|
Originally Posted by Aett_Thorn
It's okay, we both schooled Claws.
 |
I hope you used herostats to countthat 15,000 defeats.
I've heard rumour a few times that the random number generator is seeded for your character based on name at generation, creating lucky and unlucky characters. But I dont believe that myself
|
Originally Posted by sblinks
I've heard rumour a few times that the random number generator is seeded for your character based on name at generation, creating lucky and unlucky characters. But I dont believe that myself
|
|
Originally Posted by Dispari
I don't know why Dink thinks she's not as sexy as Jay was. In 5 posts she's already upstaged his entire career.
|
|
Originally Posted by Aggelakis
Rumors are like statistics, 97.54317% are made up.
|
Hey guys,
Thanks for all the replies. I wrote that post in the early morning, well early for me, and didn't realize how much of a whiny baby i sounded till i read it again this evening. lol
Anyway as far as the kills. a few of you reminded me that if i'm on a team that it drops the chances of me getting the purple drops. The 15,000 that i counted to was when i was solo. if i wsa on a team, i didn't count those at all. Perhaps it's my method. i usually run lib farm, even lvl set for 6 or 7 on my fire/rad corr. i normally do like 5-6 runs in a row, all solo. I can understand the randomness of the purples, but the reason i thought it might just be something wrong with my account, i know sounds ridiculous, is because in those 5-6 runs i usually get like 1-2 non recipie io's, 1 recipie IO, normally something like crap of the hunter, and then 3-4 pieces of common salvage, i don't keep up with the insps i get.
My wife, also on a fire/rad, even lvl set for 6-7, easily beats my drop rate for 5-6 runs in 1 run of the same farm, not to mention, she get's the occasional purple drop. lol
It just makes me wonder if i should send a message to support saying check my dice roll it's broken. lol
Oh, and the number of 12,000 kills per purp drop i quoted earlier was a number quoted to me by like 6-7 different people in game, no idea where they came up with that number.
All the tips, and tricks for good drops are greatly appreciated. I won't get alot of "play time" until this weekend, but I'm deffinatley gonna try out some of the suggestions.
Thanks,
Z
|
graystar_blaster and Guildenstern are dead: rng works like this from what I was told by a programmer example a coin flip heads or tails its 50%-50% each flip. So somewhere in the universe someone has flipped 10,000 tails in a row without a heads at somepoint |
|
Originally Posted by zoltaire
but the reason i thought it might just be something wrong with my account, i know sounds ridiculous, is because in those 5-6 runs i usually get like 1-2 non recipie io's, 1 recipie IO, normally something like crap of the hunter, and then 3-4 pieces of common salvage
|
|
Originally Posted by ChaosExMachina
Keep in mind that computer random numbers are usually seeded, which makes some likely to have a sense of inertia, depending upon the seed source and the implementation.
|
I'm gonna have to call BS on the "are usually seeded". Most random number generators are exactly that, as random as the algorithm can make them.
|
Originally Posted by graystar_blaster
rng works like this from what I was told by a programmer
example a coin flip heads or tails its 50%-50% each flip. So somewhere in the universe someone has flipped 10,000 tails in a row without a heads at somepoint so the 1/12,000 rng is 1 in 12,000 each kill, and, not if u kill 15,000 u get one. Its not like lets say a spedometer where when u go 12,000 it resets to zero its actually a 1 in 12,000 chance each time u kill anything that is capable of dropping a purple recipie hence the rareness and also thats why its sucks when u get one and its a confuse or a sleep set. enjoy -neg rep anyway please! |
were these 15000 kills in a single play session? the RNG changes every time you log out and back in.
As far as I know, the RNG randomizes with every defeat. It's the only way that would make sense. Is there a citing on this otherwise?
|
Originally Posted by zoltaire
i usually run lib farm, even lvl set for 6 or 7 on my fire/rad corr. i normally do like 5-6 runs in a row, all solo. I can understand the randomness of the purples, but the reason i thought it might just be something wrong with my account, i know sounds ridiculous, is because in those 5-6 runs i usually get like 1-2 non recipie io's, 1 recipie IO, normally something like crap of the hunter, and then 3-4 pieces of common salvage, i don't keep up with the insps i get.
|
|
Originally Posted by Rajani Isa
I think the term you are looking for is a "non-cumulative 1/12,000 chance"
|
|
Originally Posted by Decorum
I'm gonna have to call BS on the "are usually seeded". Most random number generators are exactly that, as random as the algorithm can make them.
|
|
Originally Posted by Decorum
As far as I know, the RNG randomizes with every defeat. It's the only way that would make sense. Is there a citing on this otherwise?
|
|
Ironblade notes: Yes, as you yourself mention, that's exactly the point. Random number generators generally are not truly random. True randomness is very hard to achieve. |
|
Originally Posted by Decorum
Maybe I don't understand "seeded", but that sounds to me like deliberately making it non-random (making it trend to one thing or another). True randomness is very hard to achieve. But, non-randomness in random number generators is a small statistical anomaly that is almost impossible to detect through usage.
|
|
Originally Posted by Decorum
Maybe I don't understand "seeded", but that sounds to me like deliberately making it non-random (making it trend to one thing or another).
|
|
Originally Posted by Ironblade
Okay, you don't understand it. The problem is that a computer has no 'built in' method to generate numbers that are actually random. Unless perhaps you have a quantum computer. So your 'random numbers' are, in fact, guaranteed to be NON-random. The goal is to make them 'random enough' that there won't be a perceivable pattern.
One solution is 'seeding' a 'random number generator' with some starting value that is unlikely to be repeated the next time. One popular seed method in the past has been the number of seconds elapsed since midnight. |
|
Ironblade explains: Okay, you don't understand it. The problem is that a computer has no 'built in' method to generate numbers that are actually random. Unless perhaps you have a quantum computer. So your 'random numbers' are, in fact, guaranteed to be NON-random. The goal is to make them 'random enough' that there won't be a perceivable pattern. One solution is 'seeding' a 'random number generator' with some starting value that is unlikely to be repeated the next time. One popular seed method in the past has been the number of seconds elapsed since midnight. |
|
Originally Posted by Decorum
OK, I understand it now (well enough, anyway). Now, is there any indication that this is done here? Yeah, "random enough" is usually good enough to not be noticed without a bunch of datamining. I mean, we are talking about "not perfectly random" right? I'm assuming that means some tiny fraction of non-randomness exists, but is likely very insignificant. I'm assuming some kind of thing like "on a d100 roll, there is a .0001% chance of non-randomness" or something (boy, that's an awful explanation with numbers pulled out of my butt and likely doesn't work along those lines technically, but I hope what I'm saying is understandable).
Perfectly willing to be corrected, provided you can actually understand my garble. |
OK, so back to my original comment, then...does this preclude each defeat being its own random roll? Or are we back to "what do you consider random"?
