Goldbeater is one impossible equation.

[Xcom]

Been hard at work trying to figure out the exact equation of goldbeater skins.

After rigures tests I figured out that Sqrt(sewing * dex) is the softcap.

This was done by having a high Psy and Dex char softcap the goldbeater. So was the high Cooking Percep char. But when I put thimbles on the high Psy Dex char with only 1 sewing I saw a major jump in Q. Then I had a high Dex Sewing char craft them and I saw a very large jump in the Q.

To make sure no other skills were involved I put thimbles on the high Sewing Dex char and it didn't make any Q difference. With 79 Dex 90 Sewing and Q71 goldbeater I was convinced that no other skills could be involved in the softcaping of goldbeaters as it would have upped the Q if it was a 3Sqrt (dex * sew * 3rd skill) equation.

So after putting that aside I started to make Q experiments on the goldbeaters to find out the final equation. But so far I haven't been able to crack it and Im starting to get sick and tired of it. Every new data I put in makes things even worse. Maybe with some help and some more even numbers it might be solvable. But as of now Im gonna put it aside.

Although I verified that Q of fuel used will not affect the outcome. Q10 and Q101 fuel made no difference.

Maybe someone here in the community might have a crack at it and help figure it out for the wiki. I can also include that I did check the Q of all materials before each experiment. I can also make sure that the Q of caldron is what it is as I used clever Qs of different wax and leather to make hard leather to verify the Q.

My data:
Water Caldron Vingegar Intests Goldbeater
Intest
30 82 140 36 67
30 82 140 40 70
30 82 140 43 71
30 82 140 44 72
30 82 140 50 75
30 82 140 53 77

Vinegar Test
30 82 21 43 38
30 82 21 44 39
30 82 80 44 55

Water Test
10 82 140 43 70
10 82 140 44 70
60 82 140 44 75

Caldron Test
30 11 140 44 66

Edit: added numbers

Its clear that the higher the Q you go the better Q you get but the relation is hard to figure out.

I found that every Q of Intests you go you rase the goldbeater by .56 to .6
Similarly with Vinegar you rase the goldbeater by .35 to .37
Water and Caldron roughly .1 to .12

Looking at the HL equation you have the (Caldron + Water) together so it might be similar here.

Best I could come up with was:

Goldbeater = ( Caldron * .1 + Water * .1 + Intests * .6 + Vinegar .35 ) / 1.15
But nothing comes out right when you put in the numbers.

Hope someone figures it out and posts something.

[MagicManICT]

Great work on getting some numbers. I've not really checked on the quality information for it, so if nothing else, you might have produced the unknown softcap.

As far as I'm aware, the cauldron quality only matters in one or two known cases, but I can't recall those off the top of my head. It is quite possible it doesn't matter here, either. The only thing is the clay cauldron automatically causing the water quality to drop by half.

[Xcom]

Great work on getting some numbers. I've not really checked on the quality information for it, so if nothing else, you might have produced the unknown softcap.

As far as I'm aware, the cauldron quality only matters in one or two known cases, but I can't recall those off the top of my head. It is quite possible it doesn't matter here, either. The only thing is the clay cauldron automatically causing the water quality to drop by half.

I used a metal caldron in all instances but the softcap is one bugging annoying thing. Because of the impossible equation involved I feel like I want to have a high lvl crafter try and craft a known goldbeater to check if his high stats dosen't softcap as well. I'm mostly sure the softcap is correctly assumed to be dex sewing but its just bugging me that the equation is so complex that it might be softcaped in there somewhere and give strange results.

[Potjeh]

IIRC cauldron recipes don't directly use cauldron q, rather the cauldron softcaps water and clay cauldron is always q 0.

[Xcom]

After making some crazy stuff in excel I figured out an equation but not sure if its correct.

