Guild Wars 2

My Guildmate had his 10th or 11th precursor this weekend, simply logged in for an our, dropped his yellows into the forge. he doesnt even remember it’s name, because in the last half year he got one every month, so precursor is a pretty common thing to him.

I had one, Venom, almost 2 years ago

In our guild, with 30 ppl, there are not one other, with this luck.

Anyone has similar experiences?

(Yes, I’m jealous)

I’m new to GW2 and am light a moth to a flame, attracted to ‘pretty’ cosmetics—Have used the ‘mystic toliet’ on all my rares/exotics….NADA.

Stopped doing it..i just sell/salvage stuff now—at least I get something in return.

inB4itsjustRNG

yup.. yesterday one of our typical suspects pulled out yet another pre right after the patch. We only have two guys in the guild (one has 13 the other over 20 I think) that keep pulling out stuff from the toilet.

personally… nada

the worst thing is that one of them still keeps complaining about the game alllll the time….

Why has there been so many conspiracy theorists active lately?

Lazaar.9123:

Why has there been so many conspiracy theorists active lately?

Because this instance happens regularly and it usually irritates everyone.

Maybe they also do a lot more forges than your other guildies? If they do 100 forges a day on average while other members of you guild do 5-10, then they will get 10-20 times more precursors in the same amount of time (on average). Many people in my guild have had precursor drops while I never had one. However, many of those play at least twice as much a day, which does matter.

Lazaar.9123:

Why has there been so many conspiracy theorists active lately?

I think there are a lot of factors that affects the prec drop: MF%, number of tries, and so on, and I don’t think there is a numeric value in the account database, where the prec drop chance isstored and it’s different for each account.

But you cannot deny it, it’s pure luck, when you log in, go out to some map, and the you get a pre from the first mob

I made my Rodgort, with buying the pre, and grinding all the mats, money, obsidian to make it

Lazaar.9123:

Why has there been so many conspiracy theorists active lately?

It’s well known that instead of coding the one loot table for everyone, games developers prefer to spend vast amounts of time making fiendishly complicated systems so that some of their players get better stuff than others. Accounts of this are always supported by large quantities of well-documented evidence, which takes into account all other factors that might be influential, so there’s really very little doubt about it.

Mystic toilet is waste of time and nerves… just salvage everything and sell mats. Grind enough gold and buy precursors from TP. If you are not born under the lucky star, your chances with mystic toilet are close to 0… comapre it to 100% chance with gold grinding ;-)

There is nothing legendary about legendaries anyway, anyone with credit card can get them instantly so save yourself the disappointment and just buy them with gold.

Galespark.7835:

Maybe they also do a lot more forges than your other guildies? If they do 100 forges a day on average while other members of you guild do 5-10, then they will get 10-20 times more precursors in the same amount of time (on average). Many people in my guild have had precursor drops while I never had one. However, many of those play at least twice as much a day, which does matter.

My last try was 200 crafted greatswords, but thats only 50 roll in a row, i got 4-5 low cost exos

Olvendred.3027:

Lazaar.9123:

Why has there been so many conspiracy theorists active lately?

It’s well known that instead of coding the one loot table for everyone, games developers prefer to spend vast amounts of time making fiendishly complicated systems so that some of their players get better stuff than others. Accounts of this are always supported by large quantities of well-documented evidence, which takes into account all other factors that might be influential, so there’s really very little doubt about it.

On the other hand it’s pretty obvious that if someone gets 11 pre-cursors, and others who play everyday since release and got nothing. That there might be something wrong with the rng system.

aspirine.6852:

Olvendred.3027:

Lazaar.9123:

Why has there been so many conspiracy theorists active lately?

It’s well known that instead of coding the one loot table for everyone, games developers prefer to spend vast amounts of time making fiendishly complicated systems so that some of their players get better stuff than others. Accounts of this are always supported by large quantities of well-documented evidence, which takes into account all other factors that might be influential, so there’s really very little doubt about it.

On the other hand it’s pretty obvious that if someone gets 11 pre-cursors, and others who play everyday since release and got nothing. That there might be something wrong with the rng system.

I thought I just said that. The truth is, there really is no RNG system. Colin Johanson just sits at a desk and individually chooses who gets each precursor and when. Some people he likes, others he’s not so fond of.

I also heard that Colin is actually Santa Claus, but I haven’t found any evidence of that yet.

Mouse.7382:

I’m new to GW2 and am light a moth to a flame, attracted to ‘pretty’ cosmetics—Have used the ‘mystic toliet’ on all my rares/exotics….NADA.

Stopped doing it..i just sell/salvage stuff now—at least I get something in return.

