multi refine is trash

XxXOMNIXxX

i get an average of 1 gold mat per 100 refined materials. normally on average sometimes more sometimes less. but thats my average. now i get like 1 in 1000 refined materials @zos. this *** is unnaceptable. give me my old wayt of refining 1 at a time

VaranisArano

How many total mats did you refine?

Without context and more detailed data, we have no way of knowing if you found a statiscally significant deviation from what we should expect (1 in 200 for gold mats) or whether you just have a small sample and bad luck.

simox

Are we gonna get a new thread about this everytime some has bad RNG...?

The rate has always been 1 gold material per 200 raw mats on average.

The drop rate has never changed and it has not changed this time around either.

Deep_01

No, its not. I got 4 alloys from 150 rubedite and zero wax from 400 ancestor silk.

yodased

Deep_01 wrote: »

No, its not. I got 4 alloys from 150 rubedite and zero wax from 400 ancestor silk.

You are not guaranteed 1 for every 200 you open. Thats not how probability works.

XxXOMNIXxX

WHEN I DID My wekkly writ roundup i got 5 gold mats oput of like 3.5k supplies. thats horrible. actually then another 3 on 2k

redlink1979

simox wrote: »

Are we gonna get a new thread about this everytime some has bad RNG...?

The rate has always been 1 gold material per 200 raw mats on average.

The drop rate has never changed and it has not changed this time around either.

Exactly.

Taleof2Cities

redlink1979 wrote: »

simox wrote: »

Are we gonna get a new thread about this everytime some has bad RNG...?

The rate has always been 1 gold material per 200 raw mats on average.

The drop rate has never changed and it has not changed this time around either.

Exactly.

Which means we don’t need a new thread about this everytime some has bad RNG ... as was pointed out above.

Lylith

XxXOMNIXxX wrote: »

i get an average of 1 gold mat per 100 refined materials. normally on average sometimes more sometimes less. but thats my average. now i get like 1 in 1000 refined materials @zos. this *** is unnaceptable. give me my old wayt of refining 1 at a time

i agree completely; it's absolute ***.

if you fanboys are getting better results, more power to you.

some of us aren't, doesn't make a *** what you say about it.

VaranisArano

Lylith wrote: »

XxXOMNIXxX wrote: »

i get an average of 1 gold mat per 100 refined materials. normally on average sometimes more sometimes less. but thats my average. now i get like 1 in 1000 refined materials @zos. this *** is unnaceptable. give me my old wayt of refining 1 at a time

i agree completely; it's absolute ***.

if you fanboys are getting better results, more power to you.

some of us aren't, doesn't make a *** what you say about it.

Here's the deal.

If you aren't getting those 1 in 200 results when you refine in statistically significant amounts, prove it.

And no, "I refined a couple hundred or a thousand and got sucky RNG" doesnt count. That's not statistically significant. You'd have to repeat that a bunch of times in order to reach a significant level of tests.

For context, the player(s) who tested it on the PTS and on Live refined over a hundred thousand mats and found that the drop rate hasnt changed. If you want to legitimately disprove that, you are going to have to actually refine either in large amounts or refine smaller amounts many, many times.

If you don't have the proof, its pretty easy to chalk your complaints up to Confirmation Bias, Small sample sizes, and the Margin of Error.

Lylith

VaranisArano wrote: »

Lylith wrote: »

XxXOMNIXxX wrote: »

i get an average of 1 gold mat per 100 refined materials. normally on average sometimes more sometimes less. but thats my average. now i get like 1 in 1000 refined materials @zos. this *** is unnaceptable. give me my old wayt of refining 1 at a time

i agree completely; it's absolute ***.

if you fanboys are getting better results, more power to you.

some of us aren't, doesn't make a *** what you say about it.

Here's the deal.

If you aren't getting those 1 in 200 results when you refine in statistically significant amounts, prove it.

And no, "I refined a couple hundred or a thousand and got sucky RNG" doesnt count. That's not statistically significant. You'd have to repeat that a bunch of times in order to reach a significant level of tests.

For context, the player(s) who tested it on the PTS and on Live refined over a hundred thousand mats and found that the drop rate hasnt changed. If you want to legitimately disprove that, you are going to have to actually refine either in large amounts or refine smaller amounts many, many times.

