Silkroad Online Forums

^ You test is just about 3 builds different. We cant see the evolution of the multiplier when adding 1 int or 1 str.

You didn't understand. Bye

well
Here I tried with my new char on venus :
Char lvl : 21, str: 75, INT: 70,
weapon phy attack : 155
weapon mag attack : 224
weapon phy reinforce : 68.7%
weapon mag reinforce : 98.4%
base phy attack (displayed) : 246
base mag attack (displayed) : 292
phy balance ; 77% , mag balance :62%
weapon mastery lvl : 19, imbue mastery lvl : 19
Fire phy attack passive : 1%
Avatar : 5% damage

Skills (all without imbue)
Soul spear move(lvl 3) : 62(+250%) ===> 832 damge
Dancing demon spear (lvl 5) : 62(200%) ====> 666 damage
Ghost spear petal (lvl 1) : 65(220%) ====> 741 damage
wolf bite spear (lvl : 65(160%) ====> 538 damage

you can make your calculates, when char lvl up i'll re-do the test and see difference, then will add str and int points and re-do again to see difference, Isn't that what you need??

Now there is something that isnt working. You said that you have base attacks of 246 but according to the formula and doing the lower value rounding that silkroad made on C char it should be 245. Can you check that?

The rounding and approximation are not very important. We got the multiplier 1,276772606 is close enough with the one gives real dmg.
That multiplier is dependent on Int & Str. That's why i told you to test different int/str case to see how it grows proportionally.

I think the multiplier change depends on difference between char lvl and monster lvl(or between skill lvl and monster lvl) , I have noticed that on Abslute damge skills , although it's supossed to do same damge for everyone it still do less damage on higher lvl ppl and monsters (like a bit less every degree or every 10 lvls not sure exactly)

Multiplier is like 1.29 ~ 1.27 I think depends on lvl difference ,

Btw my build is same as Tomiator has (hybrid 1:1) , if the multipier affected by str or int or balance that's easy to test , Tomiator can strip his set (which have much int/str increase) and may be even his avatar, he will get balances almost as me, and re-try your damage and calculate it, if multiplier stayed same then it's not related to str nor int nor balance and will confirm the theorey about lvls, if multiplier changed then you can get your answer nuclear..

And well, my char now is lvl 22 with free 3 stats points , u think if I did test before adding them then re-test after adding we will get different multiplier??

well I also wanted to ask for the phy and mag balance accurate formula as I see displayed balance not accurate enough and can't get the exact multiplier without know the accurat numbers..

I'm sure it depends on int/str, but that's the purpose of what we're going to do.
It also depends on the difference of level. The graph Tomiotar did is quite informative, it shows that the multiplier decreases by level with a polynomial of 3rd degree.
(and strangely, the center is at lvl 45 where 45=90/2)


Or simply a line with equation Y=C*X where C is a constant, because somehow the parry ratio on high lvl didnt let him reach his real maximum dmg, or the difference of (attack rating - parry ratio) is a factor that limit the real maximum dmg.

Needed to test:
- The maximum dmg deal on a lvl 90 mob
- The maximum dmg deal on a lvl 90 mob under electrical shocked (divide parry by half)

Absolute dmg:
It's actually a bit dependent on the enemy's defense. Why i said that? Because i once tested that skill. Absolute a glaiver lvl 3x (when i was 3x bard). The absolute dmg worked normally. But after that a warrior gave him phy and mag fence and the absolute dmg decreased like 30%.
If it's done to a high lvl, dmg is divide by 10.
min muliplier = 0.1, max multiplier = 1

I retruned to these old pages and re-calculated it and found where were the problem,




Notice this char were the only one that have difference between weapon and imbue lvl (weapon lvl 7 and imbue lvl 26)

Now look :


let's give it a new try so with new theory
Int: real = 542
min: dmg = 107 + ((332 + 49) * 1.07 - 10) * 0.90 * 0.84 * 1.00 * 1.29 = 494
max: dmg = 109 + ((333 + 91) * 1.07 - 10) * 0.91 * 0.84 * 1.00 * 1.29 = 546

