Welcome Guest ( Log In | Register )

2 Pages V < 1 2  
Closed TopicStart new topic
> Forging Thread, Let's crack that formula!

 
post Jun 26 2013, 20:07
Post #21
hitokiri84



Legendary Poster
***********
Group: Gold Star Club
Posts: 10,945
Joined: 24-December 07
Level 496 (Godslayer)


QUOTE(skillchip @ Jun 20 2013, 20:30) *
CODE

1    0.018888889
2    0.036666667
3     0.052222222
4     0.067222222
5     0.081111111
6     0.093888889
7    0.106111111
8    0.117222222
9    0.128333333
10    0.138333333
11    0.148333333
12    0.157222222
13    0.166111111
14    0.175
15    0.182777778
16    0.190555556
17    0.198333333
18    0.205555556
19    0.212777778
    
39    0.321645303
50     0.354


I'm at a standstill with it currently (busy), if anyone would like to pick it up.

Not sure about those 39 and 50 points. They don't fit the curve I'm working on. 50 for sure since I still have the spreadsheet of my staff data, and I have two stats at 50. Looks to be about 0.357825. Of course there is some minor rounding error since we can only see 2 decimal places, and the new(er) fluctuations are assumed to be 100% accurate (still possible minor error.) Also, I have 24, 36, 44, 45, and 46. The curve for non-damage upgrades is looking like a high order polynomial and not a log function.

Lastly, for the ADB/MDB upgrades that go from 51-100, I only have the 100 point, but it doesn't fit the 1-50 non damage upgrade curve. So damage upgrades either have a separate function, or there could possibly be one piecewise function for 1-50 and 51-100 to cover everything.

This post has been edited by hitokiri84: Jun 26 2013, 20:30
User is offlineProfile CardPM
Go to the top of the page
+Quote Post

 
post Jun 27 2013, 02:15
Post #22
skillchip



Mathematical Trashcat
*********
Group: Catgirl Camarilla
Posts: 5,754
Joined: 31-December 06
Level 497 (Godslayer)


QUOTE(hitokiri84 @ Jun 26 2013, 18:07) *

Not sure about those 39 and 50 points. They don't fit the curve I'm working on. 50 for sure since I still have the spreadsheet of my staff data, and I have two stats at 50. Looks to be about 0.357825. Of course there is some minor rounding error since we can only see 2 decimal places, and the new(er) fluctuations are assumed to be 100% accurate (still possible minor error.) Also, I have 24, 36, 44, 45, and 46. The curve for non-damage upgrades is looking like a high order polynomial and not a log function.

Lastly, for the ADB/MDB upgrades that go from 51-100, I only have the 100 point, but it doesn't fit the 1-50 non damage upgrade curve. So damage upgrades either have a separate function, or there could possibly be one piecewise function for 1-50 and 51-100 to cover everything.


Tenboro said that 50 is 35.4% on the IRC

as for the 39 and 50 points those were from chosenuno's gear, and may be a bit off

if the two curves are similar, the damage one has different variable values

This post has been edited by skillchip: Jun 27 2013, 02:16
User is offlineProfile CardPM
Go to the top of the page
+Quote Post

 
post Jun 27 2013, 04:59
Post #23
hitokiri84



Legendary Poster
***********
Group: Gold Star Club
Posts: 10,945
Joined: 24-December 07
Level 496 (Godslayer)


QUOTE(skillchip @ Jun 26 2013, 19:15) *

Tenboro said that 50 is 35.4% on the IRC

Since when does he give straight answers when it comes to stats?

He might have said that, but that's not what my numbers are showing. Do you have links to any equipment with 50 on non-damage? Such as chosenuno's? Either he misspoke trying to recall it from memory, or the fluctuations for both EDB and Elemental Proficiency are wrong. And since the fluctuations are easily verifiable with a non-forged equipment, I'm inclined to believe it's the former.