If you don't have the proof, its pretty easy to chalk your complaints up to Confirmation Bias, Small sample sizes, and the Margin of Error.

call it whatever you like.

i'm not obliged to prove a *** thing to you.

have a nice day.

tmbrinks

Lylith wrote: »

VaranisArano wrote: »

Lylith wrote: »

XxXOMNIXxX wrote: »

i get an average of 1 gold mat per 100 refined materials. normally on average sometimes more sometimes less. but thats my average. now i get like 1 in 1000 refined materials @zos. this *** is unnaceptable. give me my old wayt of refining 1 at a time

i agree completely; it's absolute ***.

if you fanboys are getting better results, more power to you.

some of us aren't, doesn't make a *** what you say about it.

Here's the deal.

If you aren't getting those 1 in 200 results when you refine in statistically significant amounts, prove it.

And no, "I refined a couple hundred or a thousand and got sucky RNG" doesnt count. That's not statistically significant. You'd have to repeat that a bunch of times in order to reach a significant level of tests.

For context, the player(s) who tested it on the PTS and on Live refined over a hundred thousand mats and found that the drop rate hasnt changed. If you want to legitimately disprove that, you are going to have to actually refine either in large amounts or refine smaller amounts many, many times.

If you don't have the proof, its pretty easy to chalk your complaints up to Confirmation Bias, Small sample sizes, and the Margin of Error.

call it whatever you like.

i'm not obliged to prove a *** thing to you.

have a nice day.

then nobody is obliged to listen to your "evidence"

Taleof2Cities

tmbrinks wrote: »

Lylith wrote: »

VaranisArano wrote: »

Lylith wrote: »

XxXOMNIXxX wrote: »

i get an average of 1 gold mat per 100 refined materials. normally on average sometimes more sometimes less. but thats my average. now i get like 1 in 1000 refined materials @zos. this *** is unnaceptable. give me my old wayt of refining 1 at a time

i agree completely; it's absolute ***.

if you fanboys are getting better results, more power to you.

some of us aren't, doesn't make a *** what you say about it.

Here's the deal.

If you aren't getting those 1 in 200 results when you refine in statistically significant amounts, prove it.

And no, "I refined a couple hundred or a thousand and got sucky RNG" doesnt count. That's not statistically significant. You'd have to repeat that a bunch of times in order to reach a significant level of tests.

For context, the player(s) who tested it on the PTS and on Live refined over a hundred thousand mats and found that the drop rate hasnt changed. If you want to legitimately disprove that, you are going to have to actually refine either in large amounts or refine smaller amounts many, many times.

If you don't have the proof, its pretty easy to chalk your complaints up to Confirmation Bias, Small sample sizes, and the Margin of Error.

call it whatever you like.

i'm not obliged to prove a *** thing to you.

have a nice day.

then nobody is obliged to listen to your "evidence"

^This.^

have a nice day.

VaranisArano

Lylith wrote: »

VaranisArano wrote: »

Lylith wrote: »

XxXOMNIXxX wrote: »

i get an average of 1 gold mat per 100 refined materials. normally on average sometimes more sometimes less. but thats my average. now i get like 1 in 1000 refined materials @zos. this *** is unnaceptable. give me my old wayt of refining 1 at a time

i agree completely; it's absolute ***.

if you fanboys are getting better results, more power to you.

some of us aren't, doesn't make a *** what you say about it.

Here's the deal.

If you aren't getting those 1 in 200 results when you refine in statistically significant amounts, prove it.

And no, "I refined a couple hundred or a thousand and got sucky RNG" doesnt count. That's not statistically significant. You'd have to repeat that a bunch of times in order to reach a significant level of tests.

For context, the player(s) who tested it on the PTS and on Live refined over a hundred thousand mats and found that the drop rate hasnt changed. If you want to legitimately disprove that, you are going to have to actually refine either in large amounts or refine smaller amounts many, many times.

If you don't have the proof, its pretty easy to chalk your complaints up to Confirmation Bias, Small sample sizes, and the Margin of Error.

call it whatever you like.

i'm not obliged to prove a *** thing to you.

have a nice day.

Okay. I mean, I'm not your science teacher, so I can't make you prove your hypothesis.

But as the others above said, without proof, why should ZOS believe you? Why should the folks who refined a ton of mats to check the drop rate believe you?

Its a pretty basic "Anecdotes aren't data" situation.

Here's how I see it.