Yep, salvage and selling stuff much better, especially if you do the world boss rotation,
and sell/salvage every yellow.

Olvendred.3027:

I also heard that Colin is actually Santa Claus, but I haven’t found any evidence of that yet.

Is it true he also likes to swing swords?

Olvendred.3027:

aspirine.6852:

Olvendred.3027:

Lazaar.9123:

Why has there been so many conspiracy theorists active lately?

It’s well known that instead of coding the one loot table for everyone, games developers prefer to spend vast amounts of time making fiendishly complicated systems so that some of their players get better stuff than others. Accounts of this are always supported by large quantities of well-documented evidence, which takes into account all other factors that might be influential, so there’s really very little doubt about it.

On the other hand it’s pretty obvious that if someone gets 11 pre-cursors, and others who play everyday since release and got nothing. That there might be something wrong with the rng system.

I thought I just said that. The truth is, there really is no RNG system. Colin Johanson just sits at a desk and individually chooses who gets each precursor and when. Some people he likes, others he’s not so fond of.

I also heard that Colin is actually Santa Claus, but I haven’t found any evidence of that yet.

Yeah, you quite fail at trying to be funny.

Lazaar.9123:

Olvendred.3027:

I also heard that Colin is actually Santa Claus, but I haven’t found any evidence of that yet.

Is it true he also likes to swing swords?

He doesn’t always swing swords. But when he does, he swings them again.

sigh

You guys are a perfect example of how people perceive “randomness”. You think randomness means everything is distributed equally. It’s doesn’t, so stop thinking like that.

True randomness means there needs to be some extremely “unlucky” streaks, and some extremely “lucky” streaks. If there weren’t, it would not be truly random.

I’ll give you a good exercise that might or might not open your eyes.

Write down a series of 0s and 1s in what you think is a truly random sequence. Make it long enough, say for example 100 numbers.

Then, start up excel and enter this formula:

=ROUND ( RAND ( ) ; 0)

And copy it through 100 cells.

Let me know your observations when you compare your own “random” sequence and the computer generated random sequence.

Finally, I would like to say true randomness in an MMO probably isn’t the best way of handling progression and loot. A good system would be random, but would also include a so-called “streak-killer” algorithm, that would cut (at least) bad streaks eventually, guaranteeing a “hit” after enough “misses”.

Loki.8793:

Lazaar.9123:

Olvendred.3027:

I also heard that Colin is actually Santa Claus, but I haven’t found any evidence of that yet.

Is it true he also likes to swing swords?

He doesn’t always swing swords. But when he does, he swings them again.

…and again.

Tom Gore.4035:

sigh

You guys are a perfect example of how people perceive “randomness”. You think randomness means everything is distributed equally. It’s doesn’t, so stop thinking like that.

True randomness means there needs to be some extremely “unlucky” streaks, and some extremely “lucky” streaks. If there weren’t, it would not be truly random.

I’ll give you a good exercise that might or might not open your eyes.

Write down a series of 0s and 1s in what you think is a truly random sequence. Make it long enough, say for example 100 numbers.

Then, start up excel and enter this formula:

=ROUND ( RAND ( ) ; 0)

And copy it through 100 cells.

Let me know your observations when you compare your own “random” sequence and the computer generated random sequence.

Finally, I would like to say true randomness in an MMO probably isn’t the best way of handling progression and loot. A good system would be random, but would also include a so-called “streak-killer” algorithm, that would cut (at least) bad streaks eventually, guaranteeing a “hit” after enough “misses”.

We / I know what RNG means, but if always suck for me, and always not for him, thats not truly RNG, RNG affects him too, then why RNG is his side always?
;)

After a year and a half of playing, i got “Howl” from the Silverwastes lastnight. Thefirst Pre was from putting swords in the MF and getting Spark. Pres do drop, but you just have to not make it the focal point of your gamin experience. On the other hand, i am dreading that T6 mats farm. >_<

Fallen Healer Hun.9638:

Tom Gore.4035:

sigh

You guys are a perfect example of how people perceive “randomness”. You think randomness means everything is distributed equally. It’s doesn’t, so stop thinking like that.

True randomness means there needs to be some extremely “unlucky” streaks, and some extremely “lucky” streaks. If there weren’t, it would not be truly random.

I’ll give you a good exercise that might or might not open your eyes.

Write down a series of 0s and 1s in what you think is a truly random sequence. Make it long enough, say for example 100 numbers.

Then, start up excel and enter this formula:

=ROUND ( RAND ( ) ; 0)

And copy it through 100 cells.

Let me know your observations when you compare your own “random” sequence and the computer generated random sequence.