Or it's a discontinuous piecewise function @ 50 and he's screwing with you.
User is offlineProfile CardPM
Go to the top of the page
+Quote Post

 
post Jun 27 2013, 15:54
Post #24
ChosenUno



A Legendary Collector
********
Group: Gold Star Club
Posts: 4,170
Joined: 23-February 10
Level 422 (Godslayer)


QUOTE(hitokiri84 @ Jun 27 2013, 09:59) *

Since when does he give straight answers when it comes to stats?

He might have said that, but that's not what my numbers are showing. Do you have links to any equipment with 50 on non-damage? Such as chosenuno's? Either he misspoke trying to recall it from memory, or the fluctuations for both EDB and Elemental Proficiency are wrong. And since the fluctuations are easily verifiable with a non-forged equipment, I'm inclined to believe it's the former.

Or it's a discontinuous piecewise function @ 50 and he's screwing with you.


Here's my staff with all the relevant stats @ 50:
Legendary Ethereal Katalox Staff of Heaven-sent
  • Quality: 0.977
  • Base rolls:
    • MDB - 30
    • Holy EDB: 9
    • Holy prof: 34
    • Intelligence: 11
    • Wisdom: 10
  • Factors:
    • MDB: 0.74969
    • Holy EDB: 0.299
    • Holy prof: 0.306
    • Intelligence: 0.29999
    • Wisdom: 0.29999

These values were accurate 100% when I tested them before this patch, using HentaiVerse Equipment Comparison for level scaling.
User is offlineProfile CardPM
Go to the top of the page
+Quote Post

 
post Jun 27 2013, 19:31
Post #25
hitokiri84



Legendary Poster
***********
Group: Gold Star Club
Posts: 10,945
Joined: 24-December 07
Level 496 (Godslayer)


QUOTE(ChosenUno @ Jun 27 2013, 08:54) *

Here's my staff with all the relevant stats @ 50:
Legendary Ethereal Katalox Staff of Heaven-sent
  • Quality: 0.977
  • Base rolls:
    • MDB - 30 36
    • Holy EDB: 9 11
    • Holy prof: 34 41
    • Intelligence: 11 14
    • Wisdom: 10 12
  • Factors:
    • MDB: 0.74969 0.82969
    • Holy EDB: 0.299 0.804
    • Holy prof: 0.306
    • Intelligence: 0.29999
    • Wisdom: 0.29999
These values were accurate 100% when I tested them before this patch, using HentaiVerse Equipment Comparison for level scaling.

First, MDB is 0.82969 now, and EDB is 0.804. The level scaling factor for EDB was also changed from 50 to 200. The quality tier bonuses are applied to base roll and always round up, so it's easier just to think in terms of base roll with it already factored in (for me at least.)

All of the non-damage 50s show a forge factor of 0.3576-0.3581, as does my staff's 50s. The variance is due to a percent error caused by the rounding of stats to 2 decimal places, causing error in the determination of the fluctuations and in the forge data collection.

The MDB forge factor at 100 shows about 0.6705, almost exactly the same as my mine.

This post has been edited by hitokiri84: Jun 27 2013, 20:44
User is offlineProfile CardPM
Go to the top of the page
+Quote Post

 
post Jul 11 2013, 11:10
Post #26
PK678353



Old Marshal Winter
*******
Group: Gold Star Club
Posts: 1,454
Joined: 7-November 10
Level 345 (Godslayer)


Have this data, may as well drop it. Improvement has had quality factored out (though rounding may cause issues on Fire given how small it was).


0 forges
Fire Scaled: 7.23
Fire Base: 3.00
Base Roll 3, Quality 0.732

Supportive Scaled: 75.49
Supportive Base: 8.49
Base Roll 27, Quality 0.732

1 forge
Fire Scaled: 7.34
Fire Base: 3.05
Fire Improvement: 1.89%
Supportive Scaled: 76.87
Supportive Base: 8.64
Supportive Improvement: 1.88%


2 forge
Fire Scaled: 7.44
Fire Base: 3.09
Fire Improvement: 3.61%
Supportive Scaled: 78.15
Supportive Base: 8.78
Supportive Improvement: 3.62%

