Curse

rohara · Nov 24, 2007, 06:50 PM // 18:50

3 out of 5 of my birthday presents have been jungle trolls. bah.

bamm bamm bamm · Nov 24, 2007, 07:06 PM // 19:06

Quote:

Originally Posted by Crom The Pale

A random number is generated with an algorithm, given the same seed it will generate the same number every time. Most use a clock to generate the random number and some use multiple inputs(day X month X year X time = seed).

Microsoft's CryptGenRandom takes data from these sources, only one of them being time. If you set up an encrypted partition or container with TrueCrypt it gathers entropy from your continued mouse movements.

Edit:

Quote:

Originally Posted by Talach_Ninneed

Go into a cave with trolls, kill 1, chances are you got a raven staff
2nd try: Raven Staff
3rd Try:...Raven staff
... : Another white
... : Purple
... :white
...19th try: golden non-max item

you may get the gold on the 1st try or the 101st try, but golds have better drops then purples and purples have better then whites.

you were jsut unlucky that you got the most common and unwanted mini.
but hey, HoM fills quick with em

my 1 and only pressie was a seige turtle

The options you listed there are weighted differently, so they have different chances to drop. Because the OP is getting three of the same minipet, they have, by definition, the same probability of occurring. The chance of getting three the same is exactly the same as the chance of getting 3 different, (1/8)^3 (within the common set). No factorial as pets aren't removed from the set as they are received, which some people do request from time to time.

Nemo the Capitalist · Nov 24, 2007, 07:22 PM // 19:22

Op u got screwed but then again on 8 toons gotten 4 fungals 1 yr old eeach so gg

dsnesnintendo · Nov 24, 2007, 08:38 PM // 20:38

i am going to list a random series of numbers
123456789101112131415
seems odd but this outcome is just a likely as any other series of numbers
random is not subject to logical interpretation

Big_Daddy · Nov 24, 2007, 08:47 PM // 20:47

Fungals are THE ugliest yr 1 bday present. I'd rather have 4 trolls in a jungle row. :/ I feel for you man.

pumpkin pie · Nov 25, 2007, 05:26 AM // 05:26

just when i thought its safe to open a birthday present. yes one of my 16 character celebrated its birthday few days ago, and gotten a birthday present. I was reluctant to open the package, but curiosity made me do it, found out it was Black Friday yesterday when I open the package and out pop a little head, and its a cute mini fungal wallow, and I look around Kamadan every where I turn its fungal wallow... so I ran away to Shing Jea Monastery to find more fungal wallow roaming the monastery...

okay, so that was not true, but i did get my third fungal wallow yesterday, and I dare not open the next birthday package .... its sitting there in the bag. blue box with yellow ribbon. *sigh*

I hope its something I have not gotten before.

it wouldn't be as bad getting repeats mini pets if you can:
1) stack the pets.
2) let them roam in herd in town/outpost.

mwahahahahahah - i would become the necrid horse herder....

lyra_song · Nov 25, 2007, 11:55 PM // 23:55

heres my minis:

Ive had ONE repeat being the jades. The other multiples were promos.

It seems really random to me. Repetition is part of randomness, even imperfect randomness.

You can't see me · Nov 26, 2007, 01:11 AM // 01:11

When you gamble at slot machines, you have a chance of getting the cherries a lot because they're the most common even though it's a random spin.

You have a chance to get a Necrid Horseman, a Fungal Wallow, and a Whiptail Devourer a lot, because they're most common, even though it's a random draw.

Don't complain because when you draw straws there are more than one straw then the other. It may seem that the people on guru complaining about this are high in numbers, but considering the other thousands of players that are NOT complaining about it...

/notsigned.

MoldyRiceFrenzy · Nov 26, 2007, 03:54 AM // 03:54

u just got pwnt... bad...

Harmless · Nov 26, 2007, 04:12 AM // 04:12

I've gotten so many necrid horsemen I'm thinking about opening a stable.

Crom The Pale · Nov 26, 2007, 04:51 AM // 04:51

Quote:

Originally Posted by bamm bamm bamm

Microsoft's CryptGenRandom takes data from these sources, only one of them being time. If you set up an encrypted partition or container with TrueCrypt it gathers entropy from your continued mouse movements.

Edit:

The options you listed there are weighted differently, so they have different chances to drop. Because the OP is getting three of the same minipet, they have, by definition, the same probability of occurring. The chance of getting three the same is exactly the same as the chance of getting 3 different, (1/8)^3 (within the common set). No factorial as pets aren't removed from the set as they are received, which some people do request from time to time.

The question is where is the random number being calculated, on your comp or on the server? If we knew what info was being used to calculate the seed then it could, in theory, be manipulated.

Winterclaw · Nov 26, 2007, 05:09 AM // 05:09

One problem I've always had with these types of threads is unless you don't have some sort of stastical analysis of a large number of characters and gifts opened, you've got no clue if the thing is bugged or not. Then you've got some people saying something like getting three in a row is possible, which it is, but seem to think that saying good enough for the thing not to be bugged.

