"Small Refine Method" AKA Time Wasting Pseudo Science Scam

rhythmsuji · September 19

Hey all, as the title says. A very long term malignant "trick" or "hack" has continually made its round and can be seen in chat as advice, some YouTube pushing it for clicks as a "helpful" tip. Or commenters in refinement videos telling YouTubers to do it lol. SO I am here to spread the word.

Just want to confirm here and now, that ive tested this many times in various ways over the years. BUT it was prior to making my own helpful content, so all I had was my typed word to a small group of people.

This time I did a mass scale refinement of 500,000 materials, half full 10,000 stacks (the maximum) and the other half in stacks of 10 (the minimum) with no difference. All the drop percentages were in upward or downward trends of less than 1% regardless of stack size, and no pattern which could crown 1 higher than the other in any scientific way.

https://www.youtube.com/watch?v=OiD3NEk_p_Y&t=1s&ab_channel=Feanor

Hope this helps some more people to save a lot of time and tedium.

-ESO Feanor

hiyde · September 19

Thanks! If there's one thing that drives me nuts, it's useless clickbait.

SkaiFaith · September 19

With the "small technique" I once refined 2.000+ Platinum without getting a single Chromium.
Being in small sizes or big chunks, such a terrible RNG shouldn't happen IMO.
(EDIT: I think I was near 4.000 refinement when it finally dropped. Only happened one time, but still... It's a huge bummer)

joergino · September 19

Thanks for your work and your huge effort. Hopefully this will help to bust this myth for good.

LootAllTheStuff · September 19

SkaiFaith wrote: »

With the "small technique" I once refined 2.000+ Platinum without getting a single Chromium.
Being in small sizes or big chunks, such a terrible RNG shouldn't happen IMO.
(EDIT: I think I was near 4.000 refinement when it finally dropped. Only happened one time, but still... It's a huge bummer)

That is really bad. If the chance of a gold mat from refining is taken as 1% (seems to be the rate for perfect roe too) then, assuming I did my mathematics correctly, the odds of getting only 1 gold mat in 100 tries is about 37%. For 150 tries, that drops to 33%. Unfortunately, Excel can't seem to handle calculating the binomial coefficient for more than 170 tries (31%). (Perhaps there's a way to set up the calculation of larger factorials that gets around the memory limitation?). Regardless, going even 1000 tries without getting a gold mat would seem to be extremely bad luck.

Thumbless_Bot · September 19

LootAllTheStuff wrote: »

SkaiFaith wrote: »

With the "small technique" I once refined 2.000+ Platinum without getting a single Chromium.
Being in small sizes or big chunks, such a terrible RNG shouldn't happen IMO.
(EDIT: I think I was near 4.000 refinement when it finally dropped. Only happened one time, but still... It's a huge bummer)

That is really bad. If the chance of a gold mat from refining is taken as 1% (seems to be the rate for perfect roe too) then, assuming I did my mathematics correctly, the odds of getting only 1 gold mat in 100 tries is about 37%. For 150 tries, that drops to 33%. Unfortunately, Excel can't seem to handle calculating the binomial coefficient for more than 170 tries (31%). (Perhaps there's a way to set up the calculation of larger factorials that gets around the memory limitation?). Regardless, going even 1000 tries without getting a gold mat would seem to be extremely bad luck.

Excel has a Power Query addon that might help. Ive used ot in the past when excel would freeze up on calcs.

Pevey · September 19

Having some fun by putting on my statistics nerd hat on for minute, the easiest way to analyze this question would be to use the F distribution (see https://en.wikipedia.org/wiki/F-distribution). To summarize the link, if you have a random variable X such as the outcome of a coin flip (heads or tails) the cumulative F distribution generalizes the probability (y-axis) of the null hypothesis outcome after a given number of coin tosses (x-axis). The line approaches one but never quite gets there, since it is theoretically possible you could flip a coin forever and never get heads, or roll a die forever and never get a 1. But extremely unlikely. The F-distribution can be used similarly to a normal or student's t to set up hypothesis tests.

