Pointless argument about chance and randomness

"

mibuwolf wrote:

"

Deccode wrote:

You are like ignoring everything. Let me ask you again... Are manipulated events random?

In a way... yes. A biased random variable.

Wrong. It can only be biased on it's own which it does but must not be manipulated else losing the therm randomness.

"

Sickness wrote:

"

Aelloon wrote:

Since this is an active thread probably on the topic:
What is the chance of making an item 6 socket with jeweller's and what's the chance of making a 6 socket 6 link or 5 link with fusings? I've read posts about it somewhere but this forum doesn't have really have a search function.

I believe it's 0.33% to make a 6 socket with jeweler and 0.33% to link all sockets with fusings.

Not sure if this is how it is done here or how you think it should be overall. Not only could you calculate it by using your own link or (here it's easier for you http://www.youtube.com/watch?v=auAvt7CIezM), No you could also calculate it in head. I'm not sure if you can reroll the same number of sockets of after using a jeweler orb (I think it happened to me) or you can make the item without sockets at all. So if you have an item without sockets and you use an jeweler orb there are 6 possible outcomes. Therefore the chance to get a 6 socketed item is 1/6. It's simply simple even proved by the link you are praising and repeating.

"

Randomness implies that the outcome is equally likely to occur proportional to the number of outcomes, whereas general probability allows any outcome to have bias or different chances of occurring.

let's take back this picture :


A and B are the two possible cases for result 1
C, D, E, F and G are the 5 possible cases for result 3
H, I and J are the 3 possible cases for result 7

now, randomnes implies ( from what you say ) that the chance to get I is equal to the chance to get B.
Alright, this is perfectly normal, the chance getting either of these are the same.
What really matters are the odds of getting ... like H OR I OR J over the rest, and there the chances for this outcome changes.

"

Therefore the chance to get a 6 socketed item is 1/6. It's simply simple even proved by the link you are praising and repeating.

/facepalm

"

TehHammer wrote:

Jeweler's odds are correct, there. It's 1:306 to roll a 6S, 5:306 to roll a 5S, 30:306 to roll a 4S, 80:306 to roll a 3S, 90:306 to roll a 2S, and 100:306 to roll 1S (this is assuming the item is capable of rolling 6 sockets). Catch is, they can't roll the same amount of sockets as the previous roll, so when an item is sitting on one socket, odds are 1:206 for a 6S, 5:206 for a 5S, etc. I actually wrote a tool to simulate using a boatload of jeweler's on an item.

those numbers seem weird to me, didn't get into those calculations in details, so do you have the formula to calculate this ? How do you get 100:306 to get 1S item ??

"

mibuwolf wrote:

Double posted apparently, so I'll use this space just... because:

"

Kreos wrote:

Despite how horribly Deccode is presenting himself, he's actually correct. The difference you're arguing is randomness versus chance. Your picture is depicting the CHANCE that one of the outcomes occur, and the chance of these outcomes are not equal. However, just because a system relies on chance does not make it random. Randomness implies that the outcome is equally likely to occur proportional to the number of outcomes, whereas general probability allows any outcome to have bias or different chances of occurring.

http://www.stat.yale.edu/Courses/1997-98/101/ranvar.htm

If that isn't considered random AT ALL then I'm afraid I would have to be shocked and appalled.

Let me make it clear though: I'm arguing that a random variable has two ways it can be formed. On a curve which is a biased outcome (certainly not equal), or it can have a uniform curve (in which it is fair).

You are stating that we are wrong to believe random variables can have a bias, but you insist that random variables MUST ABSOLUTELY have an equal probability to be considered a random variable?

So here is the disconnect that the last 5 pages have been created for: you and Sickness are arguing random variables while Deccode is arguing randomness in general. A random variable isn't necessarily random: it is a variable that changes based off of chance. A random system by definition will have its outcomes equally likely to occur proportional to its number of outcomes. The link you linked is talking specifically about random variables that each have a different probability of being selected.

"

Fruz wrote:

/facepalm

See one more example which has no idea what he is talking about. Please use the youtube video I just posted it's so simply explained that even you could understand it. I say that the chance of getting 6 sockets must be 1/6 to be called random else it's not.

First like I found after googling :
http://www.icoachmath.com/math_dictionary/Chance.html

"

Chance

Definition of Chance

The likelihood of the occurrence of a particular event is called a Chance.

this brings me back to my previous statement

/FACEPALM

"

Fruz wrote:

let's take back this picture :

This is correct. The odds I stated for jeweler's orbs are the ones the devs gave us over a year ago, and I believe are still accurate. Basically a random number is generated, between 1 and 306, and based on that number, a specified socket count is placed on the item. Certain outcomes are blocked (rolling the same number of sockets) and therefore a new random number is generated. It is completely random, however, the results of that random number (how many sockets are rolled) are weighted.

This is what Deccode is missing, it is random, but the results of that random event are weighted.

"

Deccode wrote:

You are like ignoring everything. Let me ask you again... Are manipulated events random?

That depends how they are manipulated. Don't ignore my questions.

"

Kreos wrote:

So here is the disconnect that the last 5 pages have been created for: you and Sickness are arguing random variables while Deccode is arguing randomness in general. A random variable isn't necessarily random: it is a variable that changes based off of chance. A random system by definition will have its outcomes equally likely to occur proportional to its number of outcomes. The link you linked is talking specifically about random variables that each have a different probability of being selected.

Correct, but it is still considered random. In statistics random is defined as an equal outcome, that much is true.

Biased outcome can still be considered chosen at random. I would rather the argument just end to a "let's agree to disagree" and be done with it.

"

Fruz wrote:

"

Randomness implies that the outcome is equally likely to occur proportional to the number of outcomes, whereas general probability allows any outcome to have bias or different chances of occurring.

let's take back this picture :


A and B are the two possible cases for result 1
C, D, E, F and G are the 5 possible cases for result 3
H, I and J are the 3 possible cases for result 7

now, randomnes implies ( from what you say ) that the chance to get I is equal to the chance to get B.
Alright, this is perfectly normal, the chance getting either of these are the same.
What really matters are the odds of getting ... like H OR I OR J over the rest, and there the chances for this outcome changes.
\

But what you're doing in this example is taking a random roll and then manipulating the result into your system. For example, lets say that the .33% rule is correct, then using your example I roll use whatever chaotic function GGG is using to make a pseudo-random number and restrict it to 1000. So i have a basically random number 1-1000. That is the random aspect, however then I say if the number is 1-33, it's 6 slot, 33-100 5 slot, 100-400 4 slot, etc etc etc (made up numbers, just illustrating). So what you're actually doing is manipulating the truly random variable and then calling it still random.

"

Kreos wrote:

But what you're doing in this example is taking a random roll and then manipulating the result into your system. For example, lets say that the .33% rule is correct, then using your example I roll use whatever chaotic function GGG is using to make a pseudo-random number and restrict it to 1000. So i have a basically random number 1-1000. That is the random aspect, however then I say if the number is 1-33, it's 6 slot, 33-100 5 slot, 100-400 4 slot, etc etc etc (made up numbers, just illustrating). So what you're actually doing is manipulating the truly random variable and then calling it still random.

Correct. The outcome is still "unpredictable". Although weighted, we still cannot predict the outcome because it would be impossible.