If you are right, and the new Multicraft is screwing up, and cheating you and others out of the generally accepted RNG for improvement mats, that's a big problem. It needs to be fixed.

But you can't just say "Its a problem!" With no backup and evidence to the contrary, ZOS won't listen to that. You have to prove it.

And your proof has to be in statistically significant amounts. Otherwise its very likely to be biased by confirmation bias, small sample sizes and vulnerable to the margin of error.

I realize that's not what you want to hear, but if there really is a problem, you are going to have to show that there's a problem. Especially when all the statistically significant proof we have shows that there isn't a problem.

So I'm not trying to call you a liar. Just trying to point out how you can prove you aren't mistaken, if you are so interested.

Personally, I have yet to see any statistically significant data sets that suggest there's a problem with improvement mat RNG. I'd love to see more people complaining about it actually record their data to find out.

Lylith

tmbrinks wrote: »

Lylith wrote: »

VaranisArano wrote: »

Lylith wrote: »

XxXOMNIXxX wrote: »

i get an average of 1 gold mat per 100 refined materials. normally on average sometimes more sometimes less. but thats my average. now i get like 1 in 1000 refined materials @zos. this *** is unnaceptable. give me my old wayt of refining 1 at a time

i agree completely; it's absolute ***.

if you fanboys are getting better results, more power to you.

some of us aren't, doesn't make a *** what you say about it.

Here's the deal.

If you aren't getting those 1 in 200 results when you refine in statistically significant amounts, prove it.

And no, "I refined a couple hundred or a thousand and got sucky RNG" doesnt count. That's not statistically significant. You'd have to repeat that a bunch of times in order to reach a significant level of tests.

For context, the player(s) who tested it on the PTS and on Live refined over a hundred thousand mats and found that the drop rate hasnt changed. If you want to legitimately disprove that, you are going to have to actually refine either in large amounts or refine smaller amounts many, many times.

If you don't have the proof, its pretty easy to chalk your complaints up to Confirmation Bias, Small sample sizes, and the Margin of Error.

call it whatever you like.

i'm not obliged to prove a *** thing to you.

have a nice day.

then nobody is obliged to listen to your "evidence"

then don't.

use the ignore function.

i didn't offer any evidence beyond my own experience, the refining rng is rubbish for me compared to what it used to be.

i don't the mats or the gold to do the research being suggested; i based my post on my experience, nothing more.

if you'd like to send me a few thousand mats, though, i'll be glad to see if i get a different result and get back to you on it.

tmbrinks

Lylith wrote: »

tmbrinks wrote: »

Lylith wrote: »

VaranisArano wrote: »

Lylith wrote: »

XxXOMNIXxX wrote: »

i get an average of 1 gold mat per 100 refined materials. normally on average sometimes more sometimes less. but thats my average. now i get like 1 in 1000 refined materials @zos. this *** is unnaceptable. give me my old wayt of refining 1 at a time

i agree completely; it's absolute ***.

if you fanboys are getting better results, more power to you.

some of us aren't, doesn't make a *** what you say about it.

Here's the deal.

If you aren't getting those 1 in 200 results when you refine in statistically significant amounts, prove it.

And no, "I refined a couple hundred or a thousand and got sucky RNG" doesnt count. That's not statistically significant. You'd have to repeat that a bunch of times in order to reach a significant level of tests.

For context, the player(s) who tested it on the PTS and on Live refined over a hundred thousand mats and found that the drop rate hasnt changed. If you want to legitimately disprove that, you are going to have to actually refine either in large amounts or refine smaller amounts many, many times.

If you don't have the proof, its pretty easy to chalk your complaints up to Confirmation Bias, Small sample sizes, and the Margin of Error.

call it whatever you like.

i'm not obliged to prove a *** thing to you.

have a nice day.

then nobody is obliged to listen to your "evidence"

then don't.

use the ignore function.

i didn't offer any evidence beyond my own experience, the refining rng is rubbish for me compared to what it used to be.

i don't the mats or the gold to do the research being suggested; i based my post on my experience, nothing more.

if you'd like to send me a few thousand mats, though, i'll be glad to see if i get a different result and get back to you on it.

You're the one who came in here cussing about it, yet offering nothing other than anecdotes. That combativeness tells me either one of three things.

1) You're trolling, just trying to rile up things.