SkaiFaith · September 19

LootAllTheStuff wrote: »

SkaiFaith wrote: »

With the "small technique" I once refined 2.000+ Platinum without getting a single Chromium.
Being in small sizes or big chunks, such a terrible RNG shouldn't happen IMO.
(EDIT: I think I was near 4.000 refinement when it finally dropped. Only happened one time, but still... It's a huge bummer)

That is really bad. If the chance of a gold mat from refining is taken as 1% (seems to be the rate for perfect roe too) then, assuming I did my mathematics correctly, the odds of getting only 1 gold mat in 100 tries is about 37%. For 150 tries, that drops to 33%. Unfortunately, Excel can't seem to handle calculating the binomial coefficient for more than 170 tries (31%). (Perhaps there's a way to set up the calculation of larger factorials that gets around the memory limitation?). Regardless, going even 1000 tries without getting a gold mat would seem to be extremely bad luck.

Yeah, I got kinda ""traumatized"" that time and I've stopped refining Platinums since. It's been such a frustrating experience.
Sooner or later I'll have to retry but it's kinda scary now XD
Never had problems before that time; generally I get a gold drop every 100-200 refined mats (10-20 tries)

zaria · September 19

rhythmsuji wrote: »

This time I did a mass scale refinement of 500,000 materials, half full 10,000 stacks (the maximum) and the other half in stacks of 10 (the minimum) with no difference. All the drop percentages were in upward or downward trends of less than 1% regardless of stack size, and no pattern which could crown 1 higher than the other in any scientific way.

Hope this helps some more people to save a lot of time and tedium.
-ESO Feanor

It's that I call game superstition, as the calculations are hidden from us people come up with various superstitions.
Anybody remember the fortify luck spell you was supposed to cast before entering a location in TES 4 Oblivion to get better loot and so many said it worked well, granted it was stated somewhere that loot increased loot.
But it only affected gold dropped if anything. I tested this with an simple mod, add an barrel just inside an cave, it had 1000 level scaled armor, weapons and alchemy gear, I entered with luck 0 and luck 100 + some from enchants.
No difference at all.

Now ESO has played dirty here multiple times. Remember farming white gold tower for the spell power cure restoration staff.
so many runs, no resto staff and also no SPC jewelry outside the amulet you get from the quest but so many bows

And the undaunted shoulders, was a a post of some opening 20 chests and only getting the other set.
Who equals to flipping a coin 20 time and only getting heads, one in a million chance.

And guess they still playing these tricks for stuff like motif and blueprints. But for refining, it makes no sense at all.

Elvenheart · September 19

I recall a very long time ago, way before the sticker book and curated items, getting the Burning Spellweave Inferno Staff at lvl 45 on a first run before ever knowing how desired it was, so of course I deconstructed it. Later, when I wanted to try out that build, I cannot remember how many times I ran that dungeon with groups until I got to the point I could solo it to try and get it to drop. Two weeks before item curation was announced, it finally dropped for me, long after I was really interested in the set any more, it had just become a thing to do in the game.

LootAllTheStuff · September 19

I think the best thing to be said for refining relatively small numbers of mats at a time is that it saves the pain of hitting a really bad patch in the RNG number sequence. Go whack some trash mobs or something, then come back - anything to stay sane.

zaria wrote: »

rhythmsuji wrote: »

This time I did a mass scale refinement of 500,000 materials, half full 10,000 stacks (the maximum) and the other half in stacks of 10 (the minimum) with no difference. All the drop percentages were in upward or downward trends of less than 1% regardless of stack size, and no pattern which could crown 1 higher than the other in any scientific way.

Hope this helps some more people to save a lot of time and tedium.
-ESO Feanor

It's that I call game superstition, as the calculations are hidden from us people come up with various superstitions.
Anybody remember the fortify luck spell you was supposed to cast before entering a location in TES 4 Oblivion to get better loot and so many said it worked well, granted it was stated somewhere that loot increased loot.
But it only affected gold dropped if anything. I tested this with an simple mod, add an barrel just inside an cave, it had 1000 level scaled armor, weapons and alchemy gear, I entered with luck 0 and luck 100 + some from enchants.
No difference at all.

