Guild Wars 2

Hi, I’m half-new player here. I’m really interested to get a precursor in mystic forge.
People said mystic forge is a RNG.

However, I’m very curious whether there are some formula or mechanic hidden to trigger some adjustment so that

1. if one player is losing too much, his chance of getting a precursor next time is increased, or
2. if one player is winning too much, his chance of getting a precursor next time is reduced.
3. or something other else..

Are the mystic forge totally REALLY random?
(in the sense of there are no adjustment written in the programming code, your chance of getting a precursor are all Completely, Absolutely Equal for every single trial under the circumstances that Any Other Input Are All the Same.)

Given the reports of getting nothing from thousands of forge as well as people getting 10 in a row I would say the answer to 1 and 2 are “no”.

If there is some “secret” method to give you a higher chance of getting the precursor, it’s a tightly kept secret. I doubt there is but as I have not tried every single combination in the game, I can only assume that it’s completely random.

It’s really random.

Your questions prove you do not have a very good grasp of the concept of RANDOM (and I dare say this is the issue with LOTS of the whiny, QQ RNG posts on these forums).

It is human nature to try and associate future events with events that occurred in the past, but singular RANDOM events do NOT have any influence from previous events nor can they effect any future events. It should be noted that TRUE random events are not obvious in our everyday life, thus the natural assumption that your last result MUST effect your next one (it does not). Dice rolls and coin flips are examples of truly random singular events.

We are told that the Mystic Flusher is pure RNG but whether you choose to believe ANet or buy into some wild conspiracy theory about hidden luck based stats on characters or accounts is completely up to you.

There is nothing that’ll increase or decrease your chances.

That said, there is plenty of evidence showing that the RNG in this game is not working as intended and not much showing that it is. The same players will consistently get better drops, better forges, worse drops or worse forges. Over time these outliers should move towards the center, but this doesn’t seem to be the case as the trends continue for those effected.

Some call it “wild conspiracy”, I just call it acknowledgement of evidence and coming to a logical conclusion. If the same person wins the lotto time and time again it’s logical to acknowledge something is not working as intended.

it is possible to create a rng that will lean more towards the average than totally random.
how this might be done:
$upper_limit=50;
if(rand(1,50+$upper_limit)==1) {
$upper=50;
give precursor;
}
else {
if($upper_limit>=0) $upper_limit—;
give exotic;
}
so it starts out at 1% chance, and every failed attempt increases your chance of precursor until it bottoms out at 50% chance.
it would take a HUGE sample size to tell the difference between weighted rng and true rng. of course, there’s Occam to tell us that they likely did not go through this much trouble just to mess with your sanity.

yes, it is random

Within big random sequences you are likely to see repetition, patterns, big leads, etc. In fact, it would be very unlikely for a big random sequence to look satisfactorily random in every subset. Essentially, it means you can’t determine anything from somebody getting wonderful results from the forge while someone else gets nothing.

The forge is almost certain to be random as Anet gain nothing from designing it any other way.

It’s so easy to slip into superstition when gambling on the mystic forge.

My mostly sane analytical mind tells me that the chances are random, but it doesn’t stop me from holding off on putting things in after a long play session, saving the sacrifices to zommoros till the start of my next play session.

All it does is skew the number of rares I throw in to the beginning of a play session and – hey presto! – “I seem to get more exotics at the start of my play session!”

I guess it’s healthy that I can recognize my own fallacies at least…

Drops elsewhere in the game fit tend to mirror each other well, for example, it’s not too uncommon for two players in a party to wind up with similar rares out of a chest, but those chests already have sometimes narrowly defined loot tables, like ‘underwater weapons only’, so this can be simple chance.

And then there’s the speculation into how often the game generates specific rare items like precursors…

I’ve heard that the MF generates precursors based on market prices, and that it generates them based on a quota+time table.

I’m almost positive they’ve done the latter in GW1, but gathering the data to prove it is difficult. How do you prove ‘all widget x only drop in the first 5 minutes after the hour’? Disproving it is easy, aye, but proving it?

//Yamagawa
— betting that at most 100 mini polar bears were released a year in GW1, using a hard cap reached only days into a weeks long event. Proving it? Tough, but comfirmed reports of drops fell to nil after the first week in the multi-week event that dropped it?

