[ 3 / biz / cgl / ck / diy / fa / ic / jp / lit / sci / vr / vt ] [ index / top / reports ] [ become a patron ] [ status ]
2023-11: Warosu is now out of extended maintenance.

/sci/ - Science & Math

View post   
File: 1.96 MB, 278x177, php421.gif [View same] [iqdb] [saucenao] [google]
2320929 No.2320929 [Reply] [Original]

Let's talk about the BULLSHIT in video games called "Randomness."

Anyone who has even a rudimentary understanding of science and the universe knows that randomness doesn't exist. "So what the fuck are they using?" you ask? Why, let me tell you.

Randomness, in its literal term, is unpredictability. When something is "random" it's unpredictable. What GAME DEVELOPERS think it means is "Any outcome is possible" which isn't how the universe works, and unfortunately, video games do also have to abide by those rules as well. We can't generate "any possibility" even in programming, we can, however, generate (limited)unpredictability.

So developers, in their idiocy, decide to rely on arbitrary variables to create (limited)unpredictability. So, you get things like "This has a 75% chance to succeed." Except, what's REALLY happening, is it's either a 0% chance, or a 100% chance, the percent chance it gives you is only estimated from current information, which is what probability is for, and also why it's flawed.

The idea of probability is a good one, and a useful one, and necessary in programming, and will be a key factor in developing artificial intelligence, but in each event of its usage, it's purpose is to PREDICT, and hence the famous misunderstanding comes about that "the probability of something happening 50% means it could go either way" when in reality it's "The current data is inconclusive." It doesn't show the likelihood of something happening, although that is a convenient way to explain it, instead it shows that the current data is pointing more in one direction than another (75% chance to succeed would have the data pointing more toward it succeeding than failing."

Understanding this, the way it's being used in video games is clearly COMPLETE BULLSHIT. It's not giving you a probability of something happening - no. It's fucking lying to you. It's withholding shit from you.

>> No.2320935

shit copypasta from /v/

>> No.2320931

Let's say there's an algorithm that relies on the movement of the mouse cursor to produce unpredictability. This is already fundamentally flawed, as you're relying on how much you move your mouse to generate number blocks which are then fed through a system, and at the end, this system tells you whether or not an attack, say, will succeed. Before it reaches this end, however, it has one step - the critical step to give you that probability that you see on your screen and think "80%? THIS MIGHT WORK!" at this step, it takes the INCOMPLETE DATA and gives you the probability. The reason why this is incredibly FUCKED is because the step between that and the end adds more data, which influences it either way.

So really? It's not an 80% chance to hit, it's a fucking 0% chance. For the next 5 turns, each time you try. Because some shitty developer doesn't understand how the fuck probability works.

STOP USING THIS SHIT IN GAMES. EITHER IT WILL SUCCEED, OR IT WON'T. 0% OR 100%. ONLY THE SYSTEM THAT PRODUCED THE PROBABILITY CAN USE IT TO PREDICT SOMETHING. YOU ARE TAKING A COMPLETE SHOT IN THE DARK IF IT'S NOT 0% OR 100%.

.. AND THAT, MY DEAR FRIENDS, IS WHY YOU FAILED 3+ TIMES IN A ROW AT A 50+% CHANCE ON WHATEVER GAME YOU ARE NOW THINKING OF.

>> No.2320940

>Anyone who has even a rudimentary understanding of science and the universe knows that randomness doesn't exist.

Quantum mechanics here. You are very very very wrong.

>> No.2320953

>and hence the famous misunderstanding comes about that "the probability of something happening 50% means it could go either way" when in reality it's "The current data is inconclusive."
And what, exactly, is the difference between those two scenarios?

>> No.2320958

>>2320953

Well the data is inclusive part means we don't know.
And the 50% part means it could go either way.

>> No.2320961

>>2320958
And what, exactly, is the difference between those two scenarios?

>> No.2320975

>>2320961
1. We don't know. Basically it could go either way or be completely random or nothing happens at all.

2. It could go either way.

>> No.2320980

Lol, even the game designers know it's not actually random. It's psuedo-randomness.

Also, radioactive decay?