Now ESO has played dirty here multiple times. Remember farming white gold tower for the spell power cure restoration staff.
so many runs, no resto staff and also no SPC jewelry outside the amulet you get from the quest but so many bows

And the undaunted shoulders, was a a post of some opening 20 chests and only getting the other set.
Who equals to flipping a coin 20 time and only getting heads, one in a million chance.

And guess they still playing these tricks for stuff like motif and blueprints. But for refining, it makes no sense at all.

There's a difference between a coding error (accidentally missing entries in the loot table in the shipped game, or failing to invoke the table at all) and RNG just being RNG. The biggest issue from a player's perspective is always the combination of an inherently low drop chance AND no loot curation - especially if you're trying to get a specific piece in a set of 14. It's either feast or famine. For example, I ran SC2 last night and got three Prayer Shoes, which means I'm still missing items from the set. I guess that's why there's a trading system, but still...

Jaimeh · September 19

I've heard about this before but I'm not that patient and always refined everything in one go--good to know that at least I wasn't missing out on anything, so thank you for sharing and kudos for the big undertaking.

rhythmsuji · September 20

Jaimeh wrote: »

I've heard about this before but I'm not that patient and always refined everything in one go--good to know that at least I wasn't missing out on anything, so thank you for sharing and kudos for the big undertaking.

No problem! And im glad I could ease the potential concern at minimum.

Hotdog_23 · September 20

Elvenheart wrote: »

I recall a very long time ago, way before the sticker book and curated items, getting the Burning Spellweave Inferno Staff at lvl 45 on a first run before ever knowing how desired it was, so of course I deconstructed it. Later, when I wanted to try out that build, I cannot remember how many times I ran that dungeon with groups until I got to the point I could solo it to try and get it to drop. Two weeks before item curation was announced, it finally dropped for me, long after I was really interested in the set any more, it had just become a thing to do in the game.

I still have a Burning Spellweave Inferno Staff in my inventory from way back before item curation. Dropped in charged trait. Even if I never use the set again, and honestly haven’t in years. I refused to get rid of it just because of how hard it was to get and how happy I was to get it.

Stay safe

twisttop138 · September 20

We talked about this on the reddit thread but I will repeat it here as I would like to hear if this is wide spread.

I had a raid lead who very much believed deep down in her soul that while in a dungeon or a trial, if you were to force the lock on a chest instead of picking it, that it would lower the amount and quality of items in the chest for everyone.

Now, I do engage in superstition. i was told long ago and very much held to refining 1 at a time. Now I know it's just rng, and I'd like to pirate a response in this thread that I was just trying to mitigate a bad round of RNG but that's not true. That's my excuse now though. Thank you @LootAllTheStuff
But I would hope no one would be fooled by my story. That just seems ludicrous.

LootAllTheStuff · September 20

@twisttop138 : Agreed, that is quite ludicrous. In order for it to be actually true, the developers would have to code it in to the whole process. It's doable, but why would you add complexity to something that doesn't need it? More coding complexity = more chances for bugs + harder debugging. It fails Occam's Razor, since RNG with a low drop chance for desirable items is already all the explanation needed.

KiltMaster · September 20

eh it's kinda like people have their superstitions ... whatever works for them, right?

zaria · September 20

LootAllTheStuff wrote: »

I think the best thing to be said for refining relatively small numbers of mats at a time is that it saves the pain of hitting a really bad patch in the RNG number sequence. Go whack some trash mobs or something, then come back - anything to stay sane.

zaria wrote: »

rhythmsuji wrote: »

This time I did a mass scale refinement of 500,000 materials, half full 10,000 stacks (the maximum) and the other half in stacks of 10 (the minimum) with no difference. All the drop percentages were in upward or downward trends of less than 1% regardless of stack size, and no pattern which could crown 1 higher than the other in any scientific way.