Finally, I would like to say true randomness in an MMO probably isn’t the best way of handling progression and loot. A good system would be random, but would also include a so-called “streak-killer” algorithm, that would cut (at least) bad streaks eventually, guaranteeing a “hit” after enough “misses”.

We / I know what RNG means, but if always suck for me, and always not for him, thats not truly RNG, RNG affects him too, then why RNG is his side always?
;)

If you flip a coin, it has a 50/50 chance of landing on heads or tails. If you flip a coin 7 times, and it lands on heads every time, what’s the chance it will land on tails on the 8th flip? It will still be 50%, because each flip is independent and separate from each other.

The same logic applies to loot drops. Every time you throw in some rares/exotics into the forge, you have a set chance of getting a precursor back. If you try for a precurser 100 times and fail each and every time, your 101th try won’t have a better or worse chance because you failed 100 times before. Each and every attempt is separate.

People like to try and draw patterns when there isn’t one. And this is a case where there isn’t a pattern, no matter how much you insist there is.

Fallen Healer Hun.9638:

Tom Gore.4035:

sigh

You guys are a perfect example of how people perceive “randomness”. You think randomness means everything is distributed equally. It’s doesn’t, so stop thinking like that.

True randomness means there needs to be some extremely “unlucky” streaks, and some extremely “lucky” streaks. If there weren’t, it would not be truly random.

I’ll give you a good exercise that might or might not open your eyes.

Write down a series of 0s and 1s in what you think is a truly random sequence. Make it long enough, say for example 100 numbers.

Then, start up excel and enter this formula:

=ROUND ( RAND ( ) ; 0)

And copy it through 100 cells.

Let me know your observations when you compare your own “random” sequence and the computer generated random sequence.

Finally, I would like to say true randomness in an MMO probably isn’t the best way of handling progression and loot. A good system would be random, but would also include a so-called “streak-killer” algorithm, that would cut (at least) bad streaks eventually, guaranteeing a “hit” after enough “misses”.

We / I know what RNG means, but if always suck for me, and always not for him, thats not truly RNG, RNG affects him too, then why RNG is his side always?
;)

You still don’t understand. In order for random to be random, some of the players in a particular MMO’s population must have “rotten luck” and some of them must have “amazing luck”. Otherwise it’s not random.

Sucks to be the one getting the short stick, but hey you can at least take some pride in that you’re keeping the Cups of the Random Scale balanced.

aspirine.6852:

Olvendred.3027:

aspirine.6852:

Olvendred.3027:

Lazaar.9123:

Why has there been so many conspiracy theorists active lately?

It’s well known that instead of coding the one loot table for everyone, games developers prefer to spend vast amounts of time making fiendishly complicated systems so that some of their players get better stuff than others. Accounts of this are always supported by large quantities of well-documented evidence, which takes into account all other factors that might be influential, so there’s really very little doubt about it.

On the other hand it’s pretty obvious that if someone gets 11 pre-cursors, and others who play everyday since release and got nothing. That there might be something wrong with the rng system.

I thought I just said that. The truth is, there really is no RNG system. Colin Johanson just sits at a desk and individually chooses who gets each precursor and when. Some people he likes, others he’s not so fond of.

I also heard that Colin is actually Santa Claus, but I haven’t found any evidence of that yet.

Yeah, you quite fail at trying to be funny.

I think they brought a lot of clarity to this thread. I always wondered who Santa was now I know. Thanks.
P.S. I wanna pony!

Fallen Healer Hun.9638:

Tom Gore.4035:

sigh

You guys are a perfect example of how people perceive “randomness”. You think randomness means everything is distributed equally. It’s doesn’t, so stop thinking like that.

True randomness means there needs to be some extremely “unlucky” streaks, and some extremely “lucky” streaks. If there weren’t, it would not be truly random.

I’ll give you a good exercise that might or might not open your eyes.

Write down a series of 0s and 1s in what you think is a truly random sequence. Make it long enough, say for example 100 numbers.

Then, start up excel and enter this formula:

=ROUND ( RAND ( ) ; 0)

And copy it through 100 cells.

Let me know your observations when you compare your own “random” sequence and the computer generated random sequence.

Finally, I would like to say true randomness in an MMO probably isn’t the best way of handling progression and loot. A good system would be random, but would also include a so-called “streak-killer” algorithm, that would cut (at least) bad streaks eventually, guaranteeing a “hit” after enough “misses”.

We / I know what RNG means, but if always suck for me, and always not for him, thats not truly RNG, RNG affects him too, then why RNG is his side always?
;)

You might know what it means but you don’t understand how it works.

You put every player’s “luck” (i.e rare, exotic, precursor etc, etc drop rates considering forge attempts, monsters killed, account MF etc) on a graph and you will get something similar to a bell-curve. A bunch of people with an average drop rate, some people with absolutely horrible drop-rates and some jerks that get all the loots.