I’ve made my legendaries, bought my T3 armor, ect. I have a lot of gold and not much to spend it on, so I forge a lot. I’ve forged probably tens of thousands of rares. I’m gonna share with you all of my extensive knowledge of the mystic forge….

You put stuff in, and other stuff comes out. That’s all I’ve got so far.

I’ve got 3 divinity runes in 2 days with just 16 random rare runes.. It’s a miracle.

Essence Snow.3194:

Some call it “wild conspiracy”, I just call it acknowledgement of evidence and coming to a logical conclusion. If the same person wins the lotto time and time again it’s logical to acknowledge something is not working as intended.

In cases I have seen, one guy is buying 1,000,000 tickets and the other guy is going into the convenient store 1,000,000 times and buying a ticket every 1000 trips….. It’s easy to say you “farmed” and area for an hour but you killed 100x less enemies than the winner.

If the data is skewed, it’s easy to make invalid assumptions (or promote flawed theories).

Because there are no facts, there is no truth
Just a data to be manipulated
I can get any result you like
What’s it worth to ya?

Brother Grimm.5176:

Essence Snow.3194:

Some call it “wild conspiracy”, I just call it acknowledgement of evidence and coming to a logical conclusion. If the same person wins the lotto time and time again it’s logical to acknowledge something is not working as intended.

In cases I have seen, one guy is buying 1,000,000 tickets and the other guy is going into the convenient store 1,000,000 times and buying a ticket every 1000 trips….. It’s easy to say you “farmed” and area for an hour but you killed 100x less enemies than the winner.

If the data is skewed, it’s easy to make invalid assumptions (or promote flawed theories).

Because there are no facts, there is no truth
Just a data to be manipulated
I can get any result you like
What’s it worth to ya?

Okay, now reverse that and you’ll get what we have here.

One guy farms and kills 100x times more, but the guy that kills 100x less gets 10x the rewards.

There is no proof either way ,so observation is the best we have to go by.

The problem is that truly random occurrences don’t look random to people. People often expect random events to be spread out evenly but in actuality the events tend to cluster. Below is a picture of random dots. Imagine if the dots showed people getting a precursor and the white areas are where people are trying and failing to get one. Some people would appear lucky and others unlucky when it’s simply randomness making its appearance.

Astral Projections.7320:

The problem is that truly random occurrences don’t look random to people. People often expect random events to be spread out evenly but in actuality the events tend to cluster. Below is a picture of random dots. Imagine if the dots showed people getting a precursor and the white areas are where people are trying and failing to get one. Some people would appear lucky and others unlucky when it’s simply randomness making its appearance.

If the same scatter diagram didn’t repeat itself over the course of time I’d agree. When the same scatter diagram repeats day in and day out, then I will need to question if it is completely random on all accounts.

Essence….

Where is your data for making your scattergram? The one that repeats day after day?

I do not long tolerate trolls, and you insinuate much with your last statement. So much so that either you are a troll, do not know the meaning of the words you are using, or have access to evidence that nobody can possess.

Please present proof, or refrain from trolling.

//Portable Corpse

Essence Snow.3194:

Astral Projections.7320:

The problem is that truly random occurrences don’t look random to people. People often expect random events to be spread out evenly but in actuality the events tend to cluster. Below is a picture of random dots. Imagine if the dots showed people getting a precursor and the white areas are where people are trying and failing to get one. Some people would appear lucky and others unlucky when it’s simply randomness making its appearance.

If the same scatter diagram didn’t repeat itself over the course of time I’d agree. When the same scatter diagram repeats day in and day out, then I will need to question if it is completely random on all accounts.

Given the population, the drop chances and the nature of the rng, it is certain that some “patterns” will emerge.

Remember that old adage about monkeys, typewriters and Shakespeare? That’s exactly what happens here. Doesn’t mean there’s a conspiracy at work – only that you see only a tiny part of a greater whole, connect few dots and decide they form a pattern, and think that pattern will be repeated elsewhere.

