Reverse Engineering Augments .... It's a LIE

[Wahala]

Every time I work on re-engineering augments, it takes me 3 or 4 or 5 or 6 ... or .... Augments to re-engineer a blue to a purple. Even though it says (Like everything else) 60% chance.

 

Other items (except relics which are about the same as augments) I get a re-engineered schematic on the first or second try (Usually). I am sure I will just keep doing it but it is a pain and seems a bit unfair.

[Nightblazer]

RNG is still RNG. 60% chance doesn't mean it only takes 1 or 2 tries. It means there is an 60% but still hit that 40% each time. No different then flipping a coin 10 times and landing on heads 9 times. Welcome to RNG Edited March 11, 2018 by Nightblazer

[CrazyCT]

I don't have the energy to explain this

[Andryah]

I don't have the energy to explain this

 

heh... my feelings exactly.

[Jdast]

I don't have the energy to explain this

 

No lie. I started typing a response and got half way through the sentence and said to myself...I just can't finish the sentence.

 

Dasty

[ArdeliaAgain]

Tag...I'm in. I haven't burnt myself out with it recently.

 

Any random system (or simulated random system, like computer RNG) requires streaks to be considered truly random....several consecutive results that seem to defy the stated odds. Even a 99% chance of success must occasionally encounter several failures in a row. If the system doesn't contain those streaks, it's not random...it's deterministic.

 

That's how probability works.

 

Not even going to go into the notions of sample size as it regards statistical relevance, or confirmation bias.....Google is a robust tool.

 

All you need to understand is: Streaks mean the system's working, not broken.

Edited March 11, 2018 by ArdeliaAgain

[ArdeliaAgain]

And before anybody's tempted to remind anybody that computers can't REALLY do random numbers, or how easy it would be for there to be a mistake somewhere....

 

No. Just no.

 

This isn't a newfangled thing people started wrangling with on the advent of MMOs and game forums. Methods of generating pseudo-random numbers that very closely approximate truly random systems have been around for a long time. They're sufficiently random that a human being can't tell the difference, and they usually have less deterministic error that would result from wear on the edge of a physical die, or casting variations in a flipped coin.

 

This isn't 1975....for any given application, computer RNG can be considered more than "random enough".

[Andryah]

This isn't 1975....for any given application, computer RNG can be considered more than "random enough".

 

Yep.

 

Heh.. this triggered a flashback for me........

 

Back in the day (mid 70s) ... when I was a newbie electrical engineer, one of my first assignments was to design a random number generator. At that time the best commercial source for a random seed was a zener diode, so off I went to design a random number generator. When I submitted my design for review by a senior engineer... I got a polite schooling on the fact that true random was not required, that a zener source was over-kill and that a generated pseudo random result was fine in almost all applications. Basically, any sufficiently large number of possible variations will appear random even if it still has a "pseudo" element in it's construction.

 

Now days... computational algorithms are the name of the game.. since almost anything that requires a random result exists in close association with some form of cpu. It is even possible to create pseudo random numbers that meet cryptographically secure levels of randomness.

Edited March 11, 2018 by Andryah

[Amantionus]

This isn't the 70's now? What will I do with all my punch cards...

[brazilboy]

RNG is still RNG. 60% chance doesn't mean it only takes 1 or 2 tries. It means there is an 60% but still hit that 40% each time. No different then flipping a coin 10 times and landing on heads 9 times. Welcome to RNG

yep, nailed it. 60% chance of success = 40% chance of failure (not succeeding)

 

probability of not succeeding 2x in a row = 40% x 40% = 16%

3x in a row = 6.4%

4x in a row = 2.6%

 

So yes, by the third try you should have a close to 94% (100% - 6.4%) chance to have succeeded *at least* 1 time, but there's still that pesky chance of failing 3-4x in a row (6.4% or 2.6% chance to not succeed consecutively 3x-4x).

 

See here:

This kind of thing is called a Bernoulli Process, which is complicated and hard to ELI5. But in this specific case, if you're asking “what's the probability that I win at least once in 100 attempts” the easiest way to answer it is to reverse the question: “what's the probability that you don't win at all in 100 attempts?”

 

It's easier because now you're asking about an exact number: the probability that you'll see exactly 0 wins in 100 tries. That's just the base probability of losing (99%) multiplied by itself 100 times, or 0.99¹⁰⁰, which equals about 0.366.