2) You have absolutely no idea what the actual drops are/were, yet you "feel" they are lower, and are looking for something to blame (while having zero understanding of basic statistics)

3) You knew of an "exploit" that you were using and are upset they closed that loophole on you.

Lylith

VaranisArano wrote: »

Lylith wrote: »

VaranisArano wrote: »

Lylith wrote: »

XxXOMNIXxX wrote: »

i get an average of 1 gold mat per 100 refined materials. normally on average sometimes more sometimes less. but thats my average. now i get like 1 in 1000 refined materials @zos. this *** is unnaceptable. give me my old wayt of refining 1 at a time

i agree completely; it's absolute ***.

if you fanboys are getting better results, more power to you.

some of us aren't, doesn't make a *** what you say about it.

Here's the deal.

If you aren't getting those 1 in 200 results when you refine in statistically significant amounts, prove it.

And no, "I refined a couple hundred or a thousand and got sucky RNG" doesnt count. That's not statistically significant. You'd have to repeat that a bunch of times in order to reach a significant level of tests.

For context, the player(s) who tested it on the PTS and on Live refined over a hundred thousand mats and found that the drop rate hasnt changed. If you want to legitimately disprove that, you are going to have to actually refine either in large amounts or refine smaller amounts many, many times.

If you don't have the proof, its pretty easy to chalk your complaints up to Confirmation Bias, Small sample sizes, and the Margin of Error.

call it whatever you like.

i'm not obliged to prove a *** thing to you.

have a nice day.

Okay. I mean, I'm not your science teacher, so I can't make you prove your hypothesis.

But as the others above said, without proof, why should ZOS believe you? Why should the folks who refined a ton of mats to check the drop rate believe you?

Its a pretty basic "Anecdotes aren't data" situation.

Here's how I see it.

If you are right, and the new Multicraft is screwing up, and cheating you and others out of the generally accepted RNG for improvement mats, that's a big problem. It needs to be fixed.

But you can't just say "Its a problem!" With no backup and evidence to the contrary, ZOS won't listen to that. You have to prove it.

And your proof has to be in statistically significant amounts. Otherwise its very likely to be biased by confirmation bias, small sample sizes and vulnerable to the margin of error.

I realize that's not what you want to hear, but if there really is a problem, you are going to have to show that there's a problem. Especially when all the statistically significant proof we have shows that there isn't a problem.

So I'm not trying to call you a liar. Just trying to point out how you can prove you aren't mistaken, if you are so interested.

Personally, I have yet to see any statistically significant data sets that suggest there's a problem with improvement mat RNG. I'd love to see more people complaining about it actually record their data to find out.

that makes sense, tbh. as i stated earlier, i don't have sufficient materials to make an informed analysis, i just got really bad returns from a few thousand mats the first, and only, time i used the multicraft. (first 1000 - 1 gold, 2 purple, second 1000 4 gold, and the rest don't remember b/c i was getting truly pissed).

i'm really leery of it, now, and will probably just do single or small batch refining for a while.

maybe it's just bad luck or a bad day for me to refine mats. ¯\_(ツ)_/¯

thanks for the advice.

VaranisArano

Lylith wrote: »

VaranisArano wrote: »

Lylith wrote: »

VaranisArano wrote: »

Lylith wrote: »

XxXOMNIXxX wrote: »

i get an average of 1 gold mat per 100 refined materials. normally on average sometimes more sometimes less. but thats my average. now i get like 1 in 1000 refined materials @zos. this *** is unnaceptable. give me my old wayt of refining 1 at a time

i agree completely; it's absolute ***.

if you fanboys are getting better results, more power to you.

some of us aren't, doesn't make a *** what you say about it.

Here's the deal.

If you aren't getting those 1 in 200 results when you refine in statistically significant amounts, prove it.

And no, "I refined a couple hundred or a thousand and got sucky RNG" doesnt count. That's not statistically significant. You'd have to repeat that a bunch of times in order to reach a significant level of tests.

For context, the player(s) who tested it on the PTS and on Live refined over a hundred thousand mats and found that the drop rate hasnt changed. If you want to legitimately disprove that, you are going to have to actually refine either in large amounts or refine smaller amounts many, many times.

If you don't have the proof, its pretty easy to chalk your complaints up to Confirmation Bias, Small sample sizes, and the Margin of Error.

call it whatever you like.

i'm not obliged to prove a *** thing to you.

have a nice day.

