FR EN DE ES IT PT
Browse forums 
Ankama Trackers

Your dice rolling program is broken

By Thnag August 10, 2016, 18:10:29

yup...

As usual your dice rolling program doesn't even remotely reflect what you would expect based on probability.

After reviewing the attack and defense rolls of my past 5 games in a row I found my opponents got a magical 70% success rate on defense rolls (Averaged across 5 games, some were higher). 5 games in a row. This success rate wasn't just rolling armor, but me not rolling critical with the critical power even remotely close to the expected 75%. They magically roll armor all the time with only 1 armor roll (3 armor in the case of Gein against 2 critical dice when they finally crit.)

I'm not even going to go into the opponents magical attack rolls.

No one can be that lucky or unlucky. Your dice rolling program must be broken.

If you want people to stop complaining about your dice program, start showing all of the dice rolls again (like the original game) and prove its working properly.

2 -1
Reply
Reactions 6
Score : 166
Thnag|2016-08-10 18:10:29
[...] No one can be that lucky or unlucky. Your dice rolling program must be broken.

If you want people to stop complaining about your dice program, start showing all of the dice rolls again (like the original game) and prove its working properly.

Yes, people can actually be that lucky or unlucky; it's all a matter of variance (that is, short-term deviation from the expected value of a random variable, look it up).

The dice-rolling mechanics, as far as I know, are well implemented (excluding a possible RNG bug which could NOT be noticed from outside the server and that is not likely to even exist in the first place).

People are, most often than not, biased to believe that short-term strokes of bad luck (yes, five matches is still short term from a statistical point of view) are failures for which the RNG is at fault, when in fact they are part of its normal behaviour (on the other hand, a random number generator which gave consistent results over a given number of matches would probably be filtered and, therefore, not random at all).

As a final note, the results of the dice rolls may already be viewed from the match log (on the upper part of the screen after pressing the "log" button).
0 0
Reply
Score : 676

The actual probabilities (1/6 per die face) will start to even out with N > 300

I don't think you've done over 300 rolls in 5 matches.

Still, the way the online version works, I think it's pretty fair to assume it uses some standart pseudo-RNG algorithm which is prone to unexpectedly biased results.

TL;DR:
the grounds for your claim are probably false
however, the dice rolling program may indeed be badly implemented

0 0
Reply
Score : 166
Hmmmvot|2016-08-11 14:01:18
[...]

Still, the way the online version works, I think it's pretty fair to assume it uses some standart pseudo-RNG algorithm which is prone to unexpectedly biased results.

[...]

With only six possible results (one per die face), it is hard, even for a pseudo-RNG, to have its output affected by noticeable long-term trends.

Anyway, I completely agree with you on everything else; the original poster is probably just down on their luck.
0 0
Reply
Score : 1102
Qalandar|2016-08-11 18:54:47


Anyway, I completely agree with you on everything else; the original poster is probably just down on their luck.
So does that mean i have just been "down on my luck" since i have started playing this game? Because i have always and still have this on most of my matches and it is extremely rare when i actually get lucky like my opponents usually do.
0 0
Reply
Score : 676
wowxxxxx|2016-08-13 06:04:34
Qalandar|2016-08-11 18:54:47


Anyway, I completely agree with you on everything else; the original poster is probably just down on their luck.
So does that mean i have just been "down on my luck" since i have started playing this game? Because i have always and still have this on most of my matches and it is extremely rare when i actually get lucky like my opponents usually do.
Please record all your rolls and provide us with sufficient data to make a conclusion. Like I said, information on 300+ rolls is necessary.
Otherwise, yours is an empty claim backed only by the brain's in-built heuristics (i.e., the mind tends to forget the successful rolls because you take them for granted, while focusing on bad roll because they frustrate you).
0 0
Reply
Score : 387

In some real life games, i managed several times to deal 2 damage to an opponent's un-armored Krosmaster with 3 attacks from a Crit-equipped Ayan (so yeah... 9 dice with 50% success probability each vs 3 with 33.3% each, do the math. or don't, probability is dumb :p)
Life is broken as well it seems...

anyway, luck is luck. you can somehow manipulate your luck by not attacking Armored foes without critical or armor-piercing attacks for exhample, or build a team specialized with armor-piercing attacks + Malee Bhurhum/Kriss Krass if you really hate luck getting into your way (like i do) at the cost of having some really bad matchups against heavy-ranged/summons teams, but bad periods happens when there's a random components either in virtual or real life
So far i've been having much, let's say, mathemathically coherents results with the online RNG then i have with real life games

but hey, it's luck again in the end, every person has its own

0 0
Reply
Respond to this thread