What are odds of chancing high tier uniques now?

"

KromBoy wrote:

If GGG released the official numbers, they would be affecting the actual trade value of resulting items. I.E. If they released an official figure on number of fusing it takes to 6L an item, it takes away from the community coming up with a price that has variance. Instead, most would come to a consensus on how much it should be with no flexibility on the price.

Even with the mean figures for rolling set numbers of sockets and/or links being known I do not believe they will affect the current pricing in the market. The current pricing of items & orbs is already based on the sample mean in game - actually knowing the exact mean won't change (much) the probability known in game and, at least I believe, not affect the current pricing much. Also, importantly, even given a mean the amount of variance that actually occurs in samples the size of this player base is large and will always be taken into account by risk adverse/loving people.

Take for example rolling 4L on a 4 socket item, the mean for getting this is about 12 Fusings. Around 5% of people will succeed on their first Fusing and at the other end of the spectrum 5% of people will require about 50 Fusings. I just feel really sorry for the 1 in 100 people that needs to use about 80 or the 1 in 1000 person who needs to use about 120 Fusings. Given enough people there is technically no upper limit to this either (sort of).

"

andkamen wrote:

It has a mean- an average number that it should take.And while RNG is RNG for every individual roll, the sum of a lot of rolls should begin to form a pattern. (just like it takes roughly 130-150 fuses for 5L but a some people get it in 15 and some like me have needed 250+)

As you alluded to literally right before the bolded section every roll is independent. I can't quite tell if you are mistakenly applying the Gambler's Fallacy; you should really read about it here this article. Sufficed to say though you cannot use previous data/rolls to make assumptions about future rolls especially on a martingale.

Edit: Just for the record I'm actually mildly against GGG releasing the probabilities. We already have a fairly good idea of what they are (without use of Scraps/Whetstones) through thousands of trials in game and if the figures are released there would be a deluge of misinformed players complaining they didn't get their item X linked in Y Fusings (which we already get to a certain extent).

"

arual wrote:

"

andkamen wrote:

It has a mean- an average number that it should take.And while RNG is RNG for every individual roll, the sum of a lot of rolls should begin to form a pattern. (just like it takes roughly 130-150 fuses for 5L but a some people get it in 15 and some like me have needed 250+)

As you alluded to literally right before the bolded section every roll is independent. I can't quite tell if you are mistakenly applying the Gambler's Fallacy; you should really read about it here this article. Sufficed to say though you cannot use previous data/rolls to make assumptions about future rolls especially on a martingale.

No, he's just saying that analysis of the outcome of a high volume of independent events would suggest that there is most likely a gaussian distribution of outcomes, centering around a mean. I don't see how he alluded to the idea that analysis of past outcomes can estimate the results of one or more specific future outcomes (Gambler's Fallacy).

Edit: It's important to note that he's responded to a post stating that even if the odds are 1 in 2 one could see 2000 consecutive identical outcomes, a statement that is true but so astronomically improbable you'd have a better chance of farting rainbows.

"

superclove wrote:

No, he's just saying that analysis of the outcome of a high volume of independent events would suggest that there is most likely a gaussian distribution of outcomes, centering around a mean. I don't see how he alluded to the idea that analysis of past outcomes can estimate the results of one or more specific future outcomes (Gambler's Fallacy).

Edit: It's important to note that he's responded to a post stating that even if the odds are 1 in 2 one could see 2000 consecutive identical outcomes, a statement that is true but so astronomically improbable you'd have a better chance of farting rainbows.

Having reread the comment and what he quoted it looks like I did indeed get the wrong end of the stick. At least I left myself the wiggle room of the "I can't quite tell..."!