Okay. I mean, I'm not your science teacher, so I can't make you prove your hypothesis.

But as the others above said, without proof, why should ZOS believe you? Why should the folks who refined a ton of mats to check the drop rate believe you?

Its a pretty basic "Anecdotes aren't data" situation.

Here's how I see it.

If you are right, and the new Multicraft is screwing up, and cheating you and others out of the generally accepted RNG for improvement mats, that's a big problem. It needs to be fixed.

But you can't just say "Its a problem!" With no backup and evidence to the contrary, ZOS won't listen to that. You have to prove it.

And your proof has to be in statistically significant amounts. Otherwise its very likely to be biased by confirmation bias, small sample sizes and vulnerable to the margin of error.

I realize that's not what you want to hear, but if there really is a problem, you are going to have to show that there's a problem. Especially when all the statistically significant proof we have shows that there isn't a problem.

So I'm not trying to call you a liar. Just trying to point out how you can prove you aren't mistaken, if you are so interested.

Personally, I have yet to see any statistically significant data sets that suggest there's a problem with improvement mat RNG. I'd love to see more people complaining about it actually record their data to find out.

that makes sense, tbh. as i stated earlier, i don't have sufficient materials to make an informed analysis, i just got really bad returns from a few thousand mats the first, and only, time i used the multicraft. (first 1000 - 1 gold, 2 purple, second 1000 4 gold, and the rest don't remember b/c i was getting truly pissed).

i'm really leery of it, now, and will probably just do single or small batch refining for a while.

maybe it's just bad luck or a bad day for me to refine mats. ¯\_(ツ)_/¯

thanks for the advice.

I totally understand that it can be frustrating. I actually wound up buying a 10k woodworking mats to test Multicraft out in batches of 1000, just to see if the results were noticeably below the average.

Now, 10k is still a relatively small amount in grand scheme of things. I'll be typing up my results in a separate post, most likely, but my results with gold mats were higher than average. So I'm guessing that as much as its disappointing to get only a few gold mats back, its probably within the normal range of RNG.

Alinhbo_Tyaka

Lylith wrote: »

VaranisArano wrote: »

Lylith wrote: »

VaranisArano wrote: »

Lylith wrote: »

XxXOMNIXxX wrote: »

i get an average of 1 gold mat per 100 refined materials. normally on average sometimes more sometimes less. but thats my average. now i get like 1 in 1000 refined materials @zos. this *** is unnaceptable. give me my old wayt of refining 1 at a time

i agree completely; it's absolute ***.

if you fanboys are getting better results, more power to you.

some of us aren't, doesn't make a *** what you say about it.

Here's the deal.

If you aren't getting those 1 in 200 results when you refine in statistically significant amounts, prove it.

And no, "I refined a couple hundred or a thousand and got sucky RNG" doesnt count. That's not statistically significant. You'd have to repeat that a bunch of times in order to reach a significant level of tests.

For context, the player(s) who tested it on the PTS and on Live refined over a hundred thousand mats and found that the drop rate hasnt changed. If you want to legitimately disprove that, you are going to have to actually refine either in large amounts or refine smaller amounts many, many times.

If you don't have the proof, its pretty easy to chalk your complaints up to Confirmation Bias, Small sample sizes, and the Margin of Error.

call it whatever you like.

i'm not obliged to prove a *** thing to you.

have a nice day.

Okay. I mean, I'm not your science teacher, so I can't make you prove your hypothesis.

But as the others above said, without proof, why should ZOS believe you? Why should the folks who refined a ton of mats to check the drop rate believe you?

Its a pretty basic "Anecdotes aren't data" situation.

Here's how I see it.

If you are right, and the new Multicraft is screwing up, and cheating you and others out of the generally accepted RNG for improvement mats, that's a big problem. It needs to be fixed.

But you can't just say "Its a problem!" With no backup and evidence to the contrary, ZOS won't listen to that. You have to prove it.

And your proof has to be in statistically significant amounts. Otherwise its very likely to be biased by confirmation bias, small sample sizes and vulnerable to the margin of error.

I realize that's not what you want to hear, but if there really is a problem, you are going to have to show that there's a problem. Especially when all the statistically significant proof we have shows that there isn't a problem.

So I'm not trying to call you a liar. Just trying to point out how you can prove you aren't mistaken, if you are so interested.