With precursors, though, it’d just be heavily skewed to the right, i.e a LOT of people with 0, some with 1, 2, 3…. a very, very select few (that naturally you always hear about) with 5+.

Simple change to mystic toilet: make a craftable item for each precursor called “Prayer for —--”. When you place this plus 3 lvl 80+exotics of the same weapon type in the mystic whirlpool it returns an item plus an extra Prayer For token. The more tokens you have the better your odds of a Precursor, with a hard limit # of tokens for when you get one. So, it may be more expensive than market value in the end, but at least you won’t burn literally thousands of gold for nothing, as some people have.

Tom Gore.4035:

sigh

You guys are a perfect example of how people perceive “randomness”. You think randomness means everything is distributed equally. It’s doesn’t, so stop thinking like that.

True randomness means there needs to be some extremely “unlucky” streaks, and some extremely “lucky” streaks. If there weren’t, it would not be truly random.

I’ll give you a good exercise that might or might not open your eyes.

Write down a series of 0s and 1s in what you think is a truly random sequence. Make it long enough, say for example 100 numbers.

Then, start up excel and enter this formula:

=ROUND ( RAND ( ) ; 0)

And copy it through 100 cells.

Let me know your observations when you compare your own “random” sequence and the computer generated random sequence.

Finally, I would like to say true randomness in an MMO probably isn’t the best way of handling progression and loot. A good system would be random, but would also include a so-called “streak-killer” algorithm, that would cut (at least) bad streaks eventually, guaranteeing a “hit” after enough “misses”.

Well if the chances of each possibility is equal, then you should have an equal distribution in a large sample.

The issue is that every participant in the lottery is aggregated so who are the winning and losing outliers will cancel each other out.

A truly random RNG wont have frequent long streaks like rolling 10 6s consecutively, 10 5s, 10 4s, etc in 60 rolls as an extreme example.

Khal Drogo.9631:

Tom Gore.4035:

sigh

You guys are a perfect example of how people perceive “randomness”. You think randomness means everything is distributed equally. It’s doesn’t, so stop thinking like that.

True randomness means there needs to be some extremely “unlucky” streaks, and some extremely “lucky” streaks. If there weren’t, it would not be truly random.

I’ll give you a good exercise that might or might not open your eyes.

Write down a series of 0s and 1s in what you think is a truly random sequence. Make it long enough, say for example 100 numbers.

Then, start up excel and enter this formula:

=ROUND ( RAND ( ) ; 0)

And copy it through 100 cells.

Let me know your observations when you compare your own “random” sequence and the computer generated random sequence.

Finally, I would like to say true randomness in an MMO probably isn’t the best way of handling progression and loot. A good system would be random, but would also include a so-called “streak-killer” algorithm, that would cut (at least) bad streaks eventually, guaranteeing a “hit” after enough “misses”.

Well if the chances of each possibility is equal, then you should have an equal distribution in a large sample.

The issue is that every participant in the lottery is aggregated so who are the winning and losing outliers will cancel each other out.

A truly random RNG wont have frequent long streaks like rolling 10 6s consecutively, 10 5s, 10 4s, etc in 60 rolls as an extreme example.

No, but it would have for example a streak of 5 6s quite frequently, and a streak of 10 6s would still not be a super rare anomaly either.

The fact that we have people that have had absolutely no drops during their 1000 hours of playing, and people who have had 10 drops during the same time tells us that Anets algorithm is truly random and that they do not use any streak-breaking algorithms (but maybe they should).

How do you mystic toilet? Im feelin lucky

This doesn’t mean “lucky accounts are a thing.” This means your guildie is representative of an outlier. Nothing more.

Tom Gore.4035:

No, but it would have for example a streak of 5 6s quite frequently, and a streak of 10 6s would still not be a super rare anomaly either.

The fact that we have people that have had absolutely no drops during their 1000 hours of playing, and people who have had 10 drops during the same time tells us that Anets algorithm is truly random and that they do not use any streak-breaking algorithms (but maybe they should).

If a streak of 5 6s occur frequently, wouldnt you say the algorithm is faulty? The question is what is the threshold though.

Elegy.2159:

This doesn’t mean “lucky accounts are a thing.” This means your guildie is representative of an outlier. Nothing more.

You know what the TS should do? Get all the guildies to give all their yellows or pay the lucky guildie to do the dice rolling for them.

But since it seems you don’t understand randomness or standard distribution, let’s try to reason with you, then.

Could you tell me, from a business perspective, why would Anet EVER want to assign some accounts with extremely bad luck and some accounts with extremely good luck? Tell me ONE way that would make any sense.

