Guild Wars 2

EDIT 2:

my hypothesis:
It is assumed that swamp is rolled for for both fractals.

Persons A,B,C,D,E,F,G,H,I have not run any Fractals yet today, not even setting foot in Fractals.
Person A runs a Fractal 37 with persons B,C,D,E, and receives reward X
Person A then runs a Fractal 50 with persons F,G,H,I, and receives an eerily similar Fractal roll, and receives reward X again.

If anyone has the opportunity to test this independently and corroborate or disprove, please do!!!

I ran a 37 today, and got
Swamp (Bloomhunger)
Molten Weapon Facility (Tar Elementals, Red Oozes, Spiders, Earth Eles, Blue Oozes)
Grawl Shaman
Molten Boss
Reward: Vine of the Pale Tree (Infused)

I then ran a 50 later (with different people), and got
Swamp (Bloomhunger)
Molten Weapon Facility (Tar Elementals, Red Oozes, Spiders, Earth Eles, Blue Oozes)
Aetherblade(The ONLY difference)
Molten Boss
Reward: Vine of the Pale Tree (Infused)

Thing is, this is not the first or even second time something like this has happened- suspiciously similar fractal rolls, with the exact same reward. The probability of this happening even once is miniscule- multiple times, much much less.

There’s something going on in Fractals. Has anyone else experienced this happening, or should I go buy a lottery ticket?

EDIT: I should mention that I don’t run a 37 before my 50 often- but when I do, I am afraid of this phenomenon occurring, it’s happened that much. One time I remember in particular, I got two of the same uninfused ring after the exact same fractal rolls.

It’s just RNG. If you keep track of it more consistently you’d notice. the thing is there’s no way for the game to decide to give you the same rewards and shards. Also if you think about it…how could this happen when there are 4 other people that are likely to have different shards in their previous lower level fractal, then they all come together for a 50 which means it’s impossible for everyone to have the same shards as the previous lower level run.. I run 40 and 50 almost everyday and I’ve been doing it forever. Sometimes it seems like there is a pattern but really there isn’t.

Of course the RNG is broken, stuff like this happens way too often to be true random.

“Also if you think about it…how could this happen when there are 4 other people that are likely to have different shards in their previous lower level fractal, then they all come together for a 50 which means it’s impossible for everyone to have the same shards as the previous lower level run..”

I run a 50 with guildies. They do not run Fractals before the 50- some do wurm and teq raids, others do dailies and such.

The fact that once I saw molten weapon facility, I sarcastically said “Oh, I wonder if we’ll get Grawl shaman and then a Vine of the Pale Tree”, shows that this does in fact, happen often enough to not just be RNG.

Perhaps if one of them had also run a fractal earlier it would have made things better?

It was an example. There are plenty of people (including me) who run multiple fractal levels a day with mostly pugs. It is impossible for those people to get all the same fractals every time when doing a different tier that day as I explained in my previous post. As in: Person 1 gets- swamp, cliff side, molten. Person 2 gets swamp, ascalon, jade maw. these two people both ran level 37 in different groups. They join up to do a 50. They obviously can’t all get the same shards.

If you really want me to explain the psychology behind why you think it isn’t RNG I can but that would be asking a bit much of me at this point as its only a bug forum.

And yes. If one of them had ran a fractal earlier (in a different group than you), had the same exact shards as you had earlier/second time. And received the same reward. And this happened consistently as in more than once a week at least. Then yes I will apologize and say it is not RNG.

Ltomato: if you’re convinced it’s not random, collect the evidence. For ANet to start considering an issue, they’d have to see results for a couple of dozen runs (which could be just you or you plus guildies).

I agree with the previous posters that once you start tracking, you won’t see any pattern.

We humans are really good at spotting patterns (it’s one of our survival traits), but we’re also good at seeing patterns where there aren’t any, especially when it comes to random mechanics in games or nature. One of my favorite examples is clustering, which you can see demonstrated on wikipedia. A random distribution of stars in an imaginary sky shows patterns (like constellations), whereas a perfectly even, non-random distribution shows no patterns, i.e. we easily see a pattern in the randomness, but cannot see any pattern in a predictable mechanic.

I don’t run 37s often BECAUSE I am afraid of this EXACT PHENOMENON happening.

Like, to put that into perspective, I’ve encountered a 1/2500 chance (minimum- that’s just for the same infused ring, much less the same fractal rolls or uninfused ring repeat) often enough that it has deterred me from running multiple fractals a day.

