Originally Posted by
Hurin
But, regarding the RNG in general. . . we have been hearing these sorts of things for years. And on at least a couple of occasions, some people have tested the RNG in such a fashion that their sample size was actually appreciably large and thus useful in actually drawing some conclusions. Those conclusions were invariably that the RNG was operating as expected.
Funny, because you as player cannot test the RNG. At best you can test the /roll command or some other _application_ of RNG results, and even then you have to assume that the results are representive (not modified by untested factors). The raw RNG results are never visible to players, so it's impossible to verify any claims about correct/broken RNG. And even if the RNG itself is working as intended it's quite possible that its results are processed incorrectly for certain applications. Given that most in-game RNG applications cannot be practically tested in a controlled environment with a sufficient sample size it's again impossible to verify any claims that the results are broken/correct.
Unless Turbine releases technical specifics about the RNG and how it is used for certain applications (crafting, loot) including all influencing factors (very unlikely) this will always be a point of contention.
At some point Turbine stated that the same RNG state is used for the entire server. That would mean that most player-perceived streaks aren't necessarily streaks of the RNG itself, and actual RNG streaks cannot be noticed by individual players (unless they are alone on a server). Which again works against any player-made verification attempts.
Used to play: 85 Champ / Captain / Runekeeper / Guardian, Guild Master of everything but cooking.
Playing now: Hellcat / King Tiger / GW Panther / IS / KV-5 / M4 Sherman and more