3 forge
Fire Scaled: 7.54
Fire Base: 3.13
Fire Improvement: 5.33%
Supportive Scaled: 79.32
Supportive Base: 8.92
Supportive Improvement: 5.21%

4 forge
Fire Scaled: 7.62
Fire Base: 3.16
Fire Improvement: 6.71%
Supportive Scaled: 80.41
Supportive Base: 9.04
Supportive Improvement: 6.69%

This post has been edited by PK678353: Jul 11 2013, 11:20
User is offlineProfile CardPM
Go to the top of the page
+Quote Post

 
post Jul 11 2013, 18:39
Post #27
hitokiri84



Legendary Poster
***********
Group: Gold Star Club
Posts: 10,945
Joined: 24-December 07
Level 496 (Godslayer)


QUOTE(PK678353 @ Jul 11 2013, 04:10) *
though rounding may cause issues on Fire given how small it was

Proficiency on a proficiency suffixed staff would be best. For the 100 point scale (ADB/MDB), a high quality "of the Slaughter" piece would work.

To anyone forging things like that, you just need to post a link to your equip, write down what the stat is at each forging, and what your HV level is while doing each forge.
User is offlineProfile CardPM
Go to the top of the page
+Quote Post

 
post Jul 11 2013, 22:49
Post #28
PK678353



Old Marshal Winter
*******
Group: Gold Star Club
Posts: 1,454
Joined: 7-November 10
Level 345 (Godslayer)


QUOTE(hitokiri84 @ Jul 11 2013, 11:39) *

Proficiency on a proficiency suffixed staff would be best. For the 100 point scale (ADB/MDB), a high quality "of the Slaughter" piece would work.

To anyone forging things like that, you just need to post a link to your equip, write down what the stat is at each forging, and what your HV level is while doing each forge.


Level 282 when I did that, on a Fine Oak Staff of Focus. Though honestly, I picked Fire and Supportive to double check if forging affected the whole Derived Base, or just the non-quality part of it (~20% of the entire Fire Derived Base comes from Quality, whereas ~3% of the Supportive Derived Base comes from Quality).

I've got a good Heaven-Sent Katalox for testing, but it wants HG after 6 forges. . . (IMG:[invalid] style_emoticons/default/dry.gif)

This post has been edited by PK678353: Jul 11 2013, 22:50
User is offlineProfile CardPM
Go to the top of the page
+Quote Post

 
post Sep 8 2013, 13:10
Post #29
varst



Peerless perverted long-hair-con of the Hentaiverse
***********
Group: Gold Star Club
Posts: 11,561
Joined: 30-March 10
Level 478 (Godslayer)


There's no promise on getting any result, but in case anyone got data on forging, you're always welcomed to post in this thread (or contact me if you want to keep that a secret). Right now I only have one selected case on ADB and there's no way to calibrate that.

It's fine if you only get a few data points, I can try to piece them together.

As for reminder, please provide the following.

Level
Equip being forged
PXP(0) if the item has been leveled if possible
Stat being forged
Forgings before -> after
Stat before forging
Stat after forging

In case you're too busy, just capture the equip stats screen.

Stats like ADB,MDB, EDB, prof, would be great. Other stats are also welcome, but those gives the best chance to find something meaningful.

This post has been edited by varst: Sep 8 2013, 13:17
User is offlineProfile CardPM
Go to the top of the page
+Quote Post

 
post Oct 5 2013, 05:03
Post #30
VriskaSerket



Veteran Poster
********
Group: Catgirl Camarilla
Posts: 4,118
Joined: 27-December 08
Level 500 (Godslayer)


forging this
EDB: scaled to 401 lvl.
0 50,8
1 51,72
2 52,57
3 53,34
4 54,04
5 54,72
6 55,34
7 55,92
8 56,48
9 56,99
10 57,49
11 57,95
12 58,41
13 58,85
14 59,26
15 59,65
16 60,03
17 60,39
18 60,73
User is offlineProfile CardPM
Go to the top of the page
+Quote Post

 
post Dec 21 2013, 20:15
Post #31
yobi00



Regular Poster
*****
Group: Members
Posts: 700
Joined: 13-August 13
Level 380 (Godslayer)


I tried playing around with data in posts #19 and #30 and the following formula seemed to work pretty well:
CODE
0.2*ln(0.1*forge_count+1)
but the 39 and 50 points from post #19 seem off.