The fact that if I run with the same person for 37 and 50 and it doesn’t happen doesn’t mean that it isn’t consistent at all.

The observation I’ve made is that it happens when one person runs with two fresh groups (groups that have not done ANY fractal content at all that day)- that one person gets the same Fractal roll twice.

It’s very hard to test what I think duplicates it, since it involves 2 completely different Fractal groups. Group 1: has not run any fractal yet that day, and runs a 37. Group 2: consists of one member from Group 1, and the rest having not run any yet.

This scenario is difficult to control for, and is not likely to happen when pugging- hence why most players who either run 37/50 with the same group, or pug one or the other, don’t encounter this.

I can try to collect more data, but it will take a long time, since I do not actively try to do 37s daily as a result of this- nor is the group that I do 37 with around often.

So let me try to clear this up- There are not a lot of points on my 37-50 in one daily line. On that line, a disproportionate amount (To the point where it is predictable) land on or near the same spot. And you guys are saying “Well, if your sample size is large enough, anything can happen.” My sample size isn’t large.

To repeat my hypothesis:
Persons A,B,C,D,E,F,G,H,I have not run any Fractals yet today, not even setting foot in Fractals.
Person A runs a Fractal 37 with persons B,C,D,E, and receives reward X
Person A then runs a Fractal 50 with persons F,G,H,I, and receives an eerily similar Fractal roll, and receives reward X again.

If you can find 9 people to disprove my hypothesis, please do. Like I’ve said before, I don’t run 37s enough to gather more data in a timely manner.

So it happenend once or twice, which is why you don’t do it often, but you’re convinced there is a pattern anyway? Even though you said it doesn’t happen all the time?

And you said your “sample size” isn’t that large. You know what’s also not a large sample size? The guy who got a precursor from a guild mission and on the same freaking day pulled another one out of the forge. Now I could call this a pattern, but I highly doubt there is one.

There are eleven fractals, not all of which can be first, second or third fractal interchangeably, plus three boss fractals. There are bound to be runs that are nearly identical. Just because I got the stupid hammer fractal three times in a row for second fractal doesn’t mean RNG does this on purpose (which would be the complete opposite of RNG). This isn’t the lottery; chances are your numbers will be pulled more than once in a blue moon. And that ring you got twice… how often did you receive it in total? Because getting the same ring twice isn’t that huge of a coincidence (I have two identical rings from fractals, and I maaaybe do a run twice a month on average. Maaaybe.)
If you try to reproduce this mystery bug, I’d wager you’ll see it doesn’t happen all the time, not even most of the time, even if the circumstances are always identical.

Twice in a row. Yknow. On the same day. Consecutively. That’s two 1/50 chances (rounded way down), in a row. That’s rolling 1d50 twice and getting the same outcome. And that’s just for the rewards, much less the Fractal encounters.

It’s happened at least 4 times that I remember. The first time, it was a coincidence. The second time, it was interesting. The third time, it was weird. The fourth time, there’s something going on.

This is like someone who doesn’t do guild missions much getting the SAME precursor from the SAME guild missions rolls 2 weeks in a row. It’s like a lightning strike. Except it’s happened four times.

I’m sorry we will have to agree to disagree. I will tell you I’ve run at least 1,750 fractals which is 7,000 shards. I ran 4 a day with different charaters before Fractured when rewards were charater bound. I’m telling you this not to brag but to just show you that I have a lot of experience in this area and my opinion should have some merit. I’ve noticed patterns before but that’s just because of RNG.

Anyway good luck to you

You guys don’t understand. I have a hypothesis on how to duplicate this that merely needs testing. If it causes this, then it’s a bug. If it doesn’t then it’s really really weird rng.

Believe me I’ve run my fair share of fractals as well. Judging by how often reward rng gets messed with unintentionally and intentionally, especially since after Fractured, I would not be surprised if there were some really obscure logic error that caused this.

Instead of being dismissive, maybe try out my hypothesis.

Like, take 2 apple trees. I’m saying this apple tree always has apples with worms. You guys are saying, oh, you must be unlucky, I’ve eaten thousands and thousands of apples from this other tree and I don’t get a lot of worms. And I’m saying go and eat an apple from the other tree.

Only the fractals are all on the same tree…

We do understand, but until you can get proof, most here will doubt there is any sort of pattern because we all experienced runs like these, and know its RNG being an idiot.

In science, you have to prove your hypothesis before it is accepted as truth. You are welcome to do so. I can’t, since my fractal level is a measly 21 (see, kids, that’s what you get when you run 19 over and over again…) and I don’t PUG them either, but maybe someone else gets curious and goes to see for themselves. Until then, I will stand by it that this is just RNG being weird.

