This game has seriously messed up RNG

[Joe777]

Say what you want and dismiss it as random all you like but you cannot deny how screwy it is.

Case and point: Why is it 80-89% chances are so prone to failure? Just today I failed Hasty Touch with Steady Hands II active every attempt (80% chance) in 1 synthesis 9 times! There have been multiple times where that same condition failed 3 times in a row, including during that synthesis. Constantly I am getting NQ results in that 80-89% HQ chance range. I don't care how two-faced RNG can be with both good and bad RNG, there should NOT be THIS much failure at such a favorable rate, not once, ever, in one sitting.

I think Square Enix needs to look over their RNG system because something is seriously not right... even if only by a little bit.


I should also like to take this moment to point out I haven't seen a 10k line in mini cactpot in at least a month now, and I used to see them often. Don't know whether it was nerfed into its now intended rate or what but just wanted to point it out.

[CookieMonsta]

Say what you want and dismiss it as random all you like but you cannot deny how screwy it is.



I think Square Enix needs to look over their RNG system because something is seriously not right... even if only by a little bit.


I should also like to take this moment to point out I haven't seen a 10k line in mini cactpot in at least a month now, and I used to see them often. Don't know whether it was nerfed into its now intended rate or what but just wanted to point it out.

IIRC this game uses an extremely high quality Mersenne Twister RnG system. This is about as close to true random as you can get from a software perspective. True random does not care about your subsequent RnG results, it only tends to average out over extremely large sample sizes.
I would argue its because most other games use shielded or lower quality RnG engines thats why you don't suffer as much confirmation bias as this game.

[JCharms]

I dislike the crafting in this game because if it.
In most games you usually get better at your craft as you go, here? they just add more layers of RNG.

[ADVSS]

IIRC this game uses an extremely high quality Mersenne Twister RnG system. This is about as close to true random as you can get from a software perspective. True random does not care about your subsequent RnG results, it only tends to average out over extremely large sample sizes.
I would argue its because most other games use shielded or lower quality RnG engines thats why you don't suffer as much confirmation bias as this game.

im sorry missing 5 hits with a 99 percent success rate is neither true nor random. most people can agree they get shocked more with impossibly low success rates succeeding than what should at least guarantee a 1 in 4 no matter what. it almost feels like the actual success rate is whats missing between the success rate and 100, untill you have 100 then it works just fine


Edit: IMO crafting and gathering should have yes the initial success rating but scale based alot like how desunthesis works. Lets be real you crated a million of these damn leather wristbands, how often are you really going to fail at it at this point? its like rolling a d 20 for something a person can do in their sleep. "im sorry sir, you failed to turn the knob on the door, you die instantly"

[MN_14]

Edit: IMO crafting and gathering should have yes the initial success rating but scale based alot like how desunthesis works. Lets be real you crated a million of these damn leather wristbands, how often are you really going to fail at it at this point? its like rolling a d 20 for something a person can do in their sleep. "im sorry sir, you failed to turn the knob on the door, you die instantly"

RNG can seem quite streaky at times from my experience and I honestly don't know whether it's due to their algorithms or confirmation bias. One thing I do know though is that the % you see is the correct one over a large enough sample size. It averages out.

If there's a scaling system like desynthesis, crafts will definitely be far far too easy. You'd be able to increase your success rates just by running a macro over and over. Currently, you can already craft 4* from all NQ mats and HQ close to 95% of the time as long as you take the time to learn the system and avoid mechanically running a rotation or macro repeatedly (rotations will teach you how the base skills work only; you won't really learn RNG management/adaptation with either).

There will be crafts where you miss 7-8 times but they're relatively rare. 5-6 misses are more common (maybe 1 in 10 crafts), but you normally have a decent shot at recovering from those. From all NQ, it's not uncommon to have 5 misses and still end up with 80-90% (4*). With higher starting quality, you can recover to 100%. Start with high quality, and even 7-8 misses is recoverable.

A crafters own understanding of the abilities plus management of RNG has an effect on their overall success rates, allowing for more player differentiation. I treat NQs as part of a system where your goal is to be able to maximize your HQ yield.

[SeriousxSarcasm]

You know how bad it is? My sis dropped any form of steady hand in her crafting rotations, because she still gets the same result when she doesn't use them, and can save CP for other abilities.

I dunno, even if it is random, it's almost like a slap in the face when you put in that much effort, get your stuff to 95% chance HQ, and RNG gives you the middle finger (multiple times, in fact!). It's one of the main reasons why I gave up on crafting seriously lol. But there's not much to be done about it really.

[Iromi]

Yup, I've been playing since day 1 in 1.0 and noticed this.

90%+ always is failure or miss....as far as crafting and gathering goes. When it comes to material melding, 5% and I get it first chance. It's stupid really...

[MN_14]

I dunno, even if it is random, it's almost like a slap in the face when you put in that much effort, get your stuff to 95% chance HQ, and RNG gives you the middle finger (multiple times, in fact!). It's one of the main reasons why I gave up on crafting seriously lol. But there's not much to be done about it really.

Yep, it's definitely frustrating sometimes. There have been a lot of crafts where missed 5 times, but managed to recover back to 90% only for the craft to NQ. However, on even a 4* craft, you can craft from 0-100% the majority of the time anyways. I dislike RNG when it seems to be overly streaky, but the thing is, RNG is also the only part of the current system that gives an incentive for a crafter to learn and improve their crafting techniques. Your own strategies really can make a difference.

[Nixxe]

The randomness of Hasty Touch can be a bit annoying, but I suppose that's the penalty for not costing CP, aside from Steady Hand I/II anyway.

Either way, there's nothing at odds with proper randomness about getting a long streak of failures with any sub-100% chance. In fact, if you couldn't get such a streak, it would be less genuinely random. Assuming the nominal odds are the actual odds (80%), then getting nine failures in a row has a 5.12 * 10^(-4)% chance of happening, but no matter how inclined you might be to round it to zero, it's still non-zero and as such a streak of that nature is part of the overall 50% chance of Hasty Touch.

[Yukiko]

The main mistake in calculating your % is that the server use ONE RNG for ALL players. If it says 99% it means not only for you, it means the "next" roll. That means you actually will never have the 99% because there will always be some others who finish the same craft/dungeon/loot before you! That results in seeing from your perspective many more odds than it should have! SE does not have any net to catch odd bias. The next mistake is that SE uses an algorithm for high numbers (you remember, one RNG for all players) but you are just a single player from that high number. As example you are number 254 from a total number of 123.456.789 rolls made at same time! Statistically when counting all rolls, RNG is working fine, but for YOU it will result in seeing more odd bias.

There isnt an easy way to fix that because we need a high number RNG, you cant have "one little server" for each player! Thats why most games use a program code that works as filter (or net) to catch odd bias results.

Edit: In add we have a Server/Client problem, if you give the client more "power" it will result in cheating. You need to balance and ask if players who cheat will mess up the overall game or not? (Example: RNG is not only for crafting/loot, its for damage calculations or any other mechanic too. So, does it hurt more when players cheat best gear or when they cheat to auto kill boss? You need to find balance and ask yourself what is actually more gamebreaking)