>> No.2320985

>Basically it could go either way
Yes...
>be completely random
(another way to say it could go either way)
>or nothing happens at all.
which, surely, is included in "it could go either way". So (this is getting tiresome) what, exactly, is the difference between those two scenarios?

>> No.2320986

>>2320929
ITT: people that don't understand statistics, programming, or breathing with their mouth closed.

>> No.2320990

STOP RESPONDING TO THIS COPYPASTA FROM /v/. THIS WAS POSTED NOT EVEN 2 HOURS AGO

>> No.2321003

>>2320931
People actually believe this. It really stems from a basic misunderstanding of probability
Just to address the final sentence
Lets say you are flipping a coin. What are the odds you get heads 3 times in a row?
It's 0.5^3=0.125
Thats 12.5%

So lets say heads is failing your 50% chance. It would appear that you will fail 3 times in a row once out of every 8 times you do this.

This is again wrong. What are the odds that you fail one out of 8 times? We're going into binomial here.
let X be the chance you fail 3 times in a row and we've performed 8 independant trials

X~Bin(8, 0.125)
The expected value of successes is 1 right?
What is the probability X=1 though?
P(X=1)=8C1*0.125*0.875^7=0.392695904

So the actual chance of failing 3 times in a row every 8 times you try is less than a half.
And I've lost sight of the point i was trying to make with this calculation...

Anyway, probability doesn't work how people expect it to.

>> No.2321028
File: 97 KB, 500x389, mechanics.jpg [View same] [iqdb] [saucenao] [google]
2321028

I did built in some random events in a specific map of mine in order to create unique game experience. This combined with various other contents in the map makes people addicted to it, resulting in a 32 players server with my map running almost all times with full slots. Apparently I have succeeded in which most game developers fail.

But to be honest, I mainly blame the strive for "casualization" in the gaming industry to have destroyed most intentions and attempts among developers in experimenting with exotic and new game mechanics.

>> No.2321031

>>2321003

What type of math is that? And what math should I learn to understand probability and randomness?

>> No.2321056

>>2321031
>What type of math is that?

Jesus Christ.


Go back to /v/, imrinfected.

>> No.2321063

>>2321031
It's called probability theory.

>> No.2321065

It sounds like OP has been playing FIFA 11

>> No.2321064

>>2321031
It's probability. You should learn probability to understand probability.

>> No.2321090

>>2321003
Just to finish my own post off, people who think like OPs post can't bring themselves to account for the fact that you could flip a coin every minute for the rest of your life and never get a heads. You could pass that coin onto your child, and he could continue flipping it every minute for his entire life, and still never get a heads. These are the same people who complained about the shuffle on their media player playing two tracks from the same artist in succession. A little education in probablity would go a long way.

>> No.2321091
File: 64 KB, 600x750, atheism1.jpg [View same] [iqdb] [saucenao] [google]
2321091

>>2320929

>Anyone who has even a rudimentary understanding of science and the universe knows that randomness doesn't exist

stopped reading there, it`s perhaps the most arrogant sentence i`ve ever see.

Anyway arrogance and "Nihilism" are is becoming the new hype of 2011 ! !

This because atheism is more accepted now, it doesn`t put you in a special position anymore.

So people who want to appear smarter then others start calling themselves nihilists and say "i don`t care if people say im arrogant."

>> No.2321112
File: 68 KB, 574x721, 1288978723731.jpg [View same] [iqdb] [saucenao] [google]
2321112

>>2320929
we dont have true random number generators, but pseudorandom and random are more than similar to the first couple of approximations.

also, true randomness does exist.
radioactive decay(at alot of atomic theory) is random.

learn 2 probability / statistics please

>> No.2321157

>>2320940
>Quantum mechanics here. You are very very very wrong.

Think again. I don't know shit about the video games but the true "randomness" is really interesting topic. I know you're going to tell about the Heisenberg uncertainty principle but where did you get the idea of it? From physics fact books..? Oh yes, so you really believe in quantum physics then?

>> No.2321173

>>2321157
Quantum physics is not something you can 'believe in'. Thats like asking 'do you believe in mechanics?'