Ltomato.8649:

You guys don’t understand. I have a hypothesis on how to duplicate this that merely needs testing. If it causes this, then it’s a bug. If it doesn’t then it’s really really weird rng.

I’d be more enthusiastic to try this if the suspected pattern led to fractal skins, and not a ring

sMihaly.1492:

Ltomato.8649:

You guys don’t understand. I have a hypothesis on how to duplicate this that merely needs testing. If it causes this, then it’s a bug. If it doesn’t then it’s really really weird rng.

I’d be more enthusiastic to try this if the suspected pattern led to fractal skins, and not a ring

And others may have been more interested in helping out if the hypothesis was based on more than 2 runs.

Fractal 1: 25% chance to get Swampland, you can’t use this one in your hypothesis if you reroll for Swampland. Note that all choices for this Fractal except Swampland and Aquatic Ruins are options for Fractal 2.

Fractal 2: 20% chance to get Molten Furnace (previous Fractal is not one of the ones repeated, if it was it would be a 25% chance get Molten Furnace). Note that the Cliffside option in Fractal 2 is also a choice for Fractal 3.

Fractal 3: 20% chance to get Volcanic (25% if the previous Fractal had been Cliffside).

Fractal 4 (Boss Fractal): 33% chance to get Molten Boss.

Since your Fractal option is truly an independent probability, none of your Fractal levels are in more than one level, the chances are pretty high that in 2 runs, even on the same day, would land on the same Fractals.

As for the reward, 2 runs is not enough to say there is anything funky with the reward RNG.

I did say that this has happened at least three times before- consecutive Fractals being ridiculously similar.

There’s also the same exact mob set in Molten, which is at least 25% on top of that.

So yes, it is enough to say that there is something weird with the rng.

50% from Bloomhunger/Mossman
20% to get Molten
25% (at most) to get the same exact mob set
33% to get Molten boss
2% to get same ring reward.

That’s 1 in 6000, which, I’ll admit, is within the realm of possibility for happening once.

If we take the amount of pristine relics I have, 564, subtract say, 20, for daily bonus relics, what are the chances in 544 rolls of getting a 1/6000 chance at least 4 times?

Probability isn’t my strong suit so lemme know if I’m doing this wrong.

Chance of getting this particular roll+reward = 1/6000.

Therefore, the chance of getting it twice in a row is (1/6000)^2.

Which is like, something *10^-8

I’m SURE I’ve gotten duplicate Fractal rolls already before, multiple times. I understand that. That’s not a problem. That’s just RNG. But when it happens consecutively multiple times, then there’s something to maybe examine and see if there is something going on.

Except you got aetherblade for your 3rd instead of shaman as described in your original post. I don’t have time to test your hypothesis. When I join groups I’ll ask if they ran fractals earlier that day or not but I am not going to leave the group if one person already ran a fractal that day. If you really want to get a Dev response or for this community to recognize your claim as at all legitimate you will test it and report back. However until that point I am done responding to this thread so good luck to you.

You’ll notice I left the 3rd one out of my calculation.

I’ll test it when I can, but the opportunity for me to test it is so rare that I can’t really predict when I next try it. Hence why I’m posting my hypothesis from my personal experience on the forums where there may be more people who may have the opportunity to test it.

Ok that makes sense. Although it’s still hard to understand your hypothesis. It becomes convoluted when you don’t have to have all shards be the same. Because if it is a bug it will follow a pattern. So your hypothesis has to be something like…. When running 2 fractal runs in a day where no one else that day had run fractals for the second run. You expect to get the same ring or rewards and the same shards for the second run except the third shard will be different.

But the thing is, 2-3 times is NOT a large enough sample to say that something is off with RNG. No where near.

Perfectly random assumes each option has an equal chance of occurring.

If I flip a coin 5 times and it lands on heads 5 times, then you’re saying because it happened more than once that the coin isn’t balanced. But that’s not enough trials to say that it’s not just pure chance. Streaks can happen. Probability is only true on the LONG run. Not in the short run. You’re still well within the realm of the short run.

If you’ve documented at least 10 trials, and got 8 out of 10 showing the same Fractals and reward, then you’ve got case for something being out of whack.

But you’ve only done 3 trials. 2 of which showed the trend. 1 did not.