Qgoldbeater = ( ( water * 9 + caldron * 8 ) / 17 + ( ( intest * 2 + vinegar ) / 3 ) * 5 ) / 6

Q of water is wayed slightly higher then caldron. No idea why that comes out that way. It could be that Im completely wrong about the equation but the only way to figure that out would be to put in more numbers. It could also be that I'm still softcaping it. Havent confirmed my theory about dex sewing. Still need a high crafter to make a GB with the mats Iv prepped. Gonna ask one of the traders to help me out. All I would ask would be if people could try it out so it could be verified by at least 3-4 people before we added it to wiki.

One thing I should add is that water and caldron could just as easily be wayed equally. But in some rare instances the numbers wont match. And it might be that the water and caldron might be as suggested by Potjeh calculated in ways I dono about. The numbers also match if Caldron Water is matched 5:6, 6:7, 7:8, 8:9. Any of those work as long as water is slightly higher wayed then caldron.

[Sevenless]

After making some crazy stuff in excel I figured out an equation but not sure if its correct.

Qgoldbeater = ( ( water * 9 + caldron * 8 ) / 17 + ( ( intest * 2 + vinegar ) / 3 ) * 5 ) / 6

Q of water is wayed slightly higher then caldron. No idea why that comes out that way. It could be that Im completely wrong about the equation but the only way to figure that out would be to put in more numbers. It could also be that I'm still softcaping it. Havent confirmed my theory about dex sewing. Still need a high crafter to make a GB with the mats Iv prepped. Gonna ask one of the traders to help me out. All I would ask would be if people could try it out so it could be verified by at least 3-4 people before we added it to wiki.

One thing I should add is that water and caldron could just as easily be wayed equally. But in some rare instances the numbers wont match. And it might be that the water and caldron might be as suggested by Potjeh calculated in ways I dono about. The numbers also match if Caldron Water is matched 5:6, 6:7, 7:8, 8:9. Any of those work as long as water is slightly higher wayed then caldron.

Part of the reason that's a little wonky would be the influence of the cauldron. As far as we've ever known, it's never directly influenced the quality. It only lowers the quality if the water is higher Q than the cauldron (possibly anyway). So in the case where the water is higher than the cauldron, you get a slightly lower Q (which is why water is weighted slightly heavier). Excel can't handle that type of thing very easily I don't think.

As for formulas, there's a certain simplicity to how Jorbtar design them usually. I think you're getting close but it needs a little more refinement.

If you want to avoid this issue, only test with water below your cauldron quality.

[mvgulik]

Bummer. Seemed close, but no match. O well.

Code: Select all

Water Caldron Vinegar Intests Goldbeater ...calc... Diff
10 82 140 43 70 70.000 0
10 82 140 44 70 70.545 0
30 11 140 44 66 65.909 -1
30 82 21 43 38 39.364 1
30 82 21 44 39 39.909 0
30 82 80 44 55 56.000 1
30 82 140 36 67 68.000 1
30 82 140 43 71 71.818 0
30 82 140 44 72 72.364 0
30 82 140 50 75 75.636 0
30 82 140 53 77 77.273 0
60 82 140 44 75 75.091 0
39 78 114 34 59 60.273 1


... Guess we have to wait for MagicManICT obvious and simple solution. (hopefully with a ref link to the source.) ...

[Xcom]

Part of the reason that's a little wonky would be the influence of the cauldron. As far as we've ever known, it's never directly influenced the quality. It only lowers the quality if the water is higher Q than the cauldron (possibly anyway). So in the case where the water is higher than the cauldron, you get a slightly lower Q (which is why water is weighted slightly heavier). Excel can't handle that type of thing very easily I don't think.

As for formulas, there's a certain simplicity to how Jorbtar design them usually. I think you're getting close but it needs a little more refinement.

If you want to avoid this issue, only test with water below your cauldron quality.

Just looking at the caldron and water I'm starting to become convinced as you mentioned that it acts similar to herbalist tables. The caldron and water most likely act similar to the equation.