 

So you have a 36.6% chance of losing, therefore you have a 63.4% of winning at least once.

Edited March 11, 2018 by brazilboy

[Jozandra]

Any random system (or simulated random system, like computer RNG) requires streaks to be considered truly random....several consecutive results that seem to defy the stated odds. Even a 99% chance of success must occasionally encounter several failures in a row. If the system doesn't contain those streaks, it's not random...it's deterministic.

 

That's how probability works.

 

All you need to understand is: Streaks mean the system's working, not broken.

 

Best example of seeing this work happens during the Life Day Event, when throwing snowballs at the overheated droids. The number of times you will actually "win" a prize is higher when there are a dozen players hitting the same droid compared to if you are the only one throwing snowballs.

 

At least that was my personal experience last time, LOL.

[psandak]

I'm having flashbacks.

 

The crew skills forum is littered with these types of OP posts across the years...ALL of them disproven. The system works as intended.

 

Wahala, what you are experiencing is the gambler's fallacy (https://en.wikipedia.org/wiki/Gambler's_fallacy); you believe that because you've failed a few times "you are due" for a success...it just does not work that way.

 

I remember when 5.0 launched, I could not get the purple attack adrenal to save my life...it took 7 tries. No other schematic took more than 2 tries. Was the RNG on the blue attack adrenal broken? No!

Edited March 11, 2018 by psandak

[Wahala]

I'm having flashbacks.

 

The crew skills forum is littered with these types of OP posts across the years...ALL of them disproven. The system works as intended.

 

Wahala, what you are experiencing is the gambler's fallacy (https://en.wikipedia.org/wiki/Gambler's_fallacy); you believe that because you've failed a few times "you are due" for a success...it just does not work that way.

 

I remember when 5.0 launched, I could not get the purple attack adrenal to save my life...it took 7 tries. No other schematic took more than 2 tries. Was the RNG on the blue attack adrenal broken? No!

 

I think I understand the random thing ... and that it is possible to go through many iterations of the process without success. However, I have been reverse engineering stuff in this game for 6 years. What I am saying is that the coincidence of reverse engineering augments and relics both seem to take more trys consistently. Perhaps my "Luck'" or my memory or something personally I have against these components is what is making it seem that way. It is just my memory telling me that these two items are just harder to get to r-n-g true for me. Maybe I am mis remembering or getting senile. I remember once it took me 9 tries to get a piece of armoring ... but that was back when they said it was 20% chance (10% for augments). I don't know anything for certain ... just seems to work this way to me. Apparently I am the only one to notice it so ..... well so nothing. I do not think I am being cheated or anything like that. Just wondered if anyone else noticed a significant difference when rev engineering augments.

Edited March 11, 2018 by Wahala

[YlliyaXor]

RNG, the only part of the game that NEVER has bugs.

[ArdeliaAgain]

I think I understand the random thing ... and that it is possible to go through many iterations of the process without success. However, I have been reverse engineering stuff in this game for 6 years. What I am saying is that the coincidence of reverse engineering augments and relics both seem to take more trys consistently. Perhaps my "Luck'" or my memory or something personally I have against these components is what is making it seem that way. It is just my memory telling me that these two items are just harder to get to r-n-g true for me. Maybe I am mis remembering or getting senile. I remember once it took me 9 tries to get a piece of armoring ... but that was back when they said it was 20% chance (10% for augments). I don't know anything for certain ... just seems to work this way to me. Apparently I am the only one to notice it so ..... well so nothing. I do not think I am being cheated or anything like that. Just wondered if anyone else noticed a significant difference when rev engineering augments.

 

I've found REing to be especially bad, in terms of bias, both because you're more focused on the result, and because bad luck hurts more.

 

Most crafting, you set it and forget it. You tend not to notice if your luck was especially good or bad when you mass-craft a stack of components, for example....when it's done, you end up with one stack, and if you had a run of good luck on crits, that stack is just a little bigger than normal; it doesn't stick out.

 

REing, you tend to craft each item one at a time (why craft more blues than you need to?). Once the item is made, you open your inventory and look for that specific item, click a separate button, then click on the item again. You look and listen for the successful learn result. Those things all focus your mind on the result more than normal.

 

If it's a success, you move on. If it's a failure you have to recraft that item, and go through all the same steps again. You have to keep repeating that more focused set of activities more times, making failures stick out more....and that's all on top of the fact that our brains are wired to pay attention to negative results more than positive ones as it is.