Scientists test their hypotheses HUNDREDS of times before saying it’s true. Because what if it was a fluke? What if there was a variable they forgot to account for? They can’t just do it two or three times and then say it’s true. It’s basis for beginning to look into the problem to come up with a hypothesis. To make sure there aren’t any other variables at play.

You haven’t even made sure that your bug is just limited to levels 37 and 50. What if it’s the 31-40 tier and the 50? What if it’s two fractals that aren’t on the same tier? So a level 9 and then a level 23 will have the same anomaly.

Does it only occur if a specific set of Fractals happen during the first one done of the day? Or does it duplicate all runs?

That’s another issue with the too low number of runs. You’ve left too many variables on the table as a cause as well as how widespread the problem is.

There are more things not truly random.
I am keeping track of the black lion goods from daily login rewards. And the distribution is very weird. Example on some of my new accounts I have 3-4 black lion salvage kits already. On others 0 kits but 3+ of a different item. So far only a single black lion key on all of my 30 accounts.
I guess each account has its own seed which affects the likelyhood of getting a specific item :/

Malediktus.3740:

There are more things not truly random.
I am keeping track of the black lion goods from daily login rewards. And the distribution is very weird. Example on some of my new accounts I have 3-4 black lion salvage kits already. On others 0 kits but 3+ of a different item. So far only a single black lion key on all of my 30 accounts.
I guess each account has its own seed which affects the likelyhood of getting a specific item :/

There is nothing account specific that makes one account any more or less lucky than another.

And unless you’ve had over 100 black lion chests from the log in on both accounts each or over 200 total, there’s not enough numbers to say with any certainty that the generator is bugged or even say it isn’t bugged. And without knowing the intended drop rate it is impossible for us players to say one way or another.

And since black lion chest daily reward is an entirely different aspect of the game to fractals, put any research or calls for assistance in the research into its own thread.

If I flip a coin 5 times and it lands on heads 5 times, then you’re saying because it happened more than once that the coin isn’t balanced. But that’s not enough trials to say that it’s not just pure chance. Streaks can happen. Probability is only true on the LONG run. Not in the short run. You’re still well within the realm of the short run.

What makes my case stand out is that it is a literal 1 in a million chance, that has happened to me multiple times.

Let’s say my # of fractals is 540. Let’s say of those, half are 37, and half are 50 (not even close). So that’s 270 trials. In less than 270 trials, I rolled 4 1 in a million chances. That’s not within the realm of physical probability.

You haven’t even made sure that your bug is just limited to levels 37 and 50. What if it’s the 31-40 tier and the 50? What if it’s two fractals that aren’t on the same tier? So a level 9 and then a level 23 will have the same anomaly.

Does it only occur if a specific set of Fractals happen during the first one done of the day? Or does it duplicate all runs?

That’s another issue with the too low number of runs. You’ve left too many variables on the table as a cause as well as how widespread the problem is.

That’s why I started with a narrow hypothesis. If it works for that, then you can add more variables and experiment. As I’ve mentioned before, even setting it up to try this is difficult as it requires 8 other participants, and quite a lot of time.

The key to bug diagnosis is getting it duplicatable. If you can find a way to duplicate it, then you can start adding or taking variables to see when it works or doesn’t, which lets you narrow down where the bug is.

EDIT: with the changes to daylight savings time I will not likely be able to test this, as my normal fractal 50 group goes 30 minutes after dailies roll over.

But I can flip an actual balanced coin 1000 times and have it land on heads every time. It is possible. Your case is possible. Therefore only a handful of times experiencing it is not enough to even warrant the devs looking into it. Especially not with the number of variables left unknown as well as how widespread the glitch really is.

But I can flip an actual balanced coin 1000 times and have it land on heads every time.

Sure, 1 time out of a number with way too many zeros.

I don’t expect it to be looked into, but I provided my report, a means that I believe will duplicate the scenario, and can start off more research should that be proven true by people other than me. I shouldn’t have to conduct a research study in order for an issue to be taken seriously.

Ltomato.8649:

But I can flip an actual balanced coin 1000 times and have it land on heads every time.

Sure, 1 time out of a number with way too many zeros.

I don’t expect it to be looked into, but I provided my report, a means that I believe will duplicate the scenario, and can start off more research should that be proven true by people other than me. I shouldn’t have to conduct a research study in order for an issue to be taken seriously.

Yes, highly unlikely, but still possible and doing it twice a row is still possible. One major outlier, but still possible. And at that point, it’s open to more testing. Same force from the same height with the coin oriented the same way each time for the flip would bias a balanced coin to one side over the other. So the coin is not imbalanced. But that’s impossible to determine by just 2 runs where you weren’t really carefully noting any and all possible variables that may have accounted for the outlier.