Hope this helps some more people to save a lot of time and tedium.
-ESO Feanor

It's that I call game superstition, as the calculations are hidden from us people come up with various superstitions.
Anybody remember the fortify luck spell you was supposed to cast before entering a location in TES 4 Oblivion to get better loot and so many said it worked well, granted it was stated somewhere that loot increased loot.
But it only affected gold dropped if anything. I tested this with an simple mod, add an barrel just inside an cave, it had 1000 level scaled armor, weapons and alchemy gear, I entered with luck 0 and luck 100 + some from enchants.
No difference at all.

Now ESO has played dirty here multiple times. Remember farming white gold tower for the spell power cure restoration staff.
so many runs, no resto staff and also no SPC jewelry outside the amulet you get from the quest but so many bows

And the undaunted shoulders, was a a post of some opening 20 chests and only getting the other set.
Who equals to flipping a coin 20 time and only getting heads, one in a million chance.

And guess they still playing these tricks for stuff like motif and blueprints. But for refining, it makes no sense at all.

There's a difference between a coding error (accidentally missing entries in the loot table in the shipped game, or failing to invoke the table at all) and RNG just being RNG. The biggest issue from a player's perspective is always the combination of an inherently low drop chance AND no loot curation - especially if you're trying to get a specific piece in a set of 14. It's either feast or famine. For example, I ran SC2 last night and got three Prayer Shoes, which means I'm still missing items from the set. I guess that's why there's a trading system, but still...

Not sure how ESO handles bulk refining materials, if their doing 1000 refinements cycles if you refine 8000 materials or if they roll once for each of the loot pools. and do RND number x drop chance x number of cycles.

And by the SPC healing staff I say they deliberately set the drop rate for SPC staffs and jewelry much lower than random would indicate.

As for getting 3 prayer shoes, if you has collected all armor in a dungeon, armor is random, chests is also random. Weapons only drop from last boss and chests. If you just missed a few items it recommended to bind armor on drop.
Say you missed shoes and belt. and you did bind shoes, next drop from a boss should be belt.

Some people sell weapon runs, they miss a bis weapon in an dungeon on purpose. You and the client queue for the dungeon, either you have two friend you pay a bit for the run, or you pug it, normal works here.
You get the bis weapon as the only missing, you then sell it to client.
Repeat.

tmbrinks · September 20

Refine all at once. Save yourself time. The game just does every refine as an individual calculation anyways. There's no "bad RNG seed", no "multiplier" factor.

I have data tabs in my long-shared spreadsheet (in my signature) that has millions of refines in it. Both in bulk and single refines before Multicraft even existed in the game. There is no statistical difference. Full stop. (The drop rates did change with Meticulous Disassembly being introduced in 2021)

zaria · September 20

twisttop138 wrote: »

We talked about this on the reddit thread but I will repeat it here as I would like to hear if this is wide spread.

I had a raid lead who very much believed deep down in her soul that while in a dungeon or a trial, if you were to force the lock on a chest instead of picking it, that it would lower the amount and quality of items in the chest for everyone.

Now, I do engage in superstition. i was told long ago and very much held to refining 1 at a time. Now I know it's just rng, and I'd like to pirate a response in this thread that I was just trying to mitigate a bad round of RNG but that's not true. That's my excuse now though. Thank you @LootAllTheStuff
But I would hope no one would be fooled by my story. That just seems ludicrous.

That one makes no sense, why should the code in loot quality based on how the chest is opened.
Loot is better in harder chests you get one level harder chests in veteran, also loot is purple in vet.
But loot in chests in trials with perfected gear is normal, this is to avoid chest farming in trials like LoM.

rhythmsuji · September 20

KiltMaster wrote: »

eh it's kinda like people have their superstitions ... whatever works for them, right?

This is a perfectly fine perspective, if people >Want< to do it anyway. But I want to make sure they are informed before hand.