BTW, I got two of the same kind of minipet on two different occasions.

Feminist Terrorist · Nov 26, 2007, 08:13 AM // 08:13

Every time I read the thread title, I think it says midgets instead of minipets, and the first thing that comes to mind is "since when does GW have midgets".
I think it's a function of getting old; the eyes are the first to go.

shadows of hob · Nov 26, 2007, 08:24 AM // 08:24

I got 5 Fungals from 6 presents and 1 burning titan.

bamm bamm bamm · Nov 26, 2007, 08:32 AM // 08:32

Quote:

Originally Posted by Crom The Pale

The question is where is the random number being calculated, on your comp or on the server? If we knew what info was being used to calculate the seed then it could, in theory, be manipulated.

There's no way in hell it's client side.

Abnaxus · Nov 26, 2007, 08:32 AM // 08:32

I repeat another time, this is not a complain thread, I started it because I noticed something actually weird happening in a short timeframe.

It's not the 1st time I have duplicates, my 1st mini was a temple guardian, a couple of months later I had my 2nd on another account and it was again temple guardian.

Two temple guardians on 2 different accounts after months means nothing, the 2 events can't be correlatated in any way. They're completely independent.
I got other duplicates, but with such difference in time and place that it's clear that can't be any correlation between them.

Three in a row, in the same place (guild hall) and nearly contemporarily, made me think that the random generator can use a seed based on day/hour or location.
Someone else posted that his experience with multiple present openings in the same time and location gave similar results.

A question which could be interesting to be answered is whether the minipet is assigned when the present appears in the inventory, or when it's opened.

Abnaxus · Nov 26, 2007, 08:35 AM // 08:35

Quote:

Originally Posted by shadows of hob

I got 5 Fungals from 6 presents and 1 burning titan.

Did you open them in the same time or with some days/week/months of interval?

strcpy · Nov 26, 2007, 08:52 AM // 08:52

Quote:

Originally Posted by Abnaxus

A question which could be interesting to be answered is whether the minipet is assigned when the present appears in the inventory, or when it's opened.

If you simply mean interesting from an operations standpoint then I would agree - anything we can learn in that sense is interesting. If you mean interesting in checking on the likely hood of getting the same pet then it is irrelevant, PRNG do not work that way.

A PRNG will *only* give consistent output if the seed is the same (even the horrid ones like the old RANDU work this way). The only way this would work is if they seed on something with a large enough granularity that they have the same seed over several present opened and *also* re-seed before every roll. There are *very* few things that have this level of granularity in any of the examples given so far (in fact, I find none of them do however I will assume there is somethings I have not thought of).

Once again, more than a simple basic understanding of PRNG's would help quite a bit. While it is always possible they do something that stupid (or even meant it to be that way) it generally follows that if one is capable of writing this video game they are more than knowledgeable enough to avoid that bug. In fact, this would be along the lines of a car designer forgetting to add an engine to the car. While anything is always possible I highly doubt it.

Now, if the afore mentioned Schroedinger's Cat experiment is to be believed then the answer is apparent - it is decided once opened and someone self-aware looks at the outcome.

tmakinen · Nov 26, 2007, 09:11 AM // 09:11

Quote:

Originally Posted by Abnaxus

A question which could be interesting to be answered is whether the minipet is assigned when the present appears in the inventory, or when it's opened.

This is the 'Schrodinger's wallow' question, and the answer is that the contents of an unopened Birthday Present are in a superposition of all possible minipets for that particular year, and the wavefunction only collapses to a single minipet when the present is opened. I accidentally proved this by opening a present just before a game wide rollback. Naturally, after the rollback I had the same unopened present in my inventory, but this time it produced a different minipet than on the first time. Had the contents been determined when the present was created, I would have gotten the same minipet both times. And yeah, if the game determines such outcomes on client side then my name is Gaile Gray

In contrast, the loot possessed by monsters is determined when you enter a zone. This was proven elegantly by receiving time restricted drops outside the valid time frame because the zone was entered (and thus loot created) while still within the time frame.

strcpy · Nov 26, 2007, 10:10 AM // 10:10

Quote:

Originally Posted by tmakinen

In contrast, the loot possessed by monsters is determined when you enter a zone. This was proven elegantly by receiving time restricted drops outside the valid time frame because the zone was entered (and thus loot created) while still within the time frame.

Ah, but you see this is like the cat that the atom didn't decay - it would only remember being alive once observed and forget any of the other states it was in whilst in the box (and thus appear to have always been in a single state to the observer). In another universe you got a different piece of loot and and wrote something totally different

The rollback thing is interesting and obviously proves that the contents are only decided when opened - there can not have been very many times that experiment has been done. Would you care to expand on the latter? I would be interested to know the exact circumstances and if you have repeated it more than once. I would have imagined that different types of loot have different ways they are determined.