My three oldest characters, what did they get?
For their first birthday they got jade armors, now for their second, they got windriders.
Yes, all three of them.
Now people will say "hey, it's a gift, don't complain", and I have to say, that's true, but you can tell me what you want. It still shows that there is something really bad coded. I'd bet that's the same reason why the "simulated luck" in gw just boils down to streaks.
Lockpicks? broke all, except for one evening, where six in a row did not break (17% chance to not break).
Chests? An evening just purple ones? Yeahyeah...just "bad luck"
Weird drop distribution?
Monster drops?
(...)
Anyone else experienced this? Or is just my account cursed and to be deleted?
Birthdaypresents and a crappy "randomnumbergenerator"?
2 pages • Page 1
K
v
n
Well, I tend to be a bit aggressive without my morningcoffee
I'm not complaining about getting windriders because they are white. In fact I basically can't complain about that. I rarely use minis, I don't buy/sell them and I vastly preffer creatures over npcs. What I'm complaining about is the randomness, or more the lack of it.
Getting just jades, whiptails or windriders is kind of boring and not fun.
Also I want to point out the main reason for the complaint:
I guess they did not code an extra routine to determine what mini you get. If the same random generator is used to determine all other "random" things in gw, that might be the reason why people experience weird streakes. Like my example with the lockpicks above. Or unequal lootdistribution. Or it might be the reason why some say "hard mode is like paradise, rares everywhere" and others say "Duh? What? I don't even get gold".
Something is definately broken with it, as the number of streakes I encountered in this game can't be explained with pure "luck/ bad luck" anymore and at least some others have experienced it too.
It sometimes feels like the algorithm is using the account name and the current date as input value
I'm not complaining about getting windriders because they are white. In fact I basically can't complain about that. I rarely use minis, I don't buy/sell them and I vastly preffer creatures over npcs. What I'm complaining about is the randomness, or more the lack of it.
Getting just jades, whiptails or windriders is kind of boring and not fun.
Also I want to point out the main reason for the complaint:
I guess they did not code an extra routine to determine what mini you get. If the same random generator is used to determine all other "random" things in gw, that might be the reason why people experience weird streakes. Like my example with the lockpicks above. Or unequal lootdistribution. Or it might be the reason why some say "hard mode is like paradise, rares everywhere" and others say "Duh? What? I don't even get gold".
Something is definately broken with it, as the number of streakes I encountered in this game can't be explained with pure "luck/ bad luck" anymore and at least some others have experienced it too.
It sometimes feels like the algorithm is using the account name and the current date as input value