Good morrow! I am Arrabian PrInce that can help you bigtime with mystic forge yoU speak off. If you send smAll fee of 50g to my accountant I can make show you that forge is nOt so “mystic” they speak off. My accountant has account name Shadyharry.1234 and he is veRy trustworth. He also has good wines you can buy at 1g each that Help you do mystic forge forcursor twenty times in your closet!

Essence Snow.3194:

Brother Grimm.5176:

Essence Snow.3194:

Some call it “wild conspiracy”, I just call it acknowledgement of evidence and coming to a logical conclusion. If the same person wins the lotto time and time again it’s logical to acknowledge something is not working as intended.

In cases I have seen, one guy is buying 1,000,000 tickets and the other guy is going into the convenient store 1,000,000 times and buying a ticket every 1000 trips….. It’s easy to say you “farmed” and area for an hour but you killed 100x less enemies than the winner.

If the data is skewed, it’s easy to make invalid assumptions (or promote flawed theories).

Because there are no facts, there is no truth
Just a data to be manipulated
I can get any result you like
What’s it worth to ya?

Okay, now reverse that and you’ll get what we have here.

One guy farms and kills 100x times more, but the guy that kills 100x less gets 10x the rewards.

There is no proof either way ,so observation is the best we have to go by.

You can twist it any way you like but you are just making my point. The data is NOT there to make a case either way, so your belief in some “hidden” stat within the game is just superstitious mumbo-jumbo and trying to “convert” people to your conspiracy theory makes you appear to be a few donuts shy of a bakers dozen.

To be honest, I find it an amusing distraction to observe how these “causes” outside the realm of anything that really matters on this Earth develop. Keep up the good fight!

A friend of mine managed to get 2 Precursors within 100 rare staves. I spent my last 9g on shields and luckily managed to get ‘The chosen’ on my very last roll. I had to farm 20g or so in order to sell it overnight for 260g. Since then I’ve spent over 300g on forge and not had any precursors back, so the chance is really random.

Zomorros just likes me more than you.

This is all I will say. It is not random, and I have made over 1000g from the Forge. Some people’s Toilet is another mans… something like that.

People are generally thinking ‘uniform distribution’ when they say ‘random chance’.

The Mystic Forge does NOT run on uniform distribution.

I was told the ‘best’ method of ‘crafting’ precursors in the forge by someone sitting on about 13K gold and produced my first in about 400 attempts. The second took another 500 or so. After another 900+, still NO third precursor so I have pretty much joined ‘Mystic Toilet Anonymous’ while I am still up about 100g. Clearly I am not lucky, but what if I am???

Nike.2631:

People are generally thinking ‘uniform distribution’ when they say ‘random chance’.

The Mystic Forge does NOT run on uniform distribution.

I am pretty sure nobody in this thread thought it does. Of course different outcomes have different chances of happening. You have a much better chance of getting a rare than exotic from four rares, after all. What people mean by saing that it is random is that there are no “hidden recipes” for precursors, and that the chance is dependant only on the recipe used (not affected by DR, gems bought, time played, account id or anything like that).

Since when does believing there may be a bug or mistake in the coding = conspiracy theory, or is that just a general reply?

Essence Snow.3194:

Since when does believing there may be a bug or mistake in the coding = conspiracy theory, or is that just a general reply?

Just a general reply. Of course, lot of “reasons” supplied for “RNG favoring some people more than others” are straight from conspiracy theories (including the one about Anet just flagging people to get bad drops for being too vocal against them on forums).
But to address you specifically, there’s absolutely no reason to think there’s some bug at work as well.

No reason aside from the field research?

From talking to a guildy that has crafted dozens of precursors in the mystic forge. On avg it took him 200 attempt with random rares to get a precursor. When I crafted The Legend it took about the same amount of attempts to get it. Overall it seems like mystic forge has a 0.5% chance of creating a precursor. Save up your money and try crafting one in a single attempt for the best chances.

Mystic forging is random but based off of what my friend said 200 attempts the per precursor make crafting one cost around 200g-320g(4rares 25-40s per rare, 1g-1.6g per attempt). So I would just save up all of you money and test your luck if the precursor you want to craft is above 500g.

Essence Snow.3194:

No reason aside from the field research?