Not saying that it shouldn’t be taken seriously. But you can’t call it a bug at this point. Not after just 2 trials.

What you could do is make a post down in the Fractals and Dungeon subforum stating that you’ve had an oddity in the way Fractals and rewards are chosen for you and if any one else has had the same issue.

The fractal frequenters may have more data points to run on. I’d imagine the glitch, if it happens the way you theorize, would happen if 5 people all had the same shards in their last run. Which would happen if a group did two runs back to back. So they may already have more trials for you to add to your collection. And may be willing to test the theory.

Bug forum isn’t going to draw a lot of people who run fractals frequently. And they’re the only ones who are going to be anywhere near reliable to be able to do 2 runs per day.

Yes, highly unlikely, but still possible and doing it twice a row is still possible. One major outlier, but still possible. And at that point, it’s open to more testing.

“possible”, the amount of times you’d need to flip a coin 1000 times is at least 10^77 times more than the number of seconds the average human lives for.

I’d imagine the glitch, if it happens the way you theorize, would happen if 5 people all had the same shards in their last run. Which would happen if a group did two runs back to back. So they may already have more trials for you to add to your collection. And may be willing to test the theory.

Which shows me you didn’t read or understand my hypothesis at all. I’m not describing a commonly practiced situation. It’s very unique, and not likely to occur in most cases.

A group of 5 would not be able to contribute to my hypothesis, since you need 9 people.

I don’t know what causes it. It could have something to do with one player’s rolls carrying over and not being refreshed in some data structure or the database when other players start fresh. There’s lots of ways to approach and tackle bugs- bugs which can be very weird. Games aren’t real life. You can’t generalize bug cases without first acknowledging specifics.

For example, https://www.youtube.com/watch?v=JxgEXDnXD6M by doing a series of seemingly arbitrary frame perfect tricks, a computer can reprogram Super Mario World.

You can’t say, oh, well if it works from doing these tricks, you could do them on any level or with different items and obtain the same result.

And 2 trials successful (even though I mention that this has happened at least 4 times) is already a probability of 1 in 10^24- it would be more probable for me to spontaneously combust.

And the Fractal running groups may be able to provide more data to work with. Not to mention others interested in Fractals as well. You’re problem will be seen by more players able and willing to help you generate more data.

And you can only count the trials you were actively noting what levels you got and rewards.

But if it is a bug with a player’s rolls not refreshing, it could still affect a group of 5 who do two runs back to back. It’s worth getting some data from those players. To rule it out. Or to find out that they are included.

Bug testing needs to find out how widespread the bug is. Just what conditions trigger the bug. Will an unbugged set of rolls override a bugged set? Is it just levels 37 and 50 or could it happen to 24 then 25 or 24 then 39? Is it only when you get a magic combination of Fractals or is just one Fractal enough to bug it out?

That’s the kind of questions left unanswered by low number of trials. You aren’t generating enough data to even begin to pinpoint where the glitch occurs. Not to mention 2 runs isn’t enough data to say that your experience is such an outlier it means that something’s wrong.

You aren’t generating enough data to even begin to pinpoint where the glitch occurs. Not to mention 2 runs isn’t enough data to say that your experience is such an outlier it means that something’s wrong.

As much as I would LOVE to have a job where I can bugfix and test things for GW2, that’s not my job.

Nor should a low number of cases prevent a report from being taken seriously and recorded.

Could you imagine if a duplication glitch got that treatment?
“Uh, hi, I accidentally duplicated a stack of ectos while doing x”
“Did you try y 100 times?”
“No, I did x a couple times, but then I got scared and stopped.”
“Well, that’s not useful, try y 100 times and z 100 times, and then report it.”

I encountered this issue while doing something specific. I reported this issue.

Not to mention 2 runs isn’t enough data to say that your experience is such an outlier it means that something’s wrong.

A 1 in 1000000000000000000000000 chance occurring is not just an outlier.

An outlier is when I get 5 Fractal Tridents before getting a Fractal Pistol. An outlier is only getting ascended rings or nothing for 3 weeks of Fractal 50.

That’s RNG.

A 1 in 10^24th is not. That’s as many seconds as 14 million times over the current population of the earth will live.

That’s a once in 14 million generation achievement.