I only have 3 data points for ADB, but
CODE
0.28*ln(0.1*forge_count+1)
seems to work kind of well.

Anybody has precise data on fully forged items or more ADB/MDB points to tweak the coefficients better?

PS. Here's a nice graph comparing the values. Doesn't tell much though.
Attached Image
User is offlineProfile CardPM
Go to the top of the page
+Quote Post

 
post Dec 23 2013, 05:55
Post #32
varst



Peerless perverted long-hair-con of the Hentaiverse
***********
Group: Gold Star Club
Posts: 11,561
Joined: 30-March 10
Level 478 (Godslayer)


Skill checking with the formula.
User is offlineProfile CardPM
Go to the top of the page
+Quote Post

 
post Dec 26 2013, 01:28
Post #33
skillchip



Mathematical Trashcat
*********
Group: Catgirl Camarilla
Posts: 5,754
Joined: 31-December 06
Level 497 (Godslayer)


QUOTE(varst @ Dec 23 2013, 03:55) *

Skill checking with the formula.


You mean still right?




I got bored.

QUOTE(yobi00 @ Dec 21 2013, 18:15) *

I tried playing around with data in posts #19 and #30 and the following formula seemed to work pretty well:
CODE
0.2*ln(0.1*forge_count+1)
but the 39 and 50 points from post #19 seem off.

I only have 3 data points for ADB, but
CODE
0.28*ln(0.1*forge_count+1)
seems to work kind of well.

Anybody has precise data on fully forged items or more ADB/MDB points to tweak the coefficients better?

PS. Here's a nice graph comparing the values. Doesn't tell much though.
Attached Image



It looks close enough to me on the first one for everything but ADB/MDB. From the random data points I have laying around its off by only 0.01, which is within rounding.

I also got a value of 0.278875 for the ADB/MDB formula using Chosen's Staff and my own Staffs

I say mission accomplished.

If checking future ADB/MDB I suggest you lean heavily towards MDB because the mouseover popup might be generated by the displayed values, not actual values, thus making the whole number values of ADB not precise enough for what we need.

Edit: best part is the formula for damage fits what tenboro told us

This post has been edited by skillchip: Dec 26 2013, 21:21
User is offlineProfile CardPM
Go to the top of the page
+Quote Post

 
post Mar 16 2014, 19:58
Post #34
something



Expressionless Lolicon
*******
Group: Members
Posts: 1,106
Joined: 14-January 07
Level 355 (Godslayer)


i'm still kinda confused on how to actually use the formula.

is it
0.2*ln(0.1*forge_count+1)(base stat) or 0.2*ln(0.1*forge_count+1) + base stat?
User is offlineProfile CardPM
Go to the top of the page
+Quote Post

 
post Mar 17 2014, 16:10
Post #35
yobi00



Regular Poster
*****
Group: Members
Posts: 700
Joined: 13-August 13
Level 380 (Godslayer)


It's been a while, so I might have mixed something up, but the base stat that is displayed in the equipment pop-up of an unforged item is actually composed of:
CODE
( Base_Roll + Suffix_Roll + Quality_Bonus ) * Base_Multiplier
The equipment ranges wiki page has more detail.

The increase from forging works like this:
CODE
[ ( 1 + Forge_Gain ) * ( Base_Roll + Suffix_Roll ) + Quality_Bonus ] * Base_Multiplier
The logarithmic formulas give the gain so just substitute Forge_Gain with "0.2*ln(0.1*forge_count+1)" or "0.28*ln(0.1*forge_count+1)" and you should be good to go (IMG:[invalid] style_emoticons/default/smile.gif)
User is offlineProfile CardPM
Go to the top of the page
+Quote Post


2 Pages V < 1 2
Closed TopicStart new topic
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

 


Lo-Fi Version Time is now: 20th August 2025 - 06:13