Because that would be a surefire way to lose BOTH players. One would quit because he already got everything in the game quickly and has nothing to look forward to. Other would quit because he didn’t get anything and felt betrayed.

If anything, PREVENTING that from happening would be the only smart business decision, but as it happens, we know that they are not fudging with the RNG. It’s just RNG at its best (and worst).

Khal Drogo.9631:

Tom Gore.4035:

No, but it would have for example a streak of 5 6s quite frequently, and a streak of 10 6s would still not be a super rare anomaly either.

The fact that we have people that have had absolutely no drops during their 1000 hours of playing, and people who have had 10 drops during the same time tells us that Anets algorithm is truly random and that they do not use any streak-breaking algorithms (but maybe they should).

If a streak of 5 6s occur frequently, wouldnt you say the algorithm is faulty? The question is what is the threshold though.

We were talking about MMOs, so “frequently” could be as few as one in a thousand and it would still stick out like a sore thumb, like this thread is.

Loki.8793:

Fallen Healer Hun.9638:

Tom Gore.4035:

sigh

You guys are a perfect example of how people perceive “randomness”. You think randomness means everything is distributed equally. It’s doesn’t, so stop thinking like that.

True randomness means there needs to be some extremely “unlucky” streaks, and some extremely “lucky” streaks. If there weren’t, it would not be truly random.

I’ll give you a good exercise that might or might not open your eyes.

Write down a series of 0s and 1s in what you think is a truly random sequence. Make it long enough, say for example 100 numbers.

Then, start up excel and enter this formula:

=ROUND ( RAND ( ) ; 0)

And copy it through 100 cells.

Let me know your observations when you compare your own “random” sequence and the computer generated random sequence.

Finally, I would like to say true randomness in an MMO probably isn’t the best way of handling progression and loot. A good system would be random, but would also include a so-called “streak-killer” algorithm, that would cut (at least) bad streaks eventually, guaranteeing a “hit” after enough “misses”.

We / I know what RNG means, but if always suck for me, and always not for him, thats not truly RNG, RNG affects him too, then why RNG is his side always?
;)

If you flip a coin, it has a 50/50 chance of landing on heads or tails. If you flip a coin 7 times, and it lands on heads every time, what’s the chance it will land on tails on the 8th flip? It will still be 50%, because each flip is independent and separate from each other.

The same logic applies to loot drops. Every time you throw in some rares/exotics into the forge, you have a set chance of getting a precursor back. If you try for a precurser 100 times and fail each and every time, your 101th try won’t have a better or worse chance because you failed 100 times before. Each and every attempt is separate.

People like to try and draw patterns when there isn’t one. And this is a case where there isn’t a pattern, no matter how much you insist there is.

No, I doesn’t insist, as i wrote earlier, I know each roll is separate, and doesn’t affected by the previous outcome, this is Luck

Tom!

As i wrote earlier:

“I don’t think there is a numeric value in the account database, where the prec drop chance is stored and it’s different for each account.”

I know exactly the meaning of equal distribution, sorry if i was misunderstandable

Tom Gore.4035:

Khal Drogo.9631:

Tom Gore.4035:

No, but it would have for example a streak of 5 6s quite frequently, and a streak of 10 6s would still not be a super rare anomaly either.

The fact that we have people that have had absolutely no drops during their 1000 hours of playing, and people who have had 10 drops during the same time tells us that Anets algorithm is truly random and that they do not use any streak-breaking algorithms (but maybe they should).

If a streak of 5 6s occur frequently, wouldnt you say the algorithm is faulty? The question is what is the threshold though.

We were talking about MMOs, so “frequently” could be as few as one in a thousand and it would still stick out like a sore thumb, like this thread is.

Well it is a good question to ask statisticians: If streaks of 5 6s are frequent, whereas 1,2,3,4,5 do not have or have infrequent streaks in an equal distribution in a large sample size, would the RNG or dice roll be considered truly random?

Since we are on the subject, why do some single player non online games with RNG have the same results immediately after loading / save skitten, and different when some additional action or turn is taken?

Fallen Healer Hun.9638:

My Guildmate had his 10th or 11th precursor this weekend, simply logged in for an our, dropped his yellows into the forge. he doesnt even remember it’s name, because in the last half year he got one every month, so precursor is a pretty common thing to him.

I had one, Venom, almost 2 years ago

In our guild, with 30 ppl, there are not one other, with this luck.

Anyone has similar experiences?

(Yes, I’m jealous)

Or they are just linking them in guild chat to troll you.

