My new rotation and RNG

[Train88]

-Comfort Zone
-Inner Quiet
-Steady Hand 2
-Waste Not 1
-Hasty Touch x4
-Manipulation
-Carful Synhesis 2
-Steady Hand 2
-Hasty touch x2
-Manipuation
-Hasty Touch x2
-Steady Hand
-Hasty Touch
-Great Strides
-Byregot's Blessing
-Carful Synthesis.

(with 51-55 CP food, 345 base)

I somehow get 10 on 70 durability crafts. (did i count belssing? might have) Point is 10 touch, 80% chance, 6 fails... Now i wouldn't bother making a thread if it happened once in a while but i get this kind of stuff ALL THE TIME. Like regularly.

I did statistics in college and this kind of stuff does not happen regularly. Not unless the aliens are invading soon. Why? This is not random is it?

[Hayward]

My suspicion is that the developers' RNG has been weighted more against success than we should expect. 4/5 success rate should never allow streaks of 4 or more failures in one synth, nor should it allow those streaks over multiple synths. A lot of people on these forums have complained too much about the lack of rarity for anyone to not suspect SE is catering to these malcontents.

[Train88]

At this point i'm pretty sure i had 9 hasty touch on a 35 durability item and i failed 66%. I got those huge failure streakes before on mats, i think everyone has... but it's never been out the window like that. 33% vs 80%? Would it ever balance out on next items?

[raela]

Er, no, that's how it works. Yeah it sucks, but RNG is RNG. Read up on this https://en.m.wikipedia.org/wiki/Gambler%27s_fallacy

We're just more likely to recall when RNG was awful, rather than when things went better than expected.

[NorthernLadMSP]

It wouldn't surprise me one bit if it was more weighted to fail than succeed. It's the same with gathering.

97% success on a node

You fail
You fail
You gather
You fail

[Train88]

Someone who paid any attention to those Statistic classes in college could come up with a court class proof that there are chips missing in this bag. Just claiming 'itz RNG, lol' is a little bit hollow for an explenation. Not likely to get any anyway.

EDIT: also /tinfoil hat

[raela]

I have seen data posted before of a large number of crafts, demonstrating the %s do stand. With the number of crafts the average person does, yes, it can look skewed. You need tens if not hundreds of thousands of RNG rolls to see the how the percentages fall. People are terrible with probability.

[Train88]

Yeah, i had to take the course twice to pass. Can't say i remember anything about it really.

[Shadygrove]

Please remember that computers just can't do truly random numbers. They do psudeo random numbers. And Squenix's psudeo random numbers are known to be streaky. So missing 3 of 4 at 90% or making 2 of 4 at 10% is not at all unusual.

But in the long run, and 1,000 is NOT the long run, it will all balance out.

[LineageRazor]

Please remember that computers just can't do truly random numbers. They do psudeo random numbers. And Squenix's psudeo random numbers are known to be streaky. So missing 3 of 4 at 90% or making 2 of 4 at 10% is not at all unusual.

But in the long run, and 1,000 is NOT the long run, it will all balance out.

Also, pseudo random numbers are indistinguishable to truly random numbers to humans. They are in no way "streaky".

You say that SE's random numbers are "known to be streaky". How do you know? Has SE come forward any admitted, "Oh, yeah - our RNG is totally streaky." Of course not. We "know" because people regularly come to the forums to complain about getting six fails out of ten 80% Hasties.

I don't know what the raw numbers are of crafters who craft regularly, but I know that there's enough out there that SOME of them are going to have unlucky streaks, and of those people, some are inevitably going to call foul and come to the forums to complain. And since folks don't come to the forums to chat about the completely normal crafting experience they have, it provides the illusion that the game is chock full of below-average crafting success. It's simply not the case. SE's RNG is, as far as any human will ever be able to tell, completely random.