Fix your random number generator

I'm not a numbers person but I've seen this behavior also - especially with the Dash skill - I'll fail 4+ times in a row sometimes more..

@Chris even remarked that he thought something was bugged with that skill in particular .. perhaps it's like OP is saying with the PRNG?

 

The problem is i can show you these streaks every 5-6 sessions of play =P
So these impossible odds* arnt just happening once in a while.. they are happening regularily.

Its been pretty common knowledge that the RNG doesnt seem right to most people. Maybe people should start recording every *impossible* streak? Its literally the reason i dont craft anymore. going 20+ non exceptionals when i had 25+% chance, then getting 1, failing 15x more.

But again opposite end of spectrum, ive had 10/11 exceptionals.. and its like.. what? That shouldnt EVER happen statistically. Not to mention have happend numerous times of each example, and numerous times to people i play with.

 

Been happening to me for thousands and thousands of things I craft or try to enchant. Welcome to the club. I quit crafting.

Ive also been ranting about it here for 1.5 years.

 

I don't see reaction of @Randbo is exaggeration, rather frustration. Anyone can provide opinion in this forum and method of delivering it varies from explanation to pure ranting. Exact numbers of crafting success/failure algorithm only can be provided by development team, not by a player.

 

The sarcasm was lost on you =P.

 

If I thought that anything would be changed I'd do a video. Something is broke. 3 separate consecutively mined nodes with 94%; 1st one 3x fail, 2nd one 2x fail, 3rd one 3x fail. Just tonight while trying to skin a dragon with 78% chance; 5x fail in a row, perhaps this one is more likely but the repeated multiple failures on 94% just don't make sense.

I have drastically reduced the amount of crafting that I do because of the gross loss of materials due to lack of exceptional and waste trying to enchant/mw.

 

I have to admit that I stopped even bothering with gathering because I would fail so often... I couldn't imagine the amount of time people put into crafting...

theres no smoke without fire...

 

Hehe. I detest sarcasm, that's why

 

ya every crafter i've talked to knows about the streaks. Nothing worse than spending 50-100k on initial crafting to have nothing worth applying gold and silver to. 2 of the last 3 things i crafted took 11 and 13 normals before an exceptional, leading to having so few exceptional to enchant/masterwork that none of the 3 pieces i wanted to craft survived. of 60 armor pieces i produced. That, coupled with the incentive of waiting that 10k free craft XP provides, i got so bummed i stopped crafting too.

 

Time and resources if you are crafting, which I very rarely do, if at all. I'd be upset if I were a crafter, which I am not.. Just a miner. It's not lost on me though, it's a small unnecessary time sink to have to mine the same node 3 times in a row.

 

PRNG has his weaknesses, but i noticed it for a long time. The Chance of failing 3+ times in a raw feels higher then failing only once. It follows a algorithm, a pattern and is not that save and random as you think.
It is not only Crafting, you can see this also in the crit chance of every mob, multiple times in a raw.
And this is not the first or only Game this Problem was addressed. They have to extend the scheme or use TRNGs to get better results.

 

You are right, the odds of 3 failures in a row was more like 5000 ore nodes, not 500. Eyeball math dropped a decimal place.

 

Risk - online, lost a 32vs3 with 6sided dice. Lost a 20vs3 with 7sided dice. These are vivid remembrances for me.

Granted settlers (board-game with dice) we've had a 2 hit 5 times in a row and another game where the 8 rolled just once in the first round and never hit again!

 

And I think it's not so much as to that you have failed as to how you have failed. This game's failure is no fun. It allows no incentive to continue. Failing could at least have a chance of crafting something like a lesser product or getting some of the special resources back not just pieces of scrap. The system is faulty on how it's designed. It just angers the Crafter and makes them not want to craft. We need to fix this.

 

Does astronomy/celestial bodies currently affect the success and failure rates of crafting, gathering and other skills?

 

He was referring to the way computers generate random numbers. There are two methods as described below.

Because computers have to be inherently reliable, there are few sources (if at all) of randomness within it. Which is why true randomness is usually dependent on external sources.

https://www.howtogeek.com/183051/htg-explains-how-computers-generate-random-numbers/

True Random Numbers


You may be wondering how a computer can actually generate a random number. Where does this “randomness” come from. If it’s just a piece of computer code, isn’t it possible the numbers the computer generates could be predictable?


We generally group the random numbers computers generate into two types, depending on how they’re generated: “True” random numbers and pseudo-random numbers.


To generate a “true” random number, the computer measures some type of physical phenomenon that takes place outside of the computer. For example, the computer could measure the radioactive decay of an atom. According to quantum theory, there’s no way to know for sure when radioactive decay will occur, so this is essentially “pure randomness” from the universe. An attacker wouldn’t be able to predict when radioactive decay would occur, so they wouldn’t know the random value.

For a more day-to-day example, the computer could rely on atmospheric noise or simply use the exact time you press keys on your keyboard as a source of unpredictable data, or entropy.

Pseudorandom Numbers
Pseudorandom numbers are an alternative to “true” random numbers. A computer could use a seed value and an algorithm to generate numbers that appear to be random, but that are in fact predictable. The computer doesn’t gather any random data from the environment.

This isn’t necessarily a bad thing in every situation. For example, if you’re playing a video game, it doesn’t really matter whether the events that occur in that game are cased by “true” random numbers or pseudorandom numbers. On the other hand, if you’re using encryption, you don’t want to use pseudorandom numbers that an attacker could guess.

