No…
So the group luck is useless on right upper corner icon ?
The Strength In Numbers buff for playing in a group will increase the number of legendaries you get. It does not affect the proportion of them that are legendary, ancient or primal.
The conclution is; the more time you play the “luckier” you get
So i made my re -rolls and here are the results:
Total Re -rolls: 405
Total Ancients: 33
Total Primals: 0
!!!
Tell me how and i will post an excel file in here…
10% drop rate ? Bullshit !!!
1st one on 18th re -roll
6th ancient got on 132 re -roll
As i told u before, nobody know how this drop mechanism works ! Even D3 developers, there still may be some issues in my profile also, so please check it, broken file, corrupted, cached, i do not now, but this is starting to be very annoying !!!
Saying that u got 10% chance on ancient means that in 20 rolls ull guaranteed to get 2 ancients, but not to get 6 ancients in 132 re -rolls !!!!!!!!!!!! And u know what, im pretty tired of this, i clearly understand that there will be absolutely nothing done by this, so im gonna spread statistics and video on all communities over the internet, thats enough !
Its a second time im asking to check my diablo 3 character files, second time, and nothing in reply, so ok !
Let’s use this… Binomial Distribution Calculator
Number of Events = 405
Probability of success per event = 10%
Probability of getting between 20 and 60 successes = 99.90%
Probability of getting between 25 and 55 successes = 98.90%
Probability of getting between 30 and 50 successes = 91.77%
Probability of getting between 35 and 45 successes = 63.82%
Probability of getting exactly 40 successes = 6.608%
So, around 91-92% of people would have gotten a result in the 30-50 success range at a 10% success per event, after 405 events but only 6-7% of them would have gotten exactly 40.
No it doesn’t. Not even close. You’re describing the Expected Value.
Anyone that has even the vaguest understanding of statistics will make you look silly when they respond to you then. You might want to read these before doing that…
Law of Large Numbers
Expected Value
By the way, even if / when you get primals, they might not be useful…
1 Like
Anyone that has even the vaguest understanding of statistics will make you look silly when they respond to you then. You might want to read these before doing that…
Really ? )) dude, i do not know who u are, u are working on bliz support or u r just an advanced forum user and u may teach me on how to behave in diablo or what ever, but please do not teach me on Probability theory, and sorry that was a misclick, 10% means not 11 in 20 rolls but 2 guaranteed sorry
diablo 3 drop completely breaks all statistical physics laws ))) And please not a word about statistics im master degree in statistical physics and data analytics !
Between 51 and 132 there are 81 re -rolls which means i had to get at least 8 ancients with officially announced 10% ancient chance drop , so where are they ? Hm ?
Please to not force me to connect IBM analytic tools and whatson so prove that all this drop rate is configurable to keep user in -game !
So, around 91-92% of people would have gotten a result in the 30-50 success range at a 10% success per event, after 405 events but only 6-7% of them would have gotten exactly 40.
So that means that drop rate is not randomized, but configurable and depending on some conditions, isn’t it ? if im wrong, please share the exact drop rate calculation formula in here ! Otherwise all that u have said is completely breaking all this universe understanding of probability theory.
A little prove, just imagine i had only 152 bounty mats and got 6 ancients at roll time, is this still 10% chance ? ))) the actual chance is varies between 1.2% till 8% which means there are some conditions to up chances or get them all down, what kind of conditions there are ? what they are depending on ? I suppose nobody really knows this too, cuz d3 source code documentation was lost in flow of employment changes…
and again, why should i beg for support to check my profile ? Now this is silly, not the action of spreading stat information, that u really do not understand, all over the internet !
Please, check my account for some errors, it works incorrectly !!!
It doesn’t mean anything like that…
…which shows you’d have a 28.52% chance of getting exactly 2 successes, i.e. around 7 in 10 people wouldn’t get that result.
You have a fundamental misunderstanding of the difference between the Expected Value (i.e. the average arrived at over a huge number of events) and the short-term variances you can get from that value in small numbers of events.