Personally, I have yet to see any statistically significant data sets that suggest there's a problem with improvement mat RNG. I'd love to see more people complaining about it actually record their data to find out.

that makes sense, tbh. as i stated earlier, i don't have sufficient materials to make an informed analysis, i just got really bad returns from a few thousand mats the first, and only, time i used the multicraft. (first 1000 - 1 gold, 2 purple, second 1000 4 gold, and the rest don't remember b/c i was getting truly pissed).

i'm really leery of it, now, and will probably just do single or small batch refining for a while.

maybe it's just bad luck or a bad day for me to refine mats. ¯\_(ツ)_/¯

thanks for the advice.

I accidentally multi-refined a 1K'ish stack of platinum grains. I received 8 chromium and 8 zircon. I have other cases where I have been lucky like this but my usual take is 5 - 8 total in a combination of purple and gold. I decided to see how the other crafts fared with multi. All material stack counts were between 750 and 1K. I saw nothing out of the ordinary as the purple and gold materials appeared to be within the range I typically see. Even with the old one at a time refining I would have some runs where I didn't see more than zero, one or two higher level tempers just as I would occasionally hit the jackpot. I think you need to do more runs over a longer period to get a clearer picture of the drop rates.

katanagirl1

I’m on console so I don’t have access to PTS and the multi craft yet, but I rarely refine more than 1k mats at a time. Are you guys suggesting that I need to refine that many to have good results? That’s what‘s really bothering me when I read this. I just refined about 2k each ancestor silk and rubedo leather from the Orsinium event and had something like 11 dreugh wax. Similar results for rubedo ore and ruby ash.

I never get 0 gold mats for even 500-600k mats refining 10 at a time.

tonne.backlinderb16_ESO

Lylith wrote: »

VaranisArano wrote: »

Lylith wrote: »

XxXOMNIXxX wrote: »

i get an average of 1 gold mat per 100 refined materials. normally on average sometimes more sometimes less. but thats my average. now i get like 1 in 1000 refined materials @zos. this *** is unnaceptable. give me my old wayt of refining 1 at a time

i agree completely; it's absolute ***.

if you fanboys are getting better results, more power to you.

some of us aren't, doesn't make a *** what you say about it.

Here's the deal.

If you aren't getting those 1 in 200 results when you refine in statistically significant amounts, prove it.

And no, "I refined a couple hundred or a thousand and got sucky RNG" doesnt count. That's not statistically significant. You'd have to repeat that a bunch of times in order to reach a significant level of tests.

For context, the player(s) who tested it on the PTS and on Live refined over a hundred thousand mats and found that the drop rate hasnt changed. If you want to legitimately disprove that, you are going to have to actually refine either in large amounts or refine smaller amounts many, many times.

If you don't have the proof, its pretty easy to chalk your complaints up to Confirmation Bias, Small sample sizes, and the Margin of Error.

call it whatever you like.

i'm not obliged to prove a *** thing to you.

have a nice day.

You're a quite nice guy

VaranisArano

Here's my test with 10k mats, refined in batches of 1000.

https://forums.elderscrollsonline.com/en/discussion/490789/multicraft-refining-small-batches-test-10k-mats

I did not see a lowered drop rate. Actually, due to the relatively small sample size, my drop rate for gold improvement mats was higher than the expected result of 1 in 200. That's not to say that folks don't see worse than average results, but rather indicates that ZOS didnt lower the overall drop rate.

katanagirl1 wrote: »

I’m on console so I don’t have access to PTS and the multi craft yet, but I rarely refine more than 1k mats at a time. Are you guys suggesting that I need to refine that many to have good results? That’s what‘s really bothering me when I read this. I just refined about 2k each ancestor silk and rubedo leather from the Orsinium event and had something like 11 dreugh wax. Similar results for rubedo ore and ruby ash.

I never get 0 gold mats for even 500-600k mats refining 10 at a time.

Based on my results and the others who have refine many, many more mats than I, it shouldn't matter how many mats you refine at once. It doesn't look like the drop rate changed, so over time, your refinements will approach the expected mean of 1 in 200 for gold mats.

tmbrinks

katanagirl1 wrote: »

I’m on console so I don’t have access to PTS and the multi craft yet, but I rarely refine more than 1k mats at a time. Are you guys suggesting that I need to refine that many to have good results? That’s what‘s really bothering me when I read this. I just refined about 2k each ancestor silk and rubedo leather from the Orsinium event and had something like 11 dreugh wax. Similar results for rubedo ore and ruby ash.