Fallen Healer Hun.9638:

Galespark.7835:

Maybe they also do a lot more forges than your other guildies? If they do 100 forges a day on average while other members of you guild do 5-10, then they will get 10-20 times more precursors in the same amount of time (on average). Many people in my guild have had precursor drops while I never had one. However, many of those play at least twice as much a day, which does matter.

My last try was 200 crafted greatswords, but thats only 50 roll in a row, i got 4-5 low cost exos

Yes, and do you do that once every week, 2 weeks, month, year? And how often do those “lucky people” do this? My point still stands. If this year you MF 1000 rares, and they MF 10000, they are very likely to get more precursors than you do.

And then there is the principle that there will be some people getting very lucky and some very unlucky with a system like this, as Tom Gore has explained. In short, these anecdotes carry too little weight to be used as evidence for saying that the RNG is actually not working well.

Of course, that does not mean you have to like the current system, but there are existing threads where you can give your feedback, and the precursor problem will be adressed in some way in the expansion, though we do not know many details yet.

Khal Drogo.9631:

Tom Gore.4035:

Khal Drogo.9631:

Tom Gore.4035:

No, but it would have for example a streak of 5 6s quite frequently, and a streak of 10 6s would still not be a super rare anomaly either.

The fact that we have people that have had absolutely no drops during their 1000 hours of playing, and people who have had 10 drops during the same time tells us that Anets algorithm is truly random and that they do not use any streak-breaking algorithms (but maybe they should).

If a streak of 5 6s occur frequently, wouldnt you say the algorithm is faulty? The question is what is the threshold though.

We were talking about MMOs, so “frequently” could be as few as one in a thousand and it would still stick out like a sore thumb, like this thread is.

Well it is a good question to ask statisticians: If streaks of 5 6s are frequent, whereas 1,2,3,4,5 do not have or have infrequent streaks in an equal distribution in a large sample size, would the RNG or dice roll be considered truly random?

I took the 6s as an example. Could use any one number instead and my point would still remain.

The point is, in a large sample size, streaks happen, and they happen a lot more frequently than people generally think they do.

That was the purpose of my exercise I posted, about listing a “random” sequence of 1s and 0s and comparing it to one generated by a random number generator. This is a classic exercise by math teachers to show how people misinterpret randomness.

In most cases, people make their own “random” sequences with very little or no streaks of 5 or 6, which are almost guaranteed to happen in a truly random sequence of say, 100 numbers.

Khal Drogo.9631:

Tom Gore.4035:

Khal Drogo.9631:

Tom Gore.4035:

No, but it would have for example a streak of 5 6s quite frequently, and a streak of 10 6s would still not be a super rare anomaly either.

The fact that we have people that have had absolutely no drops during their 1000 hours of playing, and people who have had 10 drops during the same time tells us that Anets algorithm is truly random and that they do not use any streak-breaking algorithms (but maybe they should).

If a streak of 5 6s occur frequently, wouldnt you say the algorithm is faulty? The question is what is the threshold though.

We were talking about MMOs, so “frequently” could be as few as one in a thousand and it would still stick out like a sore thumb, like this thread is.

Well it is a good question to ask statisticians: If streaks of 5 6s are frequent, whereas 1,2,3,4,5 do not have or have infrequent streaks in an equal distribution in a large sample size, would the RNG or dice roll be considered truly random?

Since we are on the subject, why do some single player non online games with RNG have the same results immediately after loading / save skitten, and different when some additional action or turn is taken?

Some games have simple forms of RNG that can cause this to happen.

I have emulated some console games on computer with programs that had savestates, which save just about every variable (way more than normal saving does) and allow you to save and load anywhere in the game as if you had never stopped. In those cases most PS1 games I tried always gave the same results for the same actions done after loading.

However, I highly doubt any triple A MMO will use such simple systems, and there are no saves or savestates to be abused. If GW2 had such simple systems people would probably have found ways to exploit them by now. I do not recall any single player game released in the last 5 years that showed this behaviour using normal saves either, but granted I don’t play a lot of new games lately.

Edit: edited for citation failure

Tom Gore.4035:

I took the 6s as an example. Could use any one number instead and my point would still remain.

The point is, in a large sample size, streaks happen, and they happen a lot more frequently than people generally think they do.

That was the purpose of my exercise I posted, about listing a “random” sequence of 1s and 0s and comparing it to one generated by a random number generator. This is a classic exercise by math teachers to show how people misinterpret randomness.

In most cases, people make their own “random” sequences with very little or no streaks of 5 or 6, which are almost guaranteed to happen in a truly random sequence of say, 100 numbers.

In a large sample size, streaks do happen but infrequently if your measure is one streak in a few thousand numbers. If it happens say once in every 100 numbers, wouldnt you say the RNG is not truly random?