Q CoW = int( (Caldron + Water) / 2)

Just because the Q82 caldron and the Q30 water gives Q33 cooked mussels using Q10 raw mussels. Similarly the HL equation has a (Caldron + Water) / 2. I'm sure if I use the caldron in other recipes I can verify this.

[Xcom]

Bummer. Seemed close, but no match. O well.

Code: Select all

Water Caldron Vinegar Intests Goldbeater ...calc... Diff
10 82 140 43 70 70.000 0
10 82 140 44 70 70.545 0
30 11 140 44 66 65.909 -1
30 82 21 43 38 39.364 1
30 82 21 44 39 39.909 0
30 82 80 44 55 56.000 1
30 82 140 36 67 68.000 1
30 82 140 43 71 71.818 0
30 82 140 44 72 72.364 0
30 82 140 50 75 75.636 0
30 82 140 53 77 77.273 0
60 82 140 44 75 75.091 0
39 78 114 34 59 60.273 1


... Guess we have to wait for MagicManICT obvious and simple solution. (hopefully with a ref link to the source.) ...

This is whats bothering me. This particular calculation.
30 82 80 44 55 56.000 -1

Assuming Q of Caldron and Water discussed above is equated (82 + 30 ) / 2 = 56 and
assuming the 2nd part is ( intest * 2 + vinegar ) / 3 as it fits in with most of the calculations you get
int and vin = ( 44 * 2 + 140 ) / 3 = 56. Both comes out as 56, that means that you don't need to bother figuring out the relation between the CoW and Intest&Vin as they are equal. But why the result doesn't come out 56 is oddly strange and what throws me of every other equation thereafter.

Just because of this I made a new test:
Water Caldron Vinegar Intest Goldbeater Calulation Difference
22 82 50 53 51 52.000 -1

Same problem here
CoW = 52
Int&Vin = 52

Same issue here. The proportionality of 2 are irrelevant so it must be something to do with Int&Vin. I assumed it was the Water and Caldron that was acting up. If you alter the relation between intest and vinegar you mess up everything else, issue is that relation between them are more then likely correct so it forcefully leaves 2 other factors. Softcap issues or that caldron and water relation is not correctly assumed.

This is absolutely insane as it cant be a softcap issue or 2 different results 4Qs apart wouldn't not have softcaped by only 1Q. If in the first example would give 1Q drop your softcaping skill would have had to be around 54 as a minimum, 54 + 56 / 2 = 55. But with 54 softcap skill your drop in 1Q in the 2nd example wouldn't be explainable as 54 softcap wouldn't affect Q52 goldbeater.

The only other leftover explainable fact would be the caldron water equation. But it makes little sense if every other instance the caldron water relation is known as caldron + water / 2 leaving no more explainable quirks to why the drop in Q happens. It might be that there are unknown constants added somewhere in the equation like maybe (vin + intest*2 -1) / 3, even worse Jorb might have added a sin function in there to throw of nosy people that try to figure out the equation.


Edit: Just as I wrote that I thought why not just try that out and throw in a -1 in there and see what happens. Everything came out correctly. It just might be the right equation.

Qgoldbeater = ( int( (water + caldron) / 2 ) + ( (vinegar + intest * 2 - 1) / 3 ) *5 ) / 6

Edit 2: Actually makes sense except in this one instance.
60 82 140 44 75 74.94 -1
I'm really starting to go mad. 2days and I cant eat sleep or drink and I still cant figure out the damn formula. Evil contemplations of way to breaking in and steeling the server is starting to grow strong in my mind.

[mvgulik]

Yea, its kinda mystifying.

PS: Im using following item relation values (seems easier this way). (with, yes. CoW=unconditional averaging.)
CoW Vin Int
2 3 6 (minimized, per set)
or
4 3 4 (minimized, per item)

PS: how much water is used/needed for producing a goldbeater?

(Will play around with is some more, after giving it a little rest first.)