I never get 0 gold mats for even 500-600k mats refining 10 at a time.

You don't ever need to refine that many to get "good" results.

It's just that refining smaller amounts means there's a much larger potential spread in the data, that you got "unlucky".

Larger amounts cause the standard normal distribution to become more clustered in the center where "normal" drop rate is.

https://demonstrations.wolfram.com/DistributionOfNormalMeansWithDifferentSampleSizes/

Take a look at this and what it does to the distribution curve as you increase the sample size.

adree

Lylith wrote: »

call it whatever you like.

i'm not obliged to prove a *** thing to you.

have a nice day.

this wellknown zos's ........ trolling and bushing ppl making their posts looks like not serious with such a comments

gepe87

My test:
60 raw ancestor- 0 dreugh
30 rubedo - 2 dreugh
80 ruby - 0 rosin
30 platinum - 1 chromium
60 rubedite - 1 alloy

Not so bad!!

Parrot1986

Overall like the crafting changes in terms of multi cast and deconstruct. I wasn’t sure about the refine since it has such a high value and impact.
I’ve only done one lot of refining so far but it wasn’t as bad as I thought but will need to do a few more mass refined to be sure.

From 1.8k platinum I got 16 chromium grains
2.4k ore 18 alloy
2.3k ruby ash 16 rosin
2.5k silk and leather 16 wax

Get that not everyone is getting these results for every one but this feels ok so far and in line with what I used to get.

katanagirl1

tmbrinks wrote: »

katanagirl1 wrote: »

I’m on console so I don’t have access to PTS and the multi craft yet, but I rarely refine more than 1k mats at a time. Are you guys suggesting that I need to refine that many to have good results? That’s what‘s really bothering me when I read this. I just refined about 2k each ancestor silk and rubedo leather from the Orsinium event and had something like 11 dreugh wax. Similar results for rubedo ore and ruby ash.

I never get 0 gold mats for even 500-600k mats refining 10 at a time.

You don't ever need to refine that many to get "good" results.

It's just that refining smaller amounts means there's a much larger potential spread in the data, that you got "unlucky".

Larger amounts cause the standard normal distribution to become more clustered in the center where "normal" drop rate is.

https://demonstrations.wolfram.com/DistributionOfNormalMeansWithDifferentSampleSizes/

Take a look at this and what it does to the distribution curve as you increase the sample size.

For some reason I am unable to open that link, but you seem to get my point. Big sample sizes are better, yes. However, if people really are getting 0 gold mats for 1k mats refined or larger then it seems there’s a big problem. Is the +-3*sigma standard deviation of this normal distribution large enough to reasonably get zero in this case? I only get no gold mats if I just decon a day’s worth of farming.

tmbrinks

https://docs.google.com/spreadsheets/d/1qu4Pk3QZv6YYIEHy7oLXQU_jKxz-7IOPlxGA4eNRnQI/edit?usp=sharing

I very quickly made this for a standard binomial distribution.

It's for both N = 1000 (100 refines) and N = 10000 (1000 refines) and the odds of getting specific number of gold improvement materials while refining that much materials.

As you can see, getting 0 materials when refining 1,000 materials should happen once every 150 times it's done. Rare, yes. But certainly not impossible.

Assuming a 0.5% drop rate on the gold materials.

tmbrinks

tmbrinks wrote: »

https://docs.google.com/spreadsheets/d/1qu4Pk3QZv6YYIEHy7oLXQU_jKxz-7IOPlxGA4eNRnQI/edit?usp=sharing

I very quickly made this for a standard binomial distribution.

It's for both N = 1000 (100 refines) and N = 10000 (1000 refines) and the odds of getting specific number of gold improvement materials while refining that much materials.

As you can see, getting 0 materials when refining 1,000 materials should happen once every 150 times it's done. Rare, yes. But certainly not impossible.

Assuming a 0.5% drop rate on the gold materials.

https://docs.google.com/spreadsheets/d/1yQDUjYNQVsIFl0ktkbkSlfYPkzP6pgCOTMQ-qCuzfaI/edit?usp=sharing

I have also added a similar calculation to my data in my main spreadsheet showing the "percentile" for each of my refines.

