Quote Originally Posted by Felis View Post
You called? ^^

It would be so much easier if everyone collect some datas about succes versus fail rate (and not about fail rate only)

I did 3 leves to have always the same % chance to test it
Here my result:

8 nodes *6 strikes = 48 strikes each leve
official chance = 89%

46/48 (95,8%)
42/48 (87,5%)
43/48 (89,5%)

Total:
131/144 (90%)
The problem most people are encountering isn't the overall odds, it's the strings of good and bad 'luck' that lead to those odds. Getting a string of 5+ failures when you had a 90% success chance is very unlikely but it could happen. Having it then happen multiple times in one game session AND seeing other players also have it happen is beyond improbable. My FC chat every night gets multiple players reporting huge strings of bad luck like this yet the odds of it happening should make it so rare we only see it happen to one person a week or so...

It's probably worth pointing out though that I see equally crazy luck with successes sometimes... more or less to balance it out. To give an analogy it seems to be more like tossing a coin 100 times; you'd expect roughly 50 heads and 50 tails (give or take a few, and assuming the coin was flipped exactly the same way every time). Now the RNG here would still give us roughly 50 of each, but it would output them as sequences of 5-10 of each in a row with a rare few scattered in between. Do you see the problem? Their algorithm seems to 'stick' to a certain end of the spectrum for several rolls in a row. The overall result is correct but the way it gets there is very very wrong.