AND mostly because some creators have even profited by making false information videos suggesting or claiming the "trick" to be the "correct way"

And the main important factor there, is when you look at the amount of time the player ends up wasting in order to do it this way. I spent 10 mins doing full stacks, and over 4 and a half hours doing the other half of materials in minimum stacks.

Over time, this can end up being days or more of someones life "doing a trick" which did nothing for them. So as long as they are informed, then yeah thats fully a fine choice.

rhythmsuji · September 20

zaria wrote: »

LootAllTheStuff wrote: »

I think the best thing to be said for refining relatively small numbers of mats at a time is that it saves the pain of hitting a really bad patch in the RNG number sequence. Go whack some trash mobs or something, then come back - anything to stay sane.

zaria wrote: »

rhythmsuji wrote: »

This time I did a mass scale refinement of 500,000 materials, half full 10,000 stacks (the maximum) and the other half in stacks of 10 (the minimum) with no difference. All the drop percentages were in upward or downward trends of less than 1% regardless of stack size, and no pattern which could crown 1 higher than the other in any scientific way.

Hope this helps some more people to save a lot of time and tedium.
-ESO Feanor

It's that I call game superstition, as the calculations are hidden from us people come up with various superstitions.
Anybody remember the fortify luck spell you was supposed to cast before entering a location in TES 4 Oblivion to get better loot and so many said it worked well, granted it was stated somewhere that loot increased loot.
But it only affected gold dropped if anything. I tested this with an simple mod, add an barrel just inside an cave, it had 1000 level scaled armor, weapons and alchemy gear, I entered with luck 0 and luck 100 + some from enchants.
No difference at all.

Now ESO has played dirty here multiple times. Remember farming white gold tower for the spell power cure restoration staff.
so many runs, no resto staff and also no SPC jewelry outside the amulet you get from the quest but so many bows

And the undaunted shoulders, was a a post of some opening 20 chests and only getting the other set.
Who equals to flipping a coin 20 time and only getting heads, one in a million chance.

And guess they still playing these tricks for stuff like motif and blueprints. But for refining, it makes no sense at all.

There's a difference between a coding error (accidentally missing entries in the loot table in the shipped game, or failing to invoke the table at all) and RNG just being RNG. The biggest issue from a player's perspective is always the combination of an inherently low drop chance AND no loot curation - especially if you're trying to get a specific piece in a set of 14. It's either feast or famine. For example, I ran SC2 last night and got three Prayer Shoes, which means I'm still missing items from the set. I guess that's why there's a trading system, but still...

Not sure how ESO handles bulk refining materials, if their doing 1000 refinements cycles if you refine 8000 materials or if they roll once for each of the loot pools. and do RND number x drop chance x number of cycles.

And by the SPC healing staff I say they deliberately set the drop rate for SPC staffs and jewelry much lower than random would indicate.

As for getting 3 prayer shoes, if you has collected all armor in a dungeon, armor is random, chests is also random. Weapons only drop from last boss and chests. If you just missed a few items it recommended to bind armor on drop.
Say you missed shoes and belt. and you did bind shoes, next drop from a boss should be belt.

Some people sell weapon runs, they miss a bis weapon in an dungeon on purpose. You and the client queue for the dungeon, either you have two friend you pay a bit for the run, or you pug it, normal works here.
You get the bis weapon as the only missing, you then sell it to client.
Repeat.

Its per minimum stack.

As shown in the video, the outcomes are the same/inconsequential. So the drop chance, lets call it 5.7% will always shake out to that, a 5.7% chance on average per 10 materials refined. No matter the stack size.

PurpleScroll · September 21

Thanks for actually testing this thing out, I ended up with similar results.

rhythmsuji · September 22

PurpleScroll wrote: »

Thanks for actually testing this thing out, I ended up with similar results.

