Curse

Rhamia Darigaz · Sep 27, 2008, 05:04 AM // 05:04

Quote:

Originally Posted by Aussie Boy

Nice numbers but it's still random imo

you, sir, make me want to punch babies.
nobody said it wasn't random. the thread is about the probability of getting a specific color mini from whatever function randomly generates it.
i don't think you know what the words "random" and "probability" mean.
enjoy your high school arithmetic.

Quote:

Originally Posted by farmerfez

it's probably just random because if it was 72% to get a white and only 3% for a green greens would be worth A LOT more.

what i said applies to you as well.

bamm bamm bamm · Sep 27, 2008, 03:45 PM // 15:45

Quote:

Originally Posted by Rhamia Darigaz

nobody said it wasn't random.

They usually do. Normally every idiot conspiracy theorist comes out of the woodwork to impart their keen insights when it comes to mini chance distribution threads. Maybe they're all playing warhammer.

SkekSister · Sep 29, 2008, 09:26 AM // 09:26

Three third birthday presents. Three Cloudtouched simians, 1 in pre-searing.

Add that to yer sample group

Shasgaliel · Oct 15, 2008, 10:08 AM // 10:08

Quote:

Originally Posted by SkekSister

Three third birthday presents. Three Cloudtouched simians, 1 in pre-searing.

Add that to yer sample group

I made additional test: I was opening two birthday presents at the same time with my friend. We both got two siege turtles each. That should be quite improbable according to the probabilities but actually we expected to get similar minis (just we did not know which mini we will get). Pity that there were no 3 of us to make a bit better sample. Half an hour later I open one more present but this time it was a jade.

Regarding Mad King Thorn and Mini Gwen. I think the script for presents attributes a given white to the presents open at the same time obtained at the same date however it just attributes a random green. So for greens there can be one more randomization included. So if you happen to open presents just obtained at the same time (so obtaining and opening time should be similar) I would that with high probability you will get two exactly the same whites/purples or in the case of greens two random ones (can be the same of course). I have no theory on the gold ones. Never got any so far nor seem anyone getting one. Actually pattern I posted below contradicts my theory. But It is just a start.

I think if we gather the data including not only the numbers but also times when a given present was obtained and opened we may get a nice estimation for future and better view on the rules behind presents.

Quote:

Originally Posted by Vann Borakul

But I didn't give any factors. The factors are unknown. The probability itself is what contains the factors. But it is most definitely virtually impossible to know these factors to manipulate your chances. Even if you could, that would only change 'your' probabilities, not the static probability.

http://guildwars.wikia.com/wiki/Drop...rthday_Present

Look familiar? Only the first 2 years have an acceptable amount of sum data to actually be useful.

The problem I see here that we are encountering something which is not "independent" probability. This kind of probability would mean that regardless what you do you get a random result. Here I think it is not the case. The real factors are indeed unknown but there are some. The weirdest pattern I saw so far was:

jade
kirin
kirin
jade
jade
jade
kirin

All chars were created at same time and presents open within 10 minutes. Still I do not know what to think about it.

Numa Pompilius · Oct 15, 2008, 01:27 PM // 13:27

Quote:

Originally Posted by Shasgaliel

The weirdest pattern I saw so far was:
jade
kirin
kirin
jade
jade
jade
kirin
All chars were created at same time and presents open within 10 minutes. Still I do not know what to think about it.

The result of a "sticky" random numbers generator: if it rolls an N it's more likely to roll N again the next roll than it should be. It doesn't throw the probability distribution of rolls over large series of rolls, because all outcomes are equally likely to "stick", but gives these odd winning/losing streaks.
Not an unusual problem with random number generators, but apparently often hard to pin down the cause of or fix. Or so I've been told.

It's been pretty obvious since the start that the random number generator of GW is slightly sticky; sometimes it works against you sometimes it works for you, and in the long run it evens out.