A thought on Luck

[ARTgames]

Well we don't know ware Meiun is getting his salt or even what he is really using. I'm amusing stuff that I think may be true that will make it PC dependent which might explain why it seems some accounts are "luckyer" than others. We are just guessing. We are assuming a lot.

But I'm sure you'll agree that that last paragraph would be nice add-on to stick online 3. :D Its free and not that hard to implement to the point ware its good.

[InvisibleMan]

Hmm, I would think that it is more likely based on the server, not each client.  If you give the client the responsibility of generating the random number (or choosing the seed) it could be abused.  I would also think that it is seeded only once, not per client.

[JoEL]

One of my teachers claimed that generating a random number on a computer is based on time and computers can never "randomly" generate a number, there is ALWAYS a pattern. One of the examples she gave us was when people came home from work to play these certain games with "random" events/drops or whatever at around the exact same time, every night. A pattern within the game would form, it'd feel as if the game was repeating its self.

I think this is a pretty interesting story and theory. If the drop chances in S.O are based on random number generating (which we all know they are) then maybe there is a specific time when our "drop chance" is at its best.

Just something to think about, it sounds very probable to me.

CONSPIRACY, discuss.

[InvisibleMan]

If it's based on time, it's probably based on time in milliseconds.  Because of this, it would be impossible to aim for a specific time.  If it's based on minutes however, you could possibly do so and get a group of people in the same "lucky-slot".

[JoEL]

Well if the story is true, it'd have to depend in the whole time, hours, minutes, seconds and milliseconds.

[InvisibleMan]

That's what "based on time in milliseconds" means, heh.

[Mr Pwnage]

For anybody interested:

Kill a DS at exactly 12:03.97pm, you will get a double drop of dragoon and inferno sword.

[Scotty]

For anybody interested:

Kill a DS at exactly 12:03.97pm, you will get a double drop of dragoon and inferno sword.

Your watch was off, it's actually 12:03.969PM to be exactly precise.  You can't round when it comes down to such fine precisions!

[Mr Pwnage]

For anybody interested:

Kill a DS at exactly 12:03.97pm, you will get a double drop of dragoon and inferno sword.

Your watch was off, it's actually 12:03.969PM to be exactly precise.  You can't round when it comes down to such fine precisions!

Oh damn, good save.

[Lingus]

Well if the story is true, it'd have to depend in the whole time, hours, minutes, seconds and milliseconds.

This is the point. If it's based on milliseconds at all it is extremely unlikely that the same exact time would hit more than once. Not to mention, I assume if you throw days and years in there you'd never get the same date/time.

[ARTgames]

We don't even know what Meiun is using.

But non the less game maker default seed is the ms timer that goes on at system boot. So every time you rest your PC it also rests the timer to zero. Also you can have variations in the keys and still get similar results when using this simple generator that game maker uses. Even thought this is done at the ms level after a long time the majority of the key will remain the same because only the first few digest are moving. This is ok for a lot of games I just fell for a mmo a "more random" method would be nicer and be more fair, since "better random" stuff is still easy and cheap to use. And if anything it would be more harder to exploit.

Non of this really matters. When there are more players and a trading system I'm sure these items will get to people with work.

[Mr Pwnage]

Well guys, practically speaking here, I don't think it will ever be possible to program a "true" random generator. But the fact of the matter is the current system works, is well encrypted, and I see no way to exploit it based on tracking patterns due to the fact that would be impossible without extreme decompilation and decryption of the game code...and even if on the hair of a chance it were discovered, it would be nearly impossible to exploit due to the fact that you could go to lengths such as nano seconds or hell, even past that.

Point being as long as the random system appears to be working like it should...there's no real reason to sweat about true random...the fact is, it's a program, not a free thinking individual, machines will never be random. (and I am sure there are many arguments to this point I am making...but I personally believe that you cannot program a machine to be truly random).

[LeGuy]

Well guys, practically speaking here, I don't think it will ever be possible to program a "true" random generator.

http://www.random.org/

Uses atmospheric noise to essentially determine a random number. It's still not truly random - a certain set of conditions are still determining the numbers, but it's been improved to the point where patterns based on the timecode of the computer are nonexistent.

Of course, I'm not suggesting such a method should be used for S.O.; our current system works fine.

[ARTgames]

This is getting off topic but I would like to point out some stuff that might not be known. But we have all came to the conclusion that this is an ok how it is and I'm fine with it as I stated in my post before this one.

@Mr Pwnage
Dont assume so much about the system now. Also here is a program I made that makes what can be considered truly random numbers. Or at least really good ones.

This program need to be running near real time or as fast as the PC can provide you. Simply have a loop that keeps inverting a bit and have it check for when a key is pressed. When a key has been pressed put the current stat of your bit into a buffer and do it the processe over. Get as many bits as you want and add them up.

Now this is too slow to use in real time applications like games but is a good sources of entropy to add to a key for a prng.

This is considered physical random (like dice) and is know to be truly random. There is know way some one can press a key constantly at the level it measures, unless you got the slowest PC ever. :P  There are other ways to do this like how true crypt makes its salt for its encryption. http://www.truecrypt.org/

@LeGuy
If they are doing what they say they are that site is "truly" random or as random as physics of today have known. That side has nothing to do with any sort of time codes. 0_o I'm not sure why you said that.

But the site does hold the world truly random in italicized because whether or not true random exists is still debated.

[JoEL]

http://en.wikipedia.org/wiki/Random_number_generation#Computational_methods

I think that's called a hardware generator Art. I read about it on here:
http://www.robertnz.net/true_rng.html
And from what I read, it does actually work, if done correctly. Direct quote:

If you sample the output (not too quickly) you (hope to) get a series of bits which are statistically independent. These can be assembled into bytes, integers or floating point numbers and then, if necessary, into random numbers from other distributions using methods such as those in newran.

It doesn't seem very stable or very dependent. But this sentence worries me a bit to:

I had experimented with one of these from a Canadian company and found it to produce reasonably satisfactory numbers