the executive world, mostly because he likes it, he can do it well, and he can pull of class.

You could always be an entrepreneur... you can make a lot of money and doesn't matter what kind of degree (if any) you have. You really got to put a whole lot of planning and effort into it if you want to succeed, though.

Nah, I'm not the buisness-ish type. Mabye chemistry sounds good, but I have absoloutly no idea what I want to do. Perhaps a mad scientist would be pretty cool, I could like, get a pigeon and glue it to a cat or something then go on "dragons den" with my idea.

Chemistry, good choice...

I study chemistry as well, but I think I could actually be good in economics as well. I found the economic subjects in high school quite easy, easier then most exact sciences in fact (Chemistry excluded). I think it's usually best to follow your heart when choosing a study and placing the potential paycheck in second place.

And yes I agree, when it has science in the name its probably not real science.

Chemistry isn't what I'd call exact you know. You get isomorphs, and mixed equilibriums and multiple reaction paths. Complicated matrices are a pain in the arse with all the cross checking you need to do to make sure there are no unintended reactions or to make the most of what you want without fubaring the whole mix.

Well, some parts of chemistry aren't exact. (Biochemistry comes to my mind specifically) That it isn't is mostly because it's too complex to take everything into account so you start to make assumptions and work intuitively. It's often not impossible to calculate all kinds of equilibriums, it's just very cumbersome. So in fact the basis is exact, the way we handle it might not. I'll have to admit though that it's not as exact as physics or mathematics. (Quantumchemistry might be an exception)

Yes the maths are generally uncomplicated, just understand logarithmic functions and integration and you're probably fine for the most part. Still I'd call chemistry an exact study, though it depends heavily on the part of chemistry you're studying.

Yeah, that's true too. You can get fairly precise probabilities. Now, it's a given that your dealing with such huge quantities that the probabilities just become a measure of the concentrations at the end. I was more thinking that it gets really really hard to predict the exact outcome of a matrix of lets say two separate reactions You have to deal with the end products, the mid-products, and how they all interact, and a myriad of other factors, pressure, heat, electromagnetism. The margin of error goes up, and you try to control those factors to make the margin of error go down.

Then again, one could argue along the same lines for any experimental research. I'm just thinking that there would be a larger amount of error with complicated matrices, but I'm sure I've just not studied far enough to actually start figuring that out. Mostly I'm studying to see if one can make alloys via electrochemisty, and stuff like the salt effect and a similar kind of effect to the salt effect that happens when you make alloys. It's fun stuff.