Galespark.7835:

Khal Drogo.9631:

Tom Gore.4035:

Khal Drogo.9631:

Tom Gore.4035:

No, but it would have for example a streak of 5 6s quite frequently, and a streak of 10 6s would still not be a super rare anomaly either.

The fact that we have people that have had absolutely no drops during their 1000 hours of playing, and people who have had 10 drops during the same time tells us that Anets algorithm is truly random and that they do not use any streak-breaking algorithms (but maybe they should).

If a streak of 5 6s occur frequently, wouldnt you say the algorithm is faulty? The question is what is the threshold though.

We were talking about MMOs, so “frequently” could be as few as one in a thousand and it would still stick out like a sore thumb, like this thread is.

Well it is a good question to ask statisticians: If streaks of 5 6s are frequent, whereas 1,2,3,4,5 do not have or have infrequent streaks in an equal distribution in a large sample size, would the RNG or dice roll be considered truly random?

Since we are on the subject, why do some single player non online games with RNG have the same results immediately after loading / save skitten, and different when some additional action or turn is taken?

Some games have simple forms of RNG that can cause this to happen.

I have emulated some console games on computer with programs that had savestates, which save just about every variable (way more than normal saving does) and allow you to save and load anywhere in the game as if you had never stopped. In those cases most PS1 games I tried always gave the same results for the same actions done after loading.

However, I highly doubt any triple A MMO will use such simple systems, and there are saves or savestates to be abused. If GW2 had such simple systems people would probably have found ways to exploit them by now. I do not recall any single player game released in the last 5 years that showed this behaviour using normal saves either, but granted I don’t play a lot of new games lately.

Edit: edited for citation failure

Heh this occurs in Xcom: Enemy Unknown a recent AAA tactical game as well. Not sure if they patched this or not. Of course this would not happen in a MMO.

Tom Gore.4035:

sigh
Finally, I would like to say true randomness in an MMO probably isn’t the best way of handling progression and loot. A good system would be random, but would also include a so-called “streak-killer” algorithm, that would cut (at least) bad streaks eventually, guaranteeing a “hit” after enough “misses”.

Problem with “streak-killer” is how to sort them. I mean basically each time you use a skill you roll a dice to know if the hit crits or not, which means you would have to record a lot of data for each account to kill bad luck streaks.
In the same way you could just cater the record on mystic forging but then each time you upgrade a material it should be counted. If one system is created for precursor only then everyone will ask for the same system for chests on vinewrath, then for ecto salvage on gold, for bags during festival and so on….

RNG is so much involved everywhere in the game that killing bad luck streaks for everyone in a uniform manner would require an analysis of so many cases that the work and data handling would be overwhelming. On top of that it would actually create a “big brother” effect since now it would be true that each account would be watched to play with their luck….

get the stuff, give to your friend and ask him/her to pull a miracle from the mystic toillet.

On testing for streaks:

Whether the amount of streaks is too high would depend on many things: the possible results, the chances for those results individually and the size of the streaks come to mind. Reliably testing an RNG algorithm for such low odds as precursors already takes a lot of samples, I believe testing them for being prone to streaks of these would take many times that. Especially if you don’t really test for streaks, but rather for situations (read: individuals) that cause them to happen more/less often than they should. I don’t believe anyone has a high enough number of samples to check this, besides (probably) arenanet.

Khal Drogo.9631:

Tom Gore.4035:

I took the 6s as an example. Could use any one number instead and my point would still remain.

The point is, in a large sample size, streaks happen, and they happen a lot more frequently than people generally think they do.

That was the purpose of my exercise I posted, about listing a “random” sequence of 1s and 0s and comparing it to one generated by a random number generator. This is a classic exercise by math teachers to show how people misinterpret randomness.

In most cases, people make their own “random” sequences with very little or no streaks of 5 or 6, which are almost guaranteed to happen in a truly random sequence of say, 100 numbers.

In a large sample size, streaks do happen but infrequently if your measure is one streak in a few thousand numbers. If it happens say once in every 100 numbers, wouldnt you say the RNG is not truly random?

I’m talking about a streak of 5 or 6 1’s or 0’s, in a sequence of 100 random 1’s or 0’s. That does happen frequently, almost every time. And it’s because the RNG generating those 1s and 0s is truly random. If you don’t believe me I don’t know what I can say to convince you. Maybe you should take a few courses in statistics?

Loki.8793:

If you flip a coin, it has a 50/50 chance of landing on heads or tails. If you flip a coin 7 times, and it lands on heads every time, what’s the chance it will land on tails on the 8th flip? It will still be 50%, because each flip is independent and separate from each other.