If I toss a coin 10 times and get 10 heads, it does not mean the coin is biased. Same applies to not breaking 6 lockpicks or getting the same 2 minipets twice over.
Conversely, if I tossed a coin 100 times and got 100 heads the coin is certainly biased. Same applies if you had got ~20 lockpicks retained in a row. Granted, retaining 6 consecutive lockpicks is ~0.002% chance, but with hundreds of thousands of people using lockpicks all over the place... perhaps there's some dependance there? Either that or you got quite lucky.
You never said how many lockpicks you broke, but with a 83% chance to break each one, you are expecting to lose quite a few.
Conversely, if I tossed a coin 100 times and got 100 heads the coin is certainly biased. Same applies if you had got ~20 lockpicks retained in a row. Granted, retaining 6 consecutive lockpicks is ~0.002% chance, but with hundreds of thousands of people using lockpicks all over the place... perhaps there's some dependance there? Either that or you got quite lucky.
You never said how many lockpicks you broke, but with a 83% chance to break each one, you are expecting to lose quite a few.
B
For a start, random number algorithms are so close to random as to make no difference, including the probably standard and widely used one that Anet uses, but if you can write a better one without advocating a new PCIe card that measures background radiation, go ahead.
Secondly, your anecdotal evidence means not only less than nothing, but you seem to not understand that 'random' includes the events you describe. The whole point of 'random' and not 'fairly distributed' is so unlikely scenarios can happen. If you only want the mean outcome for everyone, complain about using random distribution at all, not the quality of the algorithm. You might have actually had a point.
Thirdly, do you think that people who got different pets come to this website and say 'Hey guys, I got different pets! I just signed up to say great work on that Random Number Generator, Anet! A+++." But because 20 people are going to post here saying they aren't on the trend line, you are going to have your ill-informed opinion reinforced. come back with 20,000 experiences and we'll have a look.
I blame Izzy. I heard he wrote a custom random number generator that only favors people who aren't me. Fire him.
Secondly, your anecdotal evidence means not only less than nothing, but you seem to not understand that 'random' includes the events you describe. The whole point of 'random' and not 'fairly distributed' is so unlikely scenarios can happen. If you only want the mean outcome for everyone, complain about using random distribution at all, not the quality of the algorithm. You might have actually had a point.
Thirdly, do you think that people who got different pets come to this website and say 'Hey guys, I got different pets! I just signed up to say great work on that Random Number Generator, Anet! A+++." But because 20 people are going to post here saying they aren't on the trend line, you are going to have your ill-informed opinion reinforced. come back with 20,000 experiences and we'll have a look.
I blame Izzy. I heard he wrote a custom random number generator that only favors people who aren't me. Fire him.
What he said^^. More specifically, this part:
Quote:
|
Originally Posted by bamm bamm bamm
Secondly, your anecdotal evidence means not only less than nothing, but you seem to not understand that 'random' includes the events you describe. The whole point of 'random' and not 'fairly distributed' is so unlikely scenarios can happen. If you only want the mean outcome for everyone, complain about using random distribution at all, not the quality of the algorithm. You might have actually had a point.
|
Well I have had four minipets as first bthday presents so far, they have all been different and all white, I would have liked better ones but its luck, yes i have had awful streaks of bad luck with chests, i opened 14 in a row, over two maps and got all purple, but thats luck, i presume you have 50/50 chance of getting one or other, my lockpicks have not broken about a third f the time with 20% chance, i agree with the other two posts its random it sucks at times, but that happens
Yes i would hate to get all the same pet for presents, but if i did i wouldnt say it was brken, just a tremendous amount of bad luck
then i would try to swap them or sell them

Yes i would hate to get all the same pet for presents, but if i did i wouldnt say it was brken, just a tremendous amount of bad luck
then i would try to swap them or sell them

Quote:
|
Originally Posted by bamm bamm bamm
For a start, random number algorithms are so close to random as to make no difference, including the probably standard and widely used one that Anet uses, but if you can write a better one without advocating a new PCIe card that measures background radiation, go ahead.
|
Quote:
|
Originally Posted by bamm bamm bamm
Secondly, your anecdotal evidence means not only less than nothing, but you seem to not understand that 'random' includes the events you describe. The whole point of 'random' and not 'fairly distributed' is so unlikely scenarios can happen. If you only want the mean outcome for everyone, complain about using random distribution at all, not the quality of the algorithm. You might have actually had a point.
|
Quote:
|
Originally Posted by bamm bamm bamm
Thirdly, do you think that people who got different pets come to this website and say 'Hey guys, I got different pets! I just signed up to say great work on that Random Number Generator, Anet! A+++." But because 20 people are going to post here saying they aren't on the trend line, you are going to have your ill-informed opinion reinforced. come back with 20,000 experiences and we'll have a look.
|
F
This link may interest you
:
http://www.rohitab.com/discuss/index...howtopic=24727
Hes using the current time to seed the random number generator. and generating the numbers before the clock ticks over a millisecond resulting in repeats.
But i would hope Anets abit more experianced than to allow that kind of bug.
:http://www.rohitab.com/discuss/index...howtopic=24727
Hes using the current time to seed the random number generator. and generating the numbers before the clock ticks over a millisecond resulting in repeats.
But i would hope Anets abit more experianced than to allow that kind of bug.
Quote:
|
Originally Posted by BlueNovember
If I toss a coin 10 times and get 10 heads, it does not mean the coin is biased. Same applies to not breaking 6 lockpicks or getting the same 2 minipets twice over.
Conversely, if I tossed a coin 100 times and got 100 heads the coin is certainly biased. Same applies if you had got ~20 lockpicks retained in a row. Granted, retaining 6 consecutive lockpicks is ~0.002% chance, but with hundreds of thousands of people using lockpicks all over the place... perhaps there's some dependance there? Either that or you got quite lucky. |
Would 100 straight heads-tosses be extraordinary? Absolutely. Likely? Not very. Possible? Yup. There are more factors that should be considered before considering the exercise biased, though.


