Think i know why RNG is inconsistent in PoE

The problem with this thread is that people assume that it really is RNG, with no other factors.

Unless GGG has explicitly said it's pure RGN, then it can't be assumed that it is. Unless GGG has said it's pure RGN, then arguments about how RNG works are meaningless.

It's quite possible that the odds increase after each failure. I've seen it done in other games where certain events became predictable, simply by counting the number of times since the last event. It never became 100% predictable, because it was still chance. It's just that the chance became very high after a while.

A simple example is on the first roll you have a 99% chance of failure. If you fail, the next roll has 98% chance of failure. And so on. Eventually it hits a cap of 5% chance of failure. There could also be a cap on the first roll, so you have 100% chance of failure for the first 10 rolls or whatever. It can also be linear or non-linear.

That type of system is still RNG. It is still random chance. It's just that the odds change as you go along.

Maybe GGG has stated that it's pure RNG. We know it's not pure RNG with evasion. So unless GGG has stated it's pure RNG with everything else in the game, then it can't be assumed that it is.

On the other hand, statistical based testing needs a hell of a lot larger sample size than what I've seen in this thread.

"
harddaysnight wrote:
Maybe GGG has stated that it's pure RNG. We know it's not pure RNG with evasion. So unless GGG has stated it's pure RNG with everything else in the game, then it can't be assumed that it is.


Of course things can change over time, but GGG has stated that the RNG-busting Evasion entropy system is only used for Evasion, and not other systems in the game.

https://www.pathofexile.com/forum/view-thread/11707/filter-account-type/staff/page/8#p656397

"

0nin wrote:
Does this even distribution system Evasion has, apply to any other percentage systems? For instance, Energy shield's innate 50% stun negation chance? If not, why not?

Mark_GGG:
No. Because for each thing that system was used for we'd have to store several variables in every game object, and evasion is the only case it's actually necessary for.


And it makes sense why: It's actually more work (and more variables to track) to modify the RNG as you go along than to just use the same straight-up pseudo-random routine for every calculation.
- here's my sig
Last edited by AcrylicHercules#1220 on Feb 2, 2021, 2:37:27 PM
"
AcrylicHercules wrote:
Of course things can change over time, but GGG has stated that the RNG-busting Evasion entropy system is only used for Evasion, and not other systems in the game.

https://www.pathofexile.com/forum/view-thread/11707/filter-account-type/staff/page/8#p656397


In my mind, this is not explicit enough, although it does imply it.

If GGG has explicitly stated that it's pure RNG, then why is this thread five pages long? Why not just link it on page one?
"
harddaysnight wrote:
"
AcrylicHercules wrote:
Of course things can change over time, but GGG has stated that the RNG-busting Evasion entropy system is only used for Evasion, and not other systems in the game.

https://www.pathofexile.com/forum/view-thread/11707/filter-account-type/staff/page/8#p656397


In my mind, this is not explicit enough, although it does imply it.

If GGG has explicitly stated that it's pure RNG, then why is this thread five pages long? Why not just link it on page one?


You're right about the "without seeing the code we'll never know for sure" part.

As for 5 pages? Who knows? I didn't see this thread until it was 4.8 pages long, so there's that. :)
- here's my sig
I came in late also. I didn't read every post, but I read a lot of them. I glanced through it specifically looking for links on how it really is done, and there were none.

People are arguing about how RGN works, when that's the wrong argument.
"
Toshis8 wrote:
Did a little experiment. Corrupted 35 gems. Red dots are bad outcome, green - good.



Imagine all 35 gems were expensive gems. If, after each successful corruption i used vaal orb on 2 trash gems, i would have saved 9 expensive gems (in other words, saved nearly 26% of gems)


Damn its almost as if good gem corruptions outcomes happened once every 6 times on average :rolleyes:
"
RedStark wrote:
what a crap game , 35 gems : 5 win , 30 lose this is straight bs


I've been measuring gem corrupts for a couple years now. I hit 1 in 6, seems right to me.


That said TC doesn't understand rng.

If there were never streaks that wouldn't be very random would it?
"
Anonymous1749704 wrote:
"
Shagsbeard wrote:
Okay. There's no such thing as "truely random", or "actually random" or "really really really really random". That's all nonsense.


This. If one manages to make a true random generator (as opposed to perceived random), they'd likely have a Nobel prize coming their way.


No. It is rather easy to make a true random number generator. They are commercially available, as far as I know. You can use either radioactive decay or cosmic radiation (there are probably a few other possible sources) to trigger them.

But that doesn't really matter - for the purposes of simulating the drops in PoE, the known deterministic pseudo-random-number generators are plenty good, and pass all the basic tests for randomness with flying colors.
Remove Horticrafting station storage limit.
brb gonna grab my tinfoil hat.
-Official Forum Dweller-
-I started the hoho movement-
-Exploit Early - Exploit Often-
-Moderators are absurd and OP-
-Heist Enjoyer-
Dare you to define what you mean by "true random".
Last edited by Shagsbeard#3964 on Feb 2, 2021, 5:29:52 PM

Report Forum Post

Report Account:

Report Type

Additional Info