Edited March 11, 2018 by ArdeliaAgain

[ivanhedgehog]

RNG, the only part of the game that NEVER has bugs.

 

Not only that, but people always assume that it is actually random, that the devs told you the truth. Did they? do you have actual proof? have you read the code? they have lied before. they have been "mistaken" before. they have changed things and "forgotten" to tell us. people assume that the guy programming the rng implimentation didnt misplace a decimal. it has happened to much more important code in the past, it doesnt make him a bad person. oops happen. just like the oops that ended up with that mars mission crashing.

[ArdeliaAgain]

Not only that, but people always assume that it is actually random, that the devs told you the truth. Did they? do you have actual proof? have you read the code? they have lied before. they have been "mistaken" before. they have changed things and "forgotten" to tell us. people assume that the guy programming the rng implimentation didnt misplace a decimal.

 

I would be stunned...and I mean absolutely beyond floored....If anybody at BW wrote a spec of "RNG code". Why would they? That would be like taking your car in for an oil change, and having the tech go drill and refine the oil, and reengineer a new filter from scratch.

 

Again, this isn't 1975; random number generation doesn't require custom code. Most languages have RNG utility functionality that works perfectly fine to generate "random enough" results. So do operating systems, DBMS.....It would be beyond ridiculous to "code" RNG.

 

Maybe you meant there was a decimal misplaced somewhere in the comparison parameter? Think that through....does anybody have any results indicating a 600% success chance? Or 6%?

 

Ever notice that the people who're most willing to assume a developer screwed up somewhere are usually the ones who least understand how things work?

Edited March 12, 2018 by ArdeliaAgain

[TrixxieTriss]

Every time I work on re-engineering augments, it takes me 3 or 4 or 5 or 6 ... or .... Augments to re-engineer a blue to a purple. Even though it says (Like everything else) 60% chance.

 

Other items (except relics which are about the same as augments) I get a re-engineered schematic on the first or second try (Usually). I am sure I will just keep doing it but it is a pain and seems a bit unfair.

 

The RNG gods hate me too. But it’s still a 60% chance each one you do. You might just be hitting the 40% of less chance more often than the 60% chance of it happening. Can’t help luck, that’s what RNG is at its core. It’s a roll of the dice, which if money was involved, would be called gambling.

Edited March 12, 2018 by TrixxieTriss

[PetFish]

I do not think I am being cheated or anything like that. Just wondered if anyone else noticed a significant difference when rev engineering augments.

 

Then why the clickbait rage thread? This statement is much cleaner and understandable than your aggressive thread title and OP.

[ivanhedgehog]

I would be stunned...and I mean absolutely beyond floored....If anybody at BW wrote a spec of "RNG code". Why would they? That would be like taking your car in for an oil change, and having the tech go drill and refine the oil, and reengineer a new filter from scratch.

 

Again, this isn't 1975; random number generation doesn't require custom code. Most languages have RNG utility functionality that works perfectly fine to generate "random enough" results. So do operating systems, DBMS.....It would be beyond ridiculous to "code" RNG.

 

Maybe you meant there was a decimal misplaced somewhere in the comparison parameter? Think that through....does anybody have any results indicating a 600% success chance? Or 6%?

 

Ever notice that the people who're most willing to assume a developer screwed up somewhere are usually the ones who least understand how things work?

 

I have seen it happen in other games. They dont have to mess up the actual rng code, just how it is applied in game. I have actually seen this happen in an MMO. people make mistakes, in real life. it happens. Dont assume that this is a special case just because you want it to. Even doctors who hold peoples lives in their hands make mistakes, why would you assume that programmers dont? If you have some sort of proof, please show us. the actual code will do.

[ArdeliaAgain]

I have seen it happen in other games. They dont have to mess up the actual rng code, just how it is applied in game. I have actually seen this happen in an MMO. people make mistakes, in real life. it happens. Dont assume that this is a special case just because you want it to. Even doctors who hold peoples lives in their hands make mistakes, why would you assume that programmers dont?

 

I've seen success chance issues in other games, too...that is, there were issues where a parameter was input wrong like I mentioned (one particular category was supposed to be 100%, while some others had a failure chance...the 100% didn't get set).

 

I've also seen issues where a player was supposed to have an improving chance over time (like the baseline is 5%, but can be increased via some rating), and the improvement wasn't reflecting.