For example, let’s say an attacker knows the algorithm and seed value a pseudorandom number generator uses. And let’s say an encryption algorithm gets a pseudorandom number from this algorithm and uses it to generate an encryption key without adding any additional randomness. If an attacker knows enough, they could work backwards and determine the pseudorandom number the encryption algorithm must have chosen in that case, breaking the encryption.

 

@Randbo
Even if your concern is valid the way you communicate it is at least in my humble opinion not justified. Just keep calm and write the problem in a constructive way.

@Topic
If failing the first time is 6% than failing a second time will be 0.36% and a third time 0.0216%. If I’m not wrong this would mean that mining a node you have such a triple fails every (round about) 4630 times.
Sound much…not really as this is the average value. If there are 100 people mining every 46.3 nodes one of these 100 miners recognizes such a triple fail.
Of course this doesn’t mean that the client calculation has to be right but at least sometimes it is just a personal imprint that there is something wrong.

@majoria70
Maybe a lesser product would be an option but generally I think failures are needed. If crafting an e.g. +10 item would be an easy task everyone would have these items and everything else would be useless.
But maybe there would be a different option possible. Currently you can choose the type of bonus you would like to add to the item…why not something similar in the opposite direction on a failure? Instead of choosing +3 strength/+10 focus/+1% crit you (or the system) have to choose between -3 strenght/-10 focus/-1% crit on a failure. The complete loss of the item could still maintain as result of a critical failure.

 

Does astronomy/celestial bodies currently affect the success and failure rates of crafting, gathering and other skills in R44?

 

Do we have enough data/evidence on this? I noticed starman mentioned this failing over 1000 times for him which is the sort of set sizes you need to compare this stuff with.

Maybe the API has data points on how many users have been successful vs failed in their attempts?

 

I didn't say failures weren't needed. One thing that makes people mad is they use for example 200 special ingots at gm on an item and they fail and get 16 scraps back or some miniscule #. True I'm not arguing about figuring out the best # combinations. I just don't think that is the whole problem with our crafting system. Like if you got the whole number combinations figured out our crafting would be more fun, hmm some say not.

Edited

 

Im challenging the pseudo random generator they are using. Most likely reusing the same seed over and over, to get the same pseudo random number sets.
I mentioned NSA above because they randomize numbers using satellite antennas that gather atmospheric noise. Thats true random.
You can help the randomizing process by RANDOMIZING THE SEED. Thats double random and it may help alot.
I want to get away from those stupid long streaks of LOW numbers that produce long streaks of FAIL.
Some of you just dont see that triple fails of 94% in mining, HAPPEN ALL THE TIME, ALL DAY, EVERY DAY. I know, i been there and see it all the time. Its not a big deal to me because i just keep on mining, but its ILLUSTRATIVE of the exact problem i am griping about. It shouldnt happen. I shouldnt see it once an hour.
And those strings of 20x fails on exceptional at 29% is just not acceptable. Huge highend resource LOSS. Its just not that i hit an unlucky streak because IT HAPPENS ALL THE TIME.
Spell casting and buff casting too. 74% and i get 12 fails in a row?
NO! 74% fails once, no problem, twice, it happens, 3x, rarely but it does happen, 12x? NEVER, NEVER, NEVER!!!! You cannot allow that to happen. You're LOTTO is RIGGED.
Sign onto a true random generator, PROVE TO US you did, and we will be quiet. I dont mind failing against that, KNOWING that i will succeed exactly 94 times out of a hundred tries at 94%
At a large statistical base, 100 is good enough, I should hit 94, 93, 95, over and over. NOT 80, 75, 74 out of 100.

MAKE RANDOM CHANCE TRULY RANDOM. or reseed the generator on every use.

 

Randomness isn't easy or cheap. You give a good example with the NSA, however I can't see satellite being a useful use of backers money and most computing calculations typically have a psuedo random generator because if they waited for their true source of randomness it would take forever for crypto and whatnot to work (It is why most Linux applications for randomness use /dev/urandom instead of /dev/random). I have heard of people using things like radio to generate noise/randomness as well, however don't know how that would work in a modern datacentre. Typically if they are utilising the random generators provided by the Operating system then they "should" be ok.

What I "think" the problem is related to is the source of the bias (Introduced when calculating your success chance). It would explain why I have issues when I mine/harvest in zones where I have 100% chance to succeed and on occasion (Once every couple of days) it fails. At the very least there is a potential rounding error which could be causing some issues.

 

And of course someone is not going to experience the norm, some people will experience the extremes and these are the ones who will react with great emotion and take to the forums. Then those others who experience the extremes will feed on this and soon you have created the perception of increased occurrence of a negative action.
This is more of a communication issue that it is a coding or mathematics one.
The process of how multiple fails occur in a row should be explained, even with actual, annonomised data from the game. It may be helpful to have some of thinking behind the probabilities more clearly explained. Although due to Ports' budget this may be tricky and is left to the community.

I feel this pain as any other does and if you are on the less probable, negative side the pain hurts even more.
I don't like the probability factor, even though I have been arguing Ports' case I don't think it's the way to go. Becuase you are going to have people who suffer outlier probabilities and that is extremely emotionally painful.

I don't think that players fear the role of the dice, it's just that pain of loss from what could be hours or days of work, when you hear your neighbour getting the positive extreme is psychologically painful enough for it to affect interaction with the game.

This is different to when you get killed by a mob or in PvP as it is a quantifiable danger and the level of calculation that Port is asking players to understand is too abstract to be held in the head by those who do not have uber math brains.

Thus it is the nature of the emotional response to this issue that is the point and not the underlying logic of the mathematical algo, which from a coding POV appears fine, but from a human POV may seem broken.

cc @Chris