Whose field research? Because so far i saw none, only hearsay and anecdotal “evidence” that proves exactly nothing.
Mind you, considering the drop chances involved, and numbers of tries in general population, to have anything even close to believable results you’d have to do a documented field research on thousands of people each doing thousands of attempts.
And then you’d have to show that the results you got are meaningfully different than expected.
But of course i guess hearsay and wild accusations are easier to make.

lol @ wild accusations…..fits perfectly with conspiracy theories

It’s evidently very easy to go to the extreme right or left. No way it could be a little off to one side or the other. It must be wild, rampant, and/or conspiracy level.

Take a look at what observational data is/does/can be used for. Then take a step back from the extreme edges and come back to reasonable.

Essence Snow.3194:

lol @ wild accusations…..fits perfectly with conspiracy theories

It’s evidently very easy to go to the extreme right or left. No way it could be a little off to one side or the other. It must be wild, rampant, and/or conspiracy level.

Take a look at what observational data is/does/can be used for. Then take a step back from the extreme edges and come back to reasonable.

That’s what i am advocating. Being reasonable. Claiming there’s something wrong with MF rng is currently not a reasonable approach, unless some proof is supplied. And so far there was neither proof nor even a believable reason for suspicion. Without either, any claims to the contrary can only be called “wild accusations”.
Don’t like that? Find proof to support your belief.

1st you need to understand what reasonable means before claiming to be as such. Trying to justify extreme criticism because something doesn’t fit into your thought process, is not reasonable. It boarders on the edge of irrational and vain naivety.

Again look up observational data. Given it is the only evidence we have access to, it fits rather well.

Mystic.5934:

it would take a HUGE sample size to tell the difference between weighted rng and true rng. of course, there’s Occam to tell us that they likely did not go through this much trouble just to mess with your sanity.

Believe it or not, it doesn’t take a very large sample set at all to tell if items in your sample are independent and random. You can’t figure it out just from looking at aggregate statistics (number of hits and misses), but you can if you record results in sequence and perform some basic time series analysis.

For example, an easy one to perform is to look at the amount of time (or number of samples) between events. If your events are independent and random, the distribution of the time between events should follow a decaying exponential function; if there is any correlation between events, it cannot follow this distribution.

I’ve seen no evidence to suggest that the output of the forge is anything other than independent and random.

One also has to look at a Bell Curve to understand how events pan out. Some people at the far right of the Bell Curve are ‘lucky’. They get more than the average number of good drops. Some people at the far left of the Bell Curve are ‘unlucky’. They get less than the average amount of good drops. Over time though, these groups will shift position within the curve.

People who are very lucky or very unlucky will talk about it. When they drift back to average, they will stop as they will either have nothing to boast about or they are now happy. The ones who are newly at the edges of the Bell Curve will then start boasting or complaining. And this keeps up a constant amount of comments about good or bad luck accounts.

Another point to consider is bias in remembering events and drops. Research has shown that people are very prone to misremembering and unconsciously changing the history of what happened. They also tend to remember the stories around them that bolster their own beliefs and to discount the stories that don’t. If someone believes that his account is unlucky he will especially notice stories of extreme luck on either side, especially since these are the stories most often told even though they might be the most rare. The majority with average luck tend not to talk about it, one way or another.

I still don’t understand how we are to believe you think a “flaw in the code” favors some players over others…. A “flaw in the code” would produce identical “flawed” results for everyone…would it not?

I have a very decent passing understanding of how computer code works (written plenty of DB front ends and various custom hardware code), but maybe you could fill us all in at how bugged code would give one player better results (consistently better, according to your claims) than another player?

I’d buy the secret luck stat (or even ANet intentional weighting) stat before I consider a “flaw” to be skewing precursor drops for some and not others. I may be looking at the problem wrong, but I don’t think so.

Grimm look up Wii flagging. It is a perfect example of unintentional rng bias that went on for over a year before being pinpointed.

Essence Snow.3194:

1st you need to understand what reasonable means before claiming to be as such. Trying to justify extreme criticism because something doesn’t fit into your thought process, is not reasonable. It boarders on the edge of irrational and vain naivety.

Not when that “extreme criticism” is in itself unreasonable. And it is.