 

What we're talking about here is a bit different...everybody has the same 60% RE chance. That chance has been at that level for quite awhile....and it's incredibly likely (not guaranteed, but likely) that's a parameter, which is unlikely to be messed up unless a change is made to it.

 

All that said, as you tried to corner me with, I don't have access to any part of the engine code...sure, it's possible the RNG is implemented in such a convoluted, unusual fashion that it applies inconsistent results to different people intermittently. The most likely behavior for anything even close to that would be everybody seeing 100% success, 100% failure, or rolls just crapping out and not happening at all, and we're not seeing that, but sure, it's possible.

 

But what's more likely?

 

Concepts like the gambler's fallacy and human beings' general struggle to deal with randomness are well documented, and computer RNG is fairly basic to implement. On the other hand, people anecdotally complain that it doesn't "feel right"....and only ever when they have poor luck, not good luck.

 

So which do you feel is the more likely cause?

 

I've never in my life claimed developers don't made mistakes. On the contrary, I'm all too painfully aware that bug-free software doesn't exist outside of 100-level classroom assignments of a hundred lines or less.

 

But I do know what's more likely here.

 

Basically, one of us is suggesting that the OP's observation is just that...an anecdote, driven by some well-understood bias phenomena. That person has provided some context and explanation for why they feel that's the case.

 

The other made the immediate assumption that the system's somehow flawed (with the extra implication that we might even have been misled). That person's supporting their assumption with demands for negative proof they know nobody can provide.

 

So, in lieu of your credible evidence (or really, any evidence) otherwise, I'm going to stick by what I said....it's unlikely to be a bad parameter specific to augments, and vanishingly unlikely to be a "broken RNG".

Edited March 12, 2018 by ArdeliaAgain

[ivanhedgehog]

I've seen success chance issues in other games, too...that is, there were issues where a parameter was input wrong like I mentioned (one particular category was supposed to be 100%, while some others had a failure chance...the 100% didn't get set).

 

snip

 

So, in lieu of your credible evidence (or really, any evidence) otherwise, I'm going to stick by what I said....it's unlikely to be a bad parameter specific to augments, and vanishingly unlikely to be a "broken RNG".

 

I agree it is very unlikely. But the people that blindly say "rng is rng" completely ignore any possibility of accidental mistakes or purposful dev changes. Just like I dont believe the devs "forget" to put things in patch notes. Change management in a software team isnt all tthat hard to implement, and failure to do so has some bad consequences.

[Xbladez]

This thread is now 3 pages long.........Just wow!

[Andryah]

Basically, one of us is suggesting that the OP's observation is just that...an anecdote, driven by some well-understood bias phenomena. That person has provided some context and explanation for why they feel that's the case.

 

The other made the immediate assumption that the system's somehow flawed (with the extra implication that we might even have been misled). That person's supporting their assumption with demands for negative proof they know nobody can provide.

 

So, in lieu of your credible evidence (or really, any evidence) otherwise, I'm going to stick by what I said....it's unlikely to be a bad parameter specific to augments, and vanishingly unlikely to be a "broken RNG".

 

I agree with your assessment here.

 

And to add another factor in all this willingness to head to conspiracy theories and supporting the unsupportable..... human nature has been pretty well laid bare via this giant social media experiment currently taking place in the internet....... https://techcrunch.com/2018/03/08/false-news-spreads-faster-than-truth-online-thanks-to-human-nature/ (a peer reviewed paper that analyzes why false or negative narratives are so readily propagated by humans)

“We have a very strong conclusion that the spread of falsity is outpacing the truth because human beings are more likely to retweet false than true news,” explained Sinan Aral, co-author of the paper.

“There’s actually extensive study in human communications in why certain news spreads faster, not just a common sense understanding of it,” explained Deb Roy, the third co-author of the paper. “It’s well understood that there’s a bias to our sharing negative over positive news, and also a bias to sharing surprising over unsurprising news.”

 

Basically, human nature, broadly speaking, tends to crave drama and nothing creates drama faster then a negative assumption or narrative propagated to others, especially if it is absent clear hard incontrovertible multi-sourced proof of it being incorrect. Any, no matter how thin or contrived, gray zone or area around a topic can and will be fed with negative energy in the process. Facts simply do not matter as they just slow down the race to full on negative drama.

Edited March 12, 2018 by Andryah

[Keta]

RNG, the only part of the game that NEVER has bugs.

 

I snorted my Red Bull reading that.