Your results are exactly what we’d expect to find in a standard binomial distribution, i.e. it’s working just fine.
This is completely wrong, except that situation if probability spreads between all users, not a single one !
And that is why im getting a primal item at 1 level of monk with 18 rings schema ? )))
And another one thing, if so, do announce that we are using binomial probabilities in-game drop rate so can i consider that the last one was an officially announcement ?
Overwise, please share drop rate calculation formula !
It doesn’t matter if 20 people have one event each, or one person has 20 events, the overall possibility of number of successes is the same for either of these.
Did you miss it the first time around…
% Probability does not guarantee anything, …not ever
1 Like
Ooh, that gets us to Confidence Levels.
Confidence levels are expressed as a percentage (for example, a 95% confidence level). It means that should you repeat an experiment or survey over and over again, 95 percent of the time your results will match the results you get from a population (in other words, your statistics would be sound!).
So, how many reforges would be required to have a 95% likelihood of obtaining a primal? We’d need to solve this equation for X…
(399/400)^X =< 0.05
The lowest value of X that satisfies this is 1197.
This means that if 1197 reforges were done, 95% of people would get at least one primal in the results.
2 Likes
My confidence level is kind of; If I’m reforging a Flavor of Time 1197 times - I will get a primal Duncraig Cross 
As we saw it did not work for me, i got 405 rolls and zero primals )
And so, if you are using binomial probabilities it must obey to the density probability and binomial distribution laws and it means that - sequence length is 150 numbers maximum, which already screws the total chance of 400 count rolls as it was divided by 150 each, 150+150+100…lets go farther…just for your information, the binomial distribution is completely discreet… for diablo 3 primal drop rate we need to use Poisson approximation cuz “n” - numbers of rolls are much higher that the probability “P” to get one.
Using Poisson formula and lets insert variables we`ll get - P(X=k) —> L^k /k!^e-L, where k=number of rolls. Lets count:
as u told me before 1/400 - 1 primal which means 0.25%
sequence length is still 150 and we get:
expected values: 3.75 dispersion: 3.65625%
1 primal in 1 roll chance is 0.22422929671430475 %
1 primal in 400 rolls chance 89.6917186857 %
1 primal in 405 rolls chance 90.731~%
It all makes sense now, we have figured out that snapshot it counted only by 150 steps, not with 400 / 500 / 1000 but 150 ! ( limits of classical binomial distribution ), but in practice we are facing absolutely different results.
Poisson approximation does not work here cuz according to my rolling statistics i get a very big dispersion:
5th and 6th ancients were made between 81 sequence length (81 rolls), let me explain that to u, it does not mean that i got 6 ancients in 81 rolls, it means that i got 5 ancient on 51th roll and 6th ancient i got on 132th roll, 81 sequence length !
29th and 30th ancients were made between 63 length (63 rolls)
With dispersion = 3.65625% it is simply impossible to have 81 sequence length per snapshot (150 rolls), it means u have about 54% of calculation error or there is something very wrong with my account. Otherwise that makes me think that all those numbers are adjustable / dependable on some unknown conditions !
If im wrong, please share the mathematical calculation formula of diablo 3 drop rate with explanation !
I suppose it should be something like this:
Mi(t)=(P/1-qe^t)^r —this is Pascal`s negative binomial distribution function with moments, just an example, and i really do not want to believe you are using this one on drop rates )))
Before answering any questions of yours, please answer to my last question, what is the exact mathematical formula of drop rate calculation ?
it took you almost a week to come up with a wall of text that still shows you don’t know that an Expected Value (such as 1 in 400) is the average when huge numbers of trials are done. Your 1600 or 405 attempts in no way even approach a sufficiently statistically viable population to interpolate to the Expected Value.
1 Like
i got more important business to attend to…Calculation formula please )
if i cannot prove u that it works wrong, will make it through that formula, otherwise there is a huge gap between us in understanding probability theory…
This is 100% correct.
2 Likes
Calculation formula please )
1 in 400…
1 Like