I know what your point is, but I wasn't going to metion heisenberg at all. Bell's Theorem. look into it.

>> No.2321258

>>2321173
Okay, I respect your opinion. For me Bell's theorem seems to be nothing more than just another added layer on the cake of quantum physics.. Sure it works if you believe in "photons" and all that.

I seriously think that flipping a coin is not true random event at all. It's only very close. We just don't have precise tools enough for measuring all variables we have. Probability statistics sure says it approaches very close to 0.5 in long run yes but never exactly 0.5. I don't see true randomness here.

>> No.2321269

>>2321258
Go say something about mathematics now. That will be even funnier, because it will be even more blatantly wrong.

>> No.2321286

>>2321258

>i dont believe in photons

>> No.2321288

I don't really see the problem.
if the game tells you this task has 75% of chance to succeed, it is just that if you repeat the task n times and you call n_succ the number of times the task succeeds, then

lim(n_succ/n,n--> +infinity) = 75%

>> No.2321313

>>2321269
>>2321286
Prove it wrong then? Without quantum physics and show me the photon you're talking about? Oh. You can't. Damn superposition collapses once again.

>> No.2321322

>>2321258
>For me Bell's theorem seems to be nothing more than just another added layer on the cake of quantum physics
In short, you don't understand it. You checked the wikipedia article, noticed quantum mechanics mentioned at the top and didn't actually bother to read anything after that. Am I right?

>Sure it works if you believe in "photons" and all that.
Wait, what?

>> No.2321330

>>2321313
Photoelectric effect. It's what got Einstein his Nobel prize.

>> No.2321331

>>2321313

can you show me gravity? magnetism? sound? heat?

>> No.2321334

>>2321322
>>2321286
>>2321269
We're feeding a troll, guys. This was originally targeted at /v/.

>> No.2321341

>OP says there is only 100% and 0% chance
>OP says randomness doesn't exist
>a random guy in this thread says photons don't exist and is pretty much saying quantum mechanics is bullshit

IF I CAN'T TEST IT OR TASTE IT, FEEL IT, SEE IT, OR SMELL IT.
IT IS DOESN'T EXIST!
>nihilistic and atheist logic no better than theist logic

>> No.2321346

>>2321341
The funny thing is that you CAN see the results of the experiments that make Quantum Mechanics necessary.

>> No.2321397

http://www.canada.com/calgaryherald/news/story.html?id=31b1ced8-0b8c-4874-b654-cefe493e2321
the universe is asymmetrical

>> No.2321403

>>2321334
Yeah, I saw it, but Imrinfected has some serious issues with paranoia and is borderline schizophrenic, so it wouldn't surprise me if he was dead serious.

>> No.2321410

>>2321397
It's talking about the matter-antimatter imbalance.

>> No.2321467

>implying determinism really matters when the details are out of our sight and control
>implying if I flip a coin right now the results won't be relatively random

>> No.2321504

im 12 but i know whats going on in here, OP is wrong

>> No.2321507

Actually, if you knew every property of every particle, you could predict precisely what would happen, theoretically with 100% accuracy over any time period. QM, chaos theory, etc. arise because we start with a system that we need to measure. "We know there's one electron orbiting that hydrogen atom, but we don't know where it is, so here's the function that says where it probably is"... and the more particles you have, the more those particles interact and magnify the uncertainty.

tl;dr: Simulated "randomness" occurs because we try to make outcomes unpredictable by obscuring patterns; observed "randomness" occurs because we try to predict outcomes that have already been obscured.

>> No.2321520

>>2321507
>implying strong determinism
Dammit all, /sci/. LEARN SOME QM
http://en.wikipedia.org/wiki/Bell%27s_theorem

>> No.2321546

you realize probability and statistics are meant to explain things that humans and computers could never want to calculate to precision. It would take too damn long.
The randomness in videogames is a "random number" generation algorithm. The numbers really are absolutely predictable because they follow a formula. The trick is, game-players don't give a shit about the algorithm (you're playing a game to have fun unless you're OP)

>> No.2321547

>>2321507
U trollin

"if we could know presicely" is a purely theoretical speculation, as the HUP applies to every part of every system.

and no. the electron in not circeling the nucleus like a ball, and we do statistics on it. it occupies an orbital, and can be in one place one instant and "far away" in the next. this is not a statistical property, but an intrinsic one of the system.

