• Dear forum reader,
    To actively participate in our forum discussions or to start your own threads, in addition to your game account, you need a forum account. You can     REGISTER HERE!
    Please ensure a translation into English is provided if your post is not in English and to respect your fellow players when posting.
  • We are looking for you!
    Always wanted to join our Support or Forum Team? We are looking for enthusiastic moderators!
    Take a look at our recruitment page for more information and how you can apply:
    Apply
  • Forum Contests

    Won't you join us for out latest contest?
    You can check out the newest one here.

Biased AI ??

Cursedveggie

Cursedveggie

Legend
StarLoad said:
i already have my arc at lv 81 lol
what i was surprised was how getting even number of bp i can get, my arc is already 81 and unlockable to over lv 500
Click to expand...
Keep going :)

20210325_062553.jpg
 

=Susharik=

Private
It does not make sense. If AI is not biased and a blueprint probability is equally distributed then a chance of getting a specific blueprint is 1/9 and not getting it is 8/9. In my case, I collected 43 Arc blueprints. On none of the occasions, I did not get the one missing blueprint. The probability for such an event is 8/9 in a power of 43, which is about a 0.006 chance. This is extremely unlikely. Looks like AI really wants you to spend diamonds on the last blueprint.

FoE Arc BPs.jpg
 
PeePee Pleb

PeePee Pleb

Lieutenant
=Susharik= said:
It does not make sense. If AI is not biased and a blueprint probability is equally distributed then a chance of getting a specific blueprint is 1/9 and not getting it is 8/9. In my case, I collected 43 Arc blueprints. On none of the occasions, I did not get the one missing blueprint. The probability for such an event is 8/9 in a power of 43, which is about a 0.006 chance. This is extremely unlikely. Looks like AI really wants you to spend diamonds on the last blueprint.

View attachment 23143
Click to expand...
Well, many times I was extrenely lucky that I got just the needed BP for a GB, so I'm pretty sure it's just a RNG thing, nothing more.
 

Paladiac the Pure

Major-General
=Susharik= said:
It does not make sense. If AI is not biased and a blueprint probability is equally distributed then a chance of getting a specific blueprint is 1/9 and not getting it is 8/9. In my case, I collected 43 Arc blueprints. On none of the occasions, I did not get the one missing blueprint. The probability for such an event is 8/9 in a power of 43, which is about a 0.006 chance. This is extremely unlikely. Looks like AI really wants you to spend diamonds on the last blueprint.
Click to expand...
Big deal, quite often by the time get the last missing BP, some of the other spots are already in double digits (i.e. 10 or more). Other times, only have two or three in each of the other slots. It is only RNG, no bias at all, except to those who do not understand what RNG is about.
 

Knight of ICE

People always complain when that last BP eludes them. In that case there must be something wrong with the RNG. There has to be something wrong with it, cause they deserve that BP, so Inno must be doing something to it, to force them to spend diamonds. Strangely enough there never is something wrong with it, when they hit that BP first time. First person to complain about that has still to be born.
 
klods hans

klods hans

Tsar
=Susharik= said:
It does not make sense. If AI is not biased and a blueprint probability is equally distributed then a chance of getting a specific blueprint is 1/9 and not getting it is 8/9. In my case, I collected 43 Arc blueprints. On none of the occasions, I did not get the one missing blueprint. The probability for such an event is 8/9 in a power of 43, which is about a 0.006 chance. This is extremely unlikely. Looks like AI really wants you to spend diamonds on the last blueprint.

View attachment 23143
Click to expand...
Randomness is not affected by what happened previously.
 

=Susharik=

Private
it is not a 6% probability, but 0.6%. For those that know statistics, if RNG is working correctly then with a lot of occurrences, you should get to exactly 1/9 of all occurrences (+- the variance). Of course, you can say that a set of 43 events is not enough for this rule to work, and I will agree. But when you multiply by many players that experience the same issue I would say that Inno RNG IS BIASED.
 

RegV

Corporal
=Susharik= said:
it is not a 6% probability, but 0.6%. For those that know statistics, if RNG is working correctly then with a lot of occurrences, you should get to exactly 1/9 of all occurrences (+- the variance). Of course, you can say that a set of 43 events is not enough for this rule to work, and I will agree. But when you multiply by many players that experience the same issue I would say that Inno RNG IS BIASED.
Click to expand...
Actually, we're both wrong. You are right, I dropped a zero. I did that because a while ago I wrote a simple simulation for this, using a simple RNG just like INNO uses, and when I plugged in 43, it came back with a 5.6% chance of having an empty slot. So I mentally dropped off a 0 and chalked off the slight difference to roundoff and only doing 100K simulations so it hadn't converged yet, It was close enough to 6%.

I bumped my simulator to 10M simulations and got 5.616% chance of having an empty slot with 43 BPs. Odd that, at 10M it should converge. Had to think about it a bit.

But your math is flawed. You are computing the odds of a *specific* slot not having a single BP. But you aren't here to complain that the last slot is still empty. It could be the first slot or middle slot or any of the others, you would still claim bias,
Compute the probability that any 1 of 9 slots is empty, and I'll bet you come up with something around 5.6%

To correct my original statement:
I don't consider something that has a 5.6% chance of happening to be "extremely unlikely". I don't even consider something that has a 0.6% chance of happening "extremely unlikely". Do you?

Never attribute to malice that which is adequately explained by math.
 

=Susharik=

Private
From information posted by other players.
Let's take this guy for example (image below). Using simple Binomial Distribution Calculator the probability for getting n or fewer successes for the blueprint marked in blue:

CdM - total occurrences 69 - probability 0.30684% - mean number of successes 7.59
CoA - total occurences 66 - probability 0.4183% - mean number of successes 7.26
Arc - total occurrences 65 - probability 0.4637% - mean number of successes 7.15
CF - total occurrences 50 - probability 7.633% - mean number of successes 5.5
HS - total occurrences 31 - probability 13.037% - mean number of successes 3.41
Traz - total occurrences 53 - probability 1.569% - mean number of successes 5.83

Let me explain what is statistical (normal) unlucky: take the mean number of successes and subtract the normal deviation of 2.5-3.0. Only CF and HS look statistically unlucky. The rest of GBs are far beyond statistically unlucky. This is a clear indication of a BIASED RNG.

FoE Arc BPs 2.jpg
 
Last edited:
You must log in or register to reply here.
Top