So yes, that is significant enough to say “something is up here”

No it is not an indication that there is definitely something wrong. It’s worth getting others to help test since it may be a bug, it is one major outlier if it isn’t a bug. What’s causing the bug? Who knows. Not enough data to really say it’s anything we can control or not until ANet decides to fix it.

And I never said it shouldn’t be taken seriously. I just don’t know how they’re going to try to find a bug when it could be affecting more than just levels 37 and 50 being done in the same day and in the same order. When the glitch could occur under any number of circumstances. Needle in a haystack comes to mind, especially if it’s not something that happens every time you do it: as then it IS related to something other than just one person doing level 37 and then level 50 with both times being in groups where all 4 other party members have not done any Fractals that day (if that was the sole cause, it would happen every single time). Or there could be nothing wrong and you’ve just hit some bad/good RNG and should buy or not buy a lottery ticket accordingly. Or the issue could be related to something entirely unrelated to the fact that during both runs none of your party members had done a Fractal that day.

So the only way is to get others to offer up their data. And the best place to get that data is the Fractals and Dungeons subforum. Not the Bug forum.

I just have an issue with you saying it is a bug and it’s caused by doing two Fractals in a row and both times having no party members who have done Fractals that day. You very likely are experiencing a bug (that’s not a certainty statement). Not enough data to really speculate on the cause. No party members having done Fractals is just a possible cause.

But it could be something that occurs only if you have those conditions and run Swamp at the beginning if your groups reroll for swamp. Or if the first run hits Swamp and Molten Boss, with no regard to having done fractals that day since bugged ones override nonbugged ones.

Ltomato.8649:

As much as I would LOVE to have a job where I can bugfix and test things for GW2, that’s not my job.

Nor should a low number of cases prevent a report from being taken seriously and recorded.

Actually, a low number of cases should prevent a bug report from getting resources deployed. I would want my QA team working on high-priority, wide-scope issues primarily. If a tiny fraction of fractal players are claiming an issue that others cannot see, I’d be inclined to think that there are a dozen other explanations that don’t involve game code: observation bias, clustering bias, and so on.

As I said above, if you really think there’s a problem, start collecting data. If you run fractals often, it won’t take very much time for you to expose a pattern (or a lack of one).

As a counter example, let’s say I had an American quarter and flipped it 10 times, with only a single head. The odds of this are less than 1 chance in 100: low, but not crazy low. Should we investigate whether my coin is unbalanced in favor of tails?

Now consider that there are 10,000 other people with the exact same coin. Some aren’t paying attention to what happens, but a few are. From those reporting back, we have a person or two seeing exactly 1 out of 10 heads, but most are seeing 4-6, as we would expect. Should that make us more or less likely to investigate the initial report?

Finally, regardless of whether any player reading this is convinced there is an issue or not, there’s also the matter of realpolitik: the powers that be are swayed by data, not by anecdotes. I’ve gone through this with many people in many situations (developers, business executives, education professionals, …) and it’s always the same: management is convinced when you can show them numbers that demonstrate there’s an issue of concern. Without it, you can type or talk forever and they won’t take action.

For the OP, it’s perfectly fine if you don’t want to run 100s of fractals on your own. But you can still coordinate efforts among others, from your guild or using this thread. Tell people exactly the steps you want them to use and how you want to collect the data (in-game mail, forum, whatever). It doesn’t necessarily take a lot of time (it’s mostly just careful bookkeeping really). Once you have some rigorously-collected data, you won’t have to talk most of us into supporting further action.

Ltomato.8649:

I don’t know what causes it. It could have something to do with one player’s rolls carrying over and not being refreshed in some data structure or the database when other players start fresh. There’s lots of ways to approach and tackle bugs- bugs which can be very weird. Games aren’t real life. You can’t generalize bug cases without first acknowledging specifics.

For example, https://www.youtube.com/watch?v=JxgEXDnXD6M by doing a series of seemingly arbitrary frame perfect tricks, a computer can reprogram Super Mario World.

As a fan of TASvideos, I can say he managed to do something that is widely demonstrated to be not only repeatable but programmable. But your analogy breaks down somewhat in that gaming console RNGenerators are often highly predictable, because they have no access (or lack the power to access) some kind of data value that can arbitrarily change. When we transfer that to a PC game, the mob of players connected to it at any point are themselves a source of RNG. The devs just have to pick one aspect of some random player’s interaction at any particular moment, scale it appropriately, maybe adjust it by the final few bytes of a clock tick, and insert it where it is needed.

And 2 trials successful (even though I mention that this has happened at least 4 times) is already a probability of 1 in 10^24- it would be more probable for me to spontaneously combust.