The same logic applies to loot drops. Every time you throw in some rares/exotics into the forge, you have a set chance of getting a precursor back. If you try for a precurser 100 times and fail each and every time, your 101th try won’t have a better or worse chance because you failed 100 times before. Each and every attempt is separate.

People like to try and draw patterns when there isn’t one. And this is a case where there isn’t a pattern, no matter how much you insist there is.

Sort of. There’s the one set of odds you’re talking about “what is the probability the 8th flip is heads”—that is 1/2, yes. But then there’s a different set of odds: “what is the probability you will get 8 heads in a row”. That is not 50%. That would be (1/2)^8 or ~.4% probability that heads will land 8 times in a row on a coin flip. While each flip can be considered independent of each other, probability is based on the larger set of data. It can happen that you get 8 heads in a row, but it is certainly not common.

That said, I do believe there is something fishy with accounts getting 10+ precursors and others getting none. Specifically because precursors have a specific % chance to drop when doing MF. It is far below 1%. If these guys have dumped thousands upon thousands of weapons in the toilet, statistically speaking, they would end up with multiple precursors. But some people get multiple precursors in 100-200 weapons into the MF. The probability of that happening is hard to swallow, let alone people that get 3, 4, or even 5 precursors within 100-200 tries. Still, it can happen, I suppose.

But I definitely see why people believe there is a conspiracy going on.

Side note: I had to buy Dusk. After 300 attempts in the MF (1200 Greatswords) and getting nothing worth above 2g, I felt my money was better spent simply saving up for it. Contrasted with one of my guildies who has now gotten his 2nd Dawn from MF and told me he’s probably done 50 attempts total. maybe. He buys 100 swords and then sees what he gets. Both sessions he’s done this he’s ended up with Dawn (although he wanted Dusk the 2nd time, so I guess the joke’s on him….). Stranger things have happened, but boy is it frustrating when I’m 2500g+ in the hole for a sword, and he’s got his sword plus a ton of profit. Best not to think about it.

^ That’s the bit a lot of people forget. As these “streaks” continue their probability chance becomes so infinitesimal that the probability that something is wrong becomes quite substantial in comparison. The probability of multiple events is different than the normal distribution curve.

So if the statistics of streaks are correct why then would the statistics of there being an error be incorrect? Food for thought

Galespark.7835:

Fallen Healer Hun.9638:

Galespark.7835:

Maybe they also do a lot more forges than your other guildies? If they do 100 forges a day on average while other members of you guild do 5-10, then they will get 10-20 times more precursors in the same amount of time (on average). Many people in my guild have had precursor drops while I never had one. However, many of those play at least twice as much a day, which does matter.

My last try was 200 crafted greatswords, but thats only 50 roll in a row, i got 4-5 low cost exos

Yes, and do you do that once every week, 2 weeks, month, year? And how often do those “lucky people” do this? My point still stands. If this year you MF 1000 rares, and they MF 10000, they are very likely to get more precursors than you do.

And then there is the principle that there will be some people getting very lucky and some very unlucky with a system like this, as Tom Gore has explained. In short, these anecdotes carry too little weight to be used as evidence for saying that the RNG is actually not working well.

Of course, that does not mean you have to like the current system, but there are existing threads where you can give your feedback, and the precursor problem will be adressed in some way in the expansion, though we do not know many details yet.

Neither me, nor him wont craft and put in into MF hundreds of rares,
he simply puts his dropped yellows from into MF.

I didn’t said i have problem with the current RNG system and I didn’t think it’s broken,
just wanted to point out the thing is happening and exists, as Tom said earlier:

“True randomness means there needs to be some extremely “unlucky” streaks, and some extremely “lucky” streaks. If there weren’t, it would not be truly random.”

and I think it would be great to make the RNG like he mentioned:

“Finally, I would like to say true randomness in an MMO probably isn’t the best way of handling progression and loot. A good system would be random, but would also include a so-called “streak-killer” algorithm, that would cut (at least) bad streaks eventually, guaranteeing a “hit” after enough “misses”.

1st precursor I purchased for 40g for Meteorlogicus
2nd precursor was storm as a drop, sold for ~300.
3rd precursor was storm as a drop 1 week later, sold for ~300.
4th precursor I purchased for I think ~200g for Moot
5th precursor was the trident by tossing in exotic AC pistols and stones into the mystic forge. Sold for 70g
6th precursor was from purchasing exotic swords with ALL my dungeon tokens (1000s upon thousands) starting from AC and going up to Arah. My last 4 swords I could purchase this way, I /bowed at the MF before accepting it, got Zap. Made Bolt
7th precursor I got as a drop was yesterday, another storm. Put it up for ~600g.

I play lots, but my time is pretty optimized too. I dungeon train most days since launch.