This is all assuming a 1.5% drop rate for Green, 1.25% for Blue, 0.75% drop rate for Purple, and 0.5% drop rate for Gold, and then comparing that to the actual data to see if there's anything statistically significant about it.

Ironically, my drop rate for tempering alloys is in the 2nd percentile (yikes), meaning I've had poor luck. It also means that the drop rate for tempering alloys might be the ONLY one that is statistically significantly off, and the drop rate MIGHT be different... going off a 95% confidence interval. At a 99% confidence interval (which I think is more commonly used now) we are still okay in saying the drop rate is 0.5%

Lylith

Alinhbo_Tyaka wrote: »

Lylith wrote: »

VaranisArano wrote: »

Lylith wrote: »

VaranisArano wrote: »

Lylith wrote: »

XxXOMNIXxX wrote: »

i get an average of 1 gold mat per 100 refined materials. normally on average sometimes more sometimes less. but thats my average. now i get like 1 in 1000 refined materials @zos. this *** is unnaceptable. give me my old wayt of refining 1 at a time

i agree completely; it's absolute ***.

if you fanboys are getting better results, more power to you.

some of us aren't, doesn't make a *** what you say about it.

Here's the deal.

If you aren't getting those 1 in 200 results when you refine in statistically significant amounts, prove it.

And no, "I refined a couple hundred or a thousand and got sucky RNG" doesnt count. That's not statistically significant. You'd have to repeat that a bunch of times in order to reach a significant level of tests.

For context, the player(s) who tested it on the PTS and on Live refined over a hundred thousand mats and found that the drop rate hasnt changed. If you want to legitimately disprove that, you are going to have to actually refine either in large amounts or refine smaller amounts many, many times.

If you don't have the proof, its pretty easy to chalk your complaints up to Confirmation Bias, Small sample sizes, and the Margin of Error.

call it whatever you like.

i'm not obliged to prove a *** thing to you.

have a nice day.

Okay. I mean, I'm not your science teacher, so I can't make you prove your hypothesis.

But as the others above said, without proof, why should ZOS believe you? Why should the folks who refined a ton of mats to check the drop rate believe you?

Its a pretty basic "Anecdotes aren't data" situation.

Here's how I see it.

If you are right, and the new Multicraft is screwing up, and cheating you and others out of the generally accepted RNG for improvement mats, that's a big problem. It needs to be fixed.

But you can't just say "Its a problem!" With no backup and evidence to the contrary, ZOS won't listen to that. You have to prove it.

And your proof has to be in statistically significant amounts. Otherwise its very likely to be biased by confirmation bias, small sample sizes and vulnerable to the margin of error.

I realize that's not what you want to hear, but if there really is a problem, you are going to have to show that there's a problem. Especially when all the statistically significant proof we have shows that there isn't a problem.

So I'm not trying to call you a liar. Just trying to point out how you can prove you aren't mistaken, if you are so interested.

Personally, I have yet to see any statistically significant data sets that suggest there's a problem with improvement mat RNG. I'd love to see more people complaining about it actually record their data to find out.

that makes sense, tbh. as i stated earlier, i don't have sufficient materials to make an informed analysis, i just got really bad returns from a few thousand mats the first, and only, time i used the multicraft. (first 1000 - 1 gold, 2 purple, second 1000 4 gold, and the rest don't remember b/c i was getting truly pissed).

i'm really leery of it, now, and will probably just do single or small batch refining for a while.

maybe it's just bad luck or a bad day for me to refine mats. ¯\_(ツ)_/¯

thanks for the advice.

I accidentally multi-refined a 1K'ish stack of platinum grains. I received 8 chromium and 8 zircon. I have other cases where I have been lucky like this but my usual take is 5 - 8 total in a combination of purple and gold. I decided to see how the other crafts fared with multi. All material stack counts were between 750 and 1K. I saw nothing out of the ordinary as the purple and gold materials appeared to be within the range I typically see. Even with the old one at a time refining I would have some runs where I didn't see more than zero, one or two higher level tempers just as I would occasionally hit the jackpot. I think you need to do more runs over a longer period to get a clearer picture of the drop rates.

thanks for the input.

like the op, i was used to getting around 1/100 or so, i didn't realize it was 1/200, so that threw me.

that, and getting 1/1000 right off the bat, ofc.