I’d comment more, but it begins to muddle what you are asserting here (these independent variables with these ranges in this combination recurring is unlikely in a random situation) with what you are trying to prove (these things aren’t random). I don’t think all of your assumptions are correct, and that some of the variables are somewhat dependent on the others. For example, by choosing a particular starting fractal, you limit the options of the ones that might follow it. That would affect the fractal order recurring, but not necessarily the final award.

Quick test. I figure you ran one of your trials with friends or guildies. Did the whole party also get rewards that matched what they got the previous time?

Sariel V.7024:

Quick test. I figure you ran one of your trials with friends or guildies. Did the whole party also get rewards that matched what they got the previous time?

That’s the thing.

In the group for 37 no one had done Fractals that day.

In the later group for 50 no one but the OP had done Fractals that day.

OP literally only has the data for two trials. And claims that’s enough to say it’s not random and that the cause has been pinpointed.

Actually, a low number of cases should prevent a bug report from getting resources deployed. I would want my QA team working on high-priority, wide-scope issues primarily.

And that’s fine. I’ve already mentioned that this isn’t a very common issue and that I don’t expect it to have resources assigned to it. I just want it recorded in case it is an issue where it’s someone who worked on the code can go “OH crap, that thing!”

And I’ve also mentioned that I’ve experienced this 4 times, all 4 times which support my hypothesis. So no, I don’t literally only have data for two trials.

I didn’t just see this happen and randomly guess at a cause. I didn’t say “Oh, what a coincidence, I bet it’s a bug.” I experienced this multiple times in the same conditions, to the point where I predicted my drop on the 4th time it happened. Cynically predicted, and hoped that it wouldn’t be true, but predicted nonetheless.

I just wanted the issue to be recorded- that’s all.

For example, by choosing a particular starting fractal, you limit the options of the ones that might follow it. That would affect the fractal order recurring, but not necessarily the final award.

I did have that in my calculations. Assumed swamp start, and only added a roll for bloomhunger vs mossman.

Ltomato.8649:

Actually, a low number of cases should prevent a bug report from getting resources deployed. I would want my QA team working on high-priority, wide-scope issues primarily.

And that’s fine. I’ve already mentioned that this isn’t a very common issue and that I don’t expect it to have resources assigned to it. I just want it recorded in case it is an issue where it’s someone who worked on the code can go “OH crap, that thing!”

And I’ve also mentioned that I’ve experienced this 4 times, all 4 times which support my hypothesis. So no, I don’t literally only have data for two trials.

I didn’t just see this happen and randomly guess at a cause. I didn’t say “Oh, what a coincidence, I bet it’s a bug.” I experienced this multiple times in the same conditions, to the point where I predicted my drop on the 4th time it happened. Cynically predicted, and hoped that it wouldn’t be true, but predicted nonetheless.

But you don’t know that that’s the cause.

And can you remember exactly what levels you got on those runs and what the reward was?

What if it’s getting a specific fractal set in the first run that bugs it out? Testers at GW2 HQ test out your hypothesis, don’t get your fractal set and don’t get the glitch so they go: just a bad run of luck. Which would be very disappointing if there was an actual glitch.

I don’t see what your aversion is to posting this over in the Fractals and Dungeon subforum. They’ve got a Forum Specialist assigned to that specific forum and if enough data was collected to give ANet a hey, there’s a bug in the RNG system under X conditions, that the forum specialist could pass that on in one of their reports to ANet.

Fractal frequenters already dislike the bad RNG for the rewards. They surely don’t want a glitch that would get them two of the same item, especially rings since one character can only wear one of each ring at a time.

I don’t see what your aversion is to posting this over in the Fractals and Dungeon subforum. They’ve got a Forum Specialist assigned to that specific forum and if enough data was collected to give ANet a hey, there’s a bug in the RNG system under X conditions, that the forum specialist could pass that on in one of their reports to ANet.

I don’t have time to manage such a thread or the data that comes from it if it kicks off (unlikely, since it takes a group of 9 people to even test it).

Nor should not pinpointing the cause be a deterrent from posting bugs. Sure, it helps majorly. It makes things much easier for everyone involved in the process of fixing the bug. But expecting a perfect bug report for something so complex to be reported at all is silly. I’ve narrowed it down as much as I reasonably can.

I saw this mostly as a curiosity, and I posted it. If you want to follow up on it, please, be my guest. I personally don’t have the time to do so.