Essence Snow.3194:

Again look up observational data. Given it is the only evidence we have access to, it fits rather well.

I did. I have also done some basic calculations. I see nothing out of the ordinary – just the normal, expected results of a high variance rng.

The only thing that i can see now is that some people just don’t understand probability and statistics – which in itself is also not suprising, seeing as they likely never learned this (beyond perhaps some simple basics).
Too bad, because every time someone goes with some unsupported claims that rng is broken somewhere, it makes seeing and reporting situations where it really is broken that much harder.

Really, if you want me to treat your claims seriously, supply some math. And this observational data (meaning: numbers) of yours.

Essence Snow.3194:

Grimm look up Wii flagging. It is a perfect example of unintentional rng bias that went on for over a year before being pinpointed.

Do you have a link as googling “wii flagging” only brings up Wii games with the word “flag” in them (mostly AC: Black Flag right now). I’m interested, but what you stated to search is either very obscure or related to something that requires additional search terms.

To add to what Astralporing is stating, my observation of the data is that many of those that complain about this lack basic knowledge of statistics AND knowledge of what to farm or how to farm them efficiently. They base their “lack of luck” on how many hours they have farmed, not on how many foes (that can actually drop what they are trying to farm) they have slain.

More to the point, since the Forge IS RNG, how many times you have dropped something in is NOT really relevant until that number reaches 1000s of times the possible outcomes due to the HUGE drop table (it’s NOT heads or tails but a mutli-thousand sided die roll).

Oops sry typo wi flag*…not wii.

As for numbers, yes I agree it comes down to number of foes killed, ranks killed, and not hours played.

The forge really gets complex in terms of input and outputs. Players can narrow chances by throwing in all of one type, higher rarity, and higher level to an extent. On the verse side chances can be diminished via throwing random combinations as we all know as well.

I am aware of how things are supposed to work in this game. It’s not hard to figure out. It’s just that what I have observed over the course of 4.6+k hours of friends and guildies, that some have remained as ouliers since the start of the game. Those who started the game and got great drop still continue to get great drops regardless of how many forges or how many foes killed. (some barely leave LA, go out once in blue moon and hit the lotto repeatedly or throw in relatively few random lvl 76 rares and hit repeatedly) On the other hand many other players farm every moment they are on and rarely see an exotic or consistently throw in similar exotic combinations in to the forge only to flush the toilet.

If one plays enough they will see patterns, that some players are either consistently lucky or unlucky over time. It’s the “consistently” and “over time” bits that make me think something is not quite right.

Essence Snow.3194:

If one plays enough they will see patterns, that some players are either consistently lucky or unlucky over time. It’s the “consistently” and “over time” bits that make me think something is not quite right.

Unfortunately, this still falls within expected results – this is just the sideeffect of it being a high deviance system. With the probability chances and deviance you have in MF, you’d have to see literally millions of results before any meaningful patterns would emerge.
For example, with the 100 tries in MF, assuming 1% chance of precursor drop on each try, there’s about 30% (look below) chance of getting no precursor. So, in a crowd of, say, 1000 people, each doing 100 tries, you might get 300 people without precursors (the unlucky ones), and several (look below, again) people with multiple “lottery wins”.
Someone looking at this from the side, without understanding how propability works, might assume that there was something wrong with RNG and that the system preferred some people over others – while in fact it would be nothing like that at all.

So again, if you want to show that there’s something wrong, you do have to supply math.

Note: The chances are sketchy, as i did the math over a month ago, and lost the results since then. Not going to repeat it.
IIRC the % chances were significant for at least 2-4 precursors.

Maybe that it’s a high deviance system that’s the problem then. After all this is a game and the objective of such does not lend itself to things which create such disparities.

Regardless, the same can be said for it working properly. “You have to supply the math that it is”.(if you posted it…plz link)

That AC incident is very interesting, and I guess it does imply that certain characters could mistakenly be handled differently than others. I still don’t buy that there is an issue here (other than in the minds that don’t feel they have been “lucky” in the game).

4600 samples is FAR to small when a random drop of 4 items in the Forge can likely generate 1000s of different items of varying probability weight. Even every one of those were drops of the of the same item (possible outcome is over a 100 items), 4600 is still several orders of magnitude too few to be considered a large enough statistical sample.