No problem! And thank you for the feedback. Happy hunting!

frogthroat · September 22

SkaiFaith wrote: »

With the "small technique" I once refined 2.000+ Platinum without getting a single Chromium.
Being in small sizes or big chunks, such a terrible RNG shouldn't happen IMO.
(EDIT: I think I was near 4.000 refinement when it finally dropped. Only happened one time, but still... It's a huge bummer)

Sorry to hear that. However...
If it is truly random, this kind of bad luck should absolutely happen. Random does not mean uniformly distributed. Sometimes there might be 10 "jackpots" in a row, other times there might be a "dry spell" for 1000 mats.

In fact, if you would look at a list of random things, say, a 1000 die rolls or coin flips, you can easily see if it is truly thrown or if a human just wrote something random down because they couldn't be bothered to actually throw a die or a coin 1000 times. If the numbers are uniformly distributed with no more than maximum 7 same results in a row, it's probably human generated. If there's strange, long streaks in the mix, that is a good indication it is randomly generated.

That's why different music player providers have to tweak their "shuffle" function because truly random would not feel random.

wolfie1.0. · September 22

Been fighting this for years. Another forum user has the data. The only difference between small batch refining and mass batch has always only been the time it took to mass refine, with the very very rare exception of bugs.

But players are superstitious by nature. I know a player that insisted that lvl1 mats gave more gold tempers than any other mat.

frogthroat · September 22

I always just click the "Mass Refine" button that comes from Mass Deconstruct addon. Never checked the options for the refinement.

Now I checked. Hey, it has an option: "Refine full stacks" with default is off.

I usually leave my crafter just refine while I go do something else, but now I don't have to. Never given it much thought, tho.

SkaiFaith · 11:07AM

frogthroat wrote: »

SkaiFaith wrote: »

With the "small technique" I once refined 2.000+ Platinum without getting a single Chromium.
Being in small sizes or big chunks, such a terrible RNG shouldn't happen IMO.
(EDIT: I think I was near 4.000 refinement when it finally dropped. Only happened one time, but still... It's a huge bummer)

Sorry to hear that. However...
If it is truly random, this kind of bad luck should absolutely happen. Random does not mean uniformly distributed. Sometimes there might be 10 "jackpots" in a row, other times there might be a "dry spell" for 1000 mats.

In fact, if you would look at a list of random things, say, a 1000 die rolls or coin flips, you can easily see if it is truly thrown or if a human just wrote something random down because they couldn't be bothered to actually throw a die or a coin 1000 times. If the numbers are uniformly distributed with no more than maximum 7 same results in a row, it's probably human generated. If there's strange, long streaks in the mix, that is a good indication it is randomly generated.

That's why different music player providers have to tweak their "shuffle" function because truly random would not feel random.

My opinion tends to be that anything that has RNG (crates, leads, plans, mats, motif, gear...) should have a pity system. Such bad RNG shouldn't happen IMO, because it pushes away players.

AzuraFan · 2:02PM

twisttop138 wrote: »

I had a raid lead who very much believed deep down in her soul that while in a dungeon or a trial, if you were to force the lock on a chest instead of picking it, that it would lower the amount and quality of items in the chest for everyone.

I always force locks on dungeon chests and get weapons/jewelry out of the non-simple ones quite often, so that's not true, but I can understand why someone would have that superstition.

SkaiFaith wrote: »

My opinion tends to be that anything that has RNG (crates, leads, plans, mats, motif, gear...) should have a pity system. Such bad RNG shouldn't happen IMO, because it pushes away players.

Absolutely. In the past, when there was still lots left for me to do, the only time I thought of quitting the game was when I'd tried for a lead for the umpteenth time and didn't get it. It's unreasonable to have such pathetically low drop rates when leads are at the end of long dungeons. Just disrespectful of players' time. Then there was the 10th anniversary fiasco when some players got the shinies almost instantly, and others tried thousands of times with no luck. Something that was supposed to be a happy celebration turned into a horrible experience for too many.

It's not fun, and while I know there have to be hamster wheels in MMOs, other games handle this much better than ESO ever has. They definitely need a pity system, or they need to increase the currently ultra-low probabilities of success. ZOS has never walked the grind vs. fun balance very well, something they should revisit if they're interested in maintaining a healthy population.