Hardened Leather Quality Formula!

[MagicManICT]

@MagicManICT: PS: When was this data collected ?

Date of the post.

[SpakyPD]

Well ..., as there are no other formula variations that are doing better than the first one here. Its getting my vote.

( ( SC((LeatherAvQ + WaxAvQ)/2)*3) + ((WaterQ + CauldronQ)/2) )/4
SC( LeatherAvQ*3/2 + WaxAvQ*3/2 )*1/4 + Water*1/8 + Cauldron*1/8

It should be noted, that when the Leather and Wax goes through softcapping, you have to truncate any decimals. Otherwise you are going to get false results.

For example on data set 43:
(16+24.5)/2 = 20.25
(20.25 *3) + (60+27)/2 = 104.25/4 = 26
This is incorrect, as it should be 25.

(16+24.5)/2 = 20
(20*3) + (60+27)/2 = 103.5/4 = 25

[Xcom]

I haven't looked into the hardened leather formula but as far as I know Loftar uses a very odd rounding system on the server. Both in the goldbeater and the hitbox calculation. Goes to say that probably the same system applies here. If the Q is off by 1 it simply means your right about the equation and also means you shouldn't try and find the exact rounding system he uses.

I tried for 2 months straight trying to figure out the goldbeater rounding and just about the same amount of time trying figure out how hitboxes rounds. The equation he uses is not a simple round down or roundup. If you want to know how he does it you simply would need to look into his code which we dont have access to. I even tried to convince people that the equation for the goldbeater was good enough and 1Q off because of rounding errors was close enough, but in the end the equation had to go into the discussion of the wiki cause it was off by 1 and people weren't happy that the EXACT equation wasn't found.

Just stick that semi accurate 1 off rounding equation on the wiki and be happy with it. You will pull your hair out trying to find how he does the exact rounding. Goes to say that its most likely something you wouldn't guess unless you had a HUGE sample size and countless of hours to waste.

Edit: Here is the goldbeater thread with that same odd 1Q round down problem.
http://www.havenandhearth.com/forum/viewtopic.php?f=2&t=27144
The wiki got tons of incorrect info without proof, or outdated info. Being anal about 1Q is just counter productive and very few people look in the discussion part of the wiki for the info. Just stick it in there and add a description on how server rounds the Q down on some instances.

[mvgulik]

@MagicManICT: PS: When was this data collected ?

Date of the post.

Roger.

[mvgulik]

Goes to say that its most likely something you wouldn't guess unless you had a HUGE sample size and countless of hours to waste.

Smart sampling best way to limit the needed samples. Current sample set kinda limited in covering potential formula aspects.

[mvgulik]

( ( SC((LeatherAvQ + WaxAvQ)/2)*3) + ((WaterQ + CauldronQ)/2) )/4
SC( LeatherAvQ*3/2 + WaxAvQ*3/2 )*1/4 + Water*1/8 + Cauldron*1/8

It should be noted, that when the Leather and Wax goes through softcapping, you have to truncate any decimals. Otherwise you are going to get false results.

Mmm, I forgot to implement that part in the calculation.

New graph based on this first formula. (its the way the calculation was setup.)
[strike]TRUNC(SC_TRUNC( LeatherAvQ*3/2 + WaxAvQ*3/2 )*1/4 + Water*1/8 + Cauldron*1/8)[/strike]
TRUNC( SC_TRUNC(LeatherQtt/6 + WaxAvQtt/4)*3/4 + WaterQ/8 + CauldronQ/8 )

I'm currently not sure if this second formula would generate the exact same results with the added truncation.
TRUNC( ( SC_TRUNC((LeatherQavg + WaxQavg)/2)*3 + (WaterQ + CauldronQ)/2 )/4 )

SC = SQRT(SEW*DEX)

http://puu.sh/4snd3.png


Mismatch cases:
- [1,2] -> Clay cauldron.
- [12..15] -> W6, Assuming results based on a different/older formula.
- [39,40] -> Clay cauldron.

Cases with unspecified cauldron types added, as Metal.
http://puu.sh/4sok1.png


Only case sticking out as a sore thumb is case 21(from OP) ... lets assume a Clay cauldron was used here, and we got a 100% fit for Metal cauldrons.

[mvgulik]

Erm. Is that Zero cauldron quality for real ?

That was a nod to statements I've read that clay cauldrons have no quality, or, they halve the quality of the water for purposes of calculation. In either repsect, for this topic it is the same as saying Qcauldron = 0.

Agreed. See my previous post in this thread using a clay cauldron.

Note:

Clay Cauldron testing.

SEW*DEX = 28 (21*40)
Leather Q22
Wax Q24
Water Q60

Cauldron Q10 = Hleather Q21

Cauldron Q29 = Hleather Q22

This is why data should be pure and not tainted/changes based on personal believes.

[MagicManICT]

What personal beliefs would have an effect on this process/formula?

[mvgulik]

Not sure why you are replying the way you are, as its not targeting the main point of my statement.

... (reported) data ...

[mvgulik]

Updated Hardened_Leather RoB wiki page.

... Over & Out ...