Take nuclear decay. if you have two nearby atoms that can decay- it is impossible to say which one will decay and when. this property can only be analyzed statistically and gives rise to the half life property.

>> No.2321567

a coin flip is not really random, as it is perfectly possible to predict the number of flips(in a controlled experiment), and hence it is slightly ambigous for using as an example.
but in ordinary conditions it is so close to random that it will come out 50/50 after N tries (for large N)

>> No.2321571

YOU ARE ALL WRONG

randomness exists
a coin flip is random

random =/= probability

>> No.2321572

>>2321520
Do you seriously believe in Wikipedia? 4chan /sci full of Einstein supporters. wtf.

>> No.2321585

>>2321572
>4chan /sci full of Einstein supporters. wtf.
hahaha what

>> No.2321598

>>2321567
Exactly. Very (very) close to 0.5 but never true random. If you really start to think about it deeply, a single coin bashes Einstein.

>> No.2321599
File: 302 KB, 2048x1536, 20lbs2.jpg [View same] [iqdb] [saucenao] [google]
2321599

>>2321571
> start with precise velocity, spin, atmospheric conditions
> out come is "random"

>> No.2321597

>>2321547
That applies to a large system, not a single particle. Those "laws" model the decay of an entire sample of radioactive material. If a single particle had a precise half life, there would be no unpredictability at ALL.

>> No.2321609

>>2321597
a single atom will decay after a random amount of time.

>> No.2321611

>>2321597
How about Scrödinger's cat? Dead and alive at the same time?

>> No.2321618

>>2321598
But Einstein didn't like QM.

Oh wait - you're talking out of your ass.

>> No.2321621

>>2321597
but the point is no particle has a set half life. The unpredictability of quantum mechanics makes the half life unpredictable. We can only say that there is a very narrow window in which half the particles of a large radioactive material will have decayed. It could take 2 seconds or 50 billion years for any one particle. It's all in the probability from quantum physics

>> No.2321623

>>2321618
>Einstein didnt like QM
>Pretty much founded it

he didnt "like" the philosophical implications - doesnt invalidate the theory in any way

>> No.2321627
File: 20 KB, 254x286, 1251527672117.jpg [View same] [iqdb] [saucenao] [google]
2321627

op has spent two weeks farming the ground for zinroc, nothing to see here.

>> No.2321628

>>2321623
Oh, I agree. But >>2321598
is thoroughly misguided.

>> No.2321638

>op has spent two weeks farming the ground for zinroc, nothing to see here.

Haha pretty much.

I'd like you "but it is not random players can predict the outcome of teh algorhitm!!1" guys go and become rich because you can predict cards in poker.

Go on, the opportunity for infinite money is there, do it faggots.

>> No.2321639

>>2321599
you don't measure all those variables when you toss a coin, therefore random

>> No.2321646

>>2320929
>Anyone who has even a rudimentary understanding of science and the universe knows that randomness doesn't exist.

But that's wrong.

>> No.2321671

>>2320929
Randomness is a part of quantum mechanics. It's kind of the basis of all quantum mechanics.

And there's a difference between practical randomness and actual randomness. There's no need to make a distinction because it doesn't matter in any application. If you can't predict the result, or even predict that the result will tend in a certain way not intended by the designers, then it's random for all practical purposes. The only true randomness lies in the very small, such as whether an atom will decay or not.

Tl;Dr semantics don't make you noble or smart, they just make you argumentative and confrontational.

>> No.2321675

>>2321671
>there's a difference between practical randomness and actual randomness
>there's no reason to make a distinction

>> No.2321682

>>2321675
Derp, I meant that you can't lump them in the same category, but you don't have to make it clear which version you're using because it's obvious.

It's ironic that my argument was that the meaning was implicit and the implicit meaning of my argument was completely ignored.

>> No.2321688

>It's fucking lying to you. It's withholding shit from you.
Also when you set skill level lower than max, the AI is just pretending to be less skilled. Fucking lies!