Ltomato.8649:

And I’ve also mentioned that I’ve experienced this 4 times, all 4 times which support my hypothesis. So no, I don’t literally only have data for two trials.

Right, I never thought you were reporting something based on zero data. What I have been trying to say is that it’s nothing close to enough data to get anyone at ANet to make a note about it. There are lots of other possible explanations.

I don’t blame you for not wanting to collect more data (I do this all the time and it’s both challenging and tedious to do it well). I do want to help you manage your own expectations about what is likely to happen without such data, which is not much.

(If you do to decide to organize collecting data from others, as well as yourself, I would be happy to assist you with putting together a post that makes it clear what would help.)

The other thing you might want to consider is that if you are able to present more data (but not enough necessarily to convince ANet), you would be able to get some of the people reading the forums to look more closely. The Great Ecto Drop Rate Nerf of 2013 was, in fact, identified through this method: a few people posted the kernel of sufficient data and it was odd enough to get other people to post their own numbers, resulting in enough data to get ANet to look closely. (Of course, it’s far easier to collect ecto drop rate data than fractal data of any kind.)

Ltomato.8649:

I don’t see what your aversion is to posting this over in the Fractals and Dungeon subforum. They’ve got a Forum Specialist assigned to that specific forum and if enough data was collected to give ANet a hey, there’s a bug in the RNG system under X conditions, that the forum specialist could pass that on in one of their reports to ANet.

I don’t have time to manage such a thread or the data that comes from it if it kicks off (unlikely, since it takes a group of 9 people to even test it).

Nor should not pinpointing the cause be a deterrent from posting bugs. Sure, it helps majorly. It makes things much easier for everyone involved in the process of fixing the bug. But expecting a perfect bug report for something so complex to be reported at all is silly. I’ve narrowed it down as much as I reasonably can.

I saw this mostly as a curiosity, and I posted it. If you want to follow up on it, please, be my guest. I personally don’t have the time to do so.

But your posting as if it is a bug and if your hypothesis is the correct one and that you have identified the cause.

And when I suggested getting data from other groups that would have similar situations as a way to help narrow down the possible cause, you said no that that wasn’t your hypothesis and that you didn’t need to do that.

And like Illconceived, I’m also willing to help collect and analyze the data. We’ll want to ask that testers not roll for Swamp since that may be part of the actual cause. Or at least we’ll need a decent number to not roll.

Seera.5916:

Malediktus.3740:

There are more things not truly random.
I am keeping track of the black lion goods from daily login rewards. And the distribution is very weird. Example on some of my new accounts I have 3-4 black lion salvage kits already. On others 0 kits but 3+ of a different item. So far only a single black lion key on all of my 30 accounts.
I guess each account has its own seed which affects the likelyhood of getting a specific item :/

There is nothing account specific that makes one account any more or less lucky than another.

And unless you’ve had over 100 black lion chests from the log in on both accounts each or over 200 total, there’s not enough numbers to say with any certainty that the generator is bugged or even say it isn’t bugged. And without knowing the intended drop rate it is impossible for us players to say one way or another.

And since black lion chest daily reward is an entirely different aspect of the game to fractals, put any research or calls for assistance in the research into its own thread.

In the case of [Chest of Black Lion Goods] there is the additional issue of the fact that each one is a different container with different item IDs. Each one having different loot tables would not be surprising(why else would you make 4 different ones?).

Khisanth.2948:

Seera.5916:

Malediktus.3740:

There are more things not truly random.
I am keeping track of the black lion goods from daily login rewards. And the distribution is very weird. Example on some of my new accounts I have 3-4 black lion salvage kits already. On others 0 kits but 3+ of a different item. So far only a single black lion key on all of my 30 accounts.
I guess each account has its own seed which affects the likelyhood of getting a specific item :/

There is nothing account specific that makes one account any more or less lucky than another.

And unless you’ve had over 100 black lion chests from the log in on both accounts each or over 200 total, there’s not enough numbers to say with any certainty that the generator is bugged or even say it isn’t bugged. And without knowing the intended drop rate it is impossible for us players to say one way or another.

And since black lion chest daily reward is an entirely different aspect of the game to fractals, put any research or calls for assistance in the research into its own thread.

In the case of [Chest of Black Lion Goods] there is the additional issue of the fact that each one is a different container with different item IDs. Each one having different loot tables would not be surprising(why else would you make 4 different ones?).

Possibly, but black lion chests are not the topic of this thread. Any research as to whether or not there is something odd with the RNG to it should be in its own thread. Like I told Malediktus.