If of those 4600 samples, a next higher tier items are dropping roughly 20% of the time, I’d say your theory is disproved right there (unless you are suggesting that the “error” is isolated to generating a precursor only….pretty FAR FAR fetched, imo).

I play a lot, I don’t use the forge all that much but in terms of drops and RNG and the like here’s what I’ve learned:
My account has more “luck” than others I play with e.g. more rare drops, more dye drops, tickets and such. I don’t know how it has happened but It does, and it make people made at me the dragon bash event with the jade tickets, I got one after only 25 coffers, other I play with it took 2000-5000 coffers. With the forge, ill throw in 4 random blue dyes and get back a green, 4 of those and up one level, then another till I have an exotic dye. It is flawed and it makes it not enjoyable to those who spend a lot of hours playing and farming for legendary weapons and such. I cant explain it but it is what it is. so many unless dyes found on a daily basis….

7 lvl 80’s one more on the way.

rng in this game (simplified):
- a random row of values, let’s take numbers(for example 1 100 25 9 89 76 1002 etc.), is attached to character and get’s reset at XY time (daily reset for example); mostlikely you get another row attached after reset
- some numbers represent getting precursor (for example 1002), some numbers represent bad luck basically (for example any number from 1 to 100)
- when you throw stuff in MF, the game goes through function and reads the number from the rng number row that is attached to your account; if the number is right you get precursor, if number is wrong you don’t get it

that is the reason why some ppl get it from 10 tries and some need to do 2000 tries before they get one just because that “lucky” number is located for some people at begin of the row and for some at the end of the row

that said, if you really want good chance for precursor from MF, save as many weapons/mats you need to make one so you can basically do as many tries as possible in short period of time (because of the reset)

keep in mind that that so called rng is obviously affected by different variables as well; for example if you farm too much on one map you get something called loot decay but that is different topic

side note: rng doesn’t really exist in currently used technologies

That is a very interesting theory. Have any proof, or is it all in your head?

(more seriously: that would be an extremely painful, convoluted and inefficient way to implement rng in a game. I’d instantly fire any programmer that would suggest it, with no questions asked).

Okay, what we have to remember here is the following:
Can a logical machine make random numbers ?

http://engineering.mit.edu/ask/can-computer-generate-truly-random-number

Now for a real life example:
ERNIE – which calculates a random number to allocate prize draws (bit like a lotto)
http://en.wikipedia.org/wiki/Premium_Bond

" ERNIE 4 uses thermal noise in transistors as its source of entropy for generating true random numbers; the original ERNIE used a gas neon diode. Pseudorandom numbers, often called simply random, can be recreated by anybody who knows the algorithm used to generate them as they are produced in a deterministic way; true random numbers cannot. "

A machine which generates RNG using a software algorithm cannot be truly random.

I would encourage you to look at: http://www.random.org as a site where the randomness comes from atmospheric noise, which for many purposes is better than the pseudo-random number algorithms typically used in computer programs.

I truly hope that we do not use a software pseudo number random generator and the game uses something like an external service. If the random numbers are generated by individual ANET hardware servers in software or by a clients PC then it is highly unlikely to be truly random.

The following site gives a good example how many software random number generators are calculated: http://www.math.utah.edu/~pa/Random/Random.html

Going even further chaos theory is often used to make sense of what may seem like random events.

I hope ANET is using an external source for randomness and not a software generated algorithm (which is often not very random).

TPMN.1483:

“Pseudorandom numbers, often called simply random, can be recreated by anybody who knows the algorithm used to generate them as they are produced in a deterministic way; true random numbers cannot. "

A machine which generates RNG using a software algorithm cannot be truly random.

What if you know the algorithm, but do not know the input data for the algorithm? Even if the algorithm would be produced in a deterministic way, you’d still be unable to recreate the end results. Most pseudorandom generators work that way. Yes, it’s not truly random. No, you can’t predict the results.

And, you know, technically things like dice throwing, roulette spin and even weather would not be truly random by that definition. It’s just that, while “algorithms” for those do exist, they are beyond observer’s ability to compute.