It works now lol, that's other decent prove that imbue mastery lvl have no effect at all and the thing that affect is only skill mastery lvl (weapon at skills and force(light,cold,fire) at nukes ..

well to make more sure I used the formula again with other number I got in that test before

Phys = (base + (skill_pow * mastery_incr) - Phys def) * balance * skill_mult * buff&passive * multiplier
= (181+ (22 * 1.07) - 7) * 0.57 * 1.5 *1.276 = 215.5 (Real 218)

All = phy + ((base + imbue_pow )* mastery_incr - Mag def) * balance * skill_mult * buff&passive * multiplier
= 215.5 + ((332 + 91) * 1.07 - 10) * 0.9 * 1.5 * 1.287 = 984.5

I used Tomiator multiplier here and numbers are close (not accurate beacuse attack and balance rounded and also char still low lvl and sound that low lvls (different builds) char have a bit higher multiplier than higher lvls when it comes for test on mangyang

Now after fixing this pure int thing I think the general formula works (In mangyang test) for all builds (we tried bow and glavie and I doubt it will make any diference with blade) , all lvls, different balances, the only thing that change is the multiplier between 1.27 and 1.29 , At worst we can use the idea of range (min/max multiplier) , let's focus and concentrate on find the variable that indicate the multiplier for mangyang for now, other higher lvls monsters might have something like damge absorb or w/e

@ absolute , I asked some high lvl before to totaly strip his set and accesories and tried the absolute and kept giving same number as it give when he has all set and accesories !

If you still believe on it's depends on str/int and wanna make that test u said come help me do it, as I told u, I has caped wiz/bard that can plvl very well but I has no 2nd pc for do it (and i'm not using illagal clients or loaders or w/e ppl use for plvling) , and friends not care about the damage formula as 99% of legits in venus is euro and euro always made as pure build beside JM turned most Variables into constants at euro race, weapon has only mag or phy attack , all phy weapons same lvls and same whites has same stats also mag weapons, all phy skills same lvl has same attack , even mastries made to be 2 only and maxed both , and all builds have 4% attack passive every 4 lvls , so if you look at the formula to compare euro builds u will find:

Phys = (base + (skill_pow * mastery_incr) - Phys def) * balance * skill_mult * buff&passive * multiplier
So only things that change the final damge is : how good is ur weapon , and how many str/int points in ur set, and skill_mult(%) so compare damage every build do will not take longer than 5 sec to look at skills bonus at Skills page , ppl realised this easily with no need for formula or anything , there is no unique builds that can be disgned as euro , the unique part you can make as euro is to customize the party formation and arranage buffs , nothing more

Edit : about parry/attack rate thing, if it make difference at high lvls it should make difference at low lvls too (things higher than mangyang), even very small difference should be noticable so u think trying with this char (lvl 22) can give the result ?

aha, but it worked only once on pure int doesnt mean it will work on general case, still need more test to prove that point.


Yeah, it should work. But it's Farking slow to get the max dmg when testing on mob same/high lvl, u need a lot of tests, unlike mangyang.

The best way is to test on mob with same parry ratio as your attack rating. Normally a mob same lvl as yours always has less parry than your weapon.

Att rate/ parry rate table : viewtopic.php?f=2&t=64659

On the test I made when I prepare that graphic Im sure that I've use my highest damage. Making the graph for monster at each level was really hard because I had to stay there lot of time to be sure that I got my highest dmg on each monster (but Im sure that I got it). I still have the lvl31 pacheon char with no masteries added, but I didnt connect to the game on las week so I couldn't check nor level it to 32 (I do have 2 pc so lvl a char from 31 to 32 with 0 gap it will not take me much time).

I deffinetly believe that att rate/parry isn't a big problem for me against monsters, because I do have a good attack rating on my bow (white stat on bow is 71%)+ blue hawk + pacheon passive + 10% on avatar and I can use +hit scroll to make the graphic (tons of this scrolls saved), so I can finally reach values near 600 attack rating (165 parry for rockys).

Is it because that your absolute skill is too strong or same lvl as the one you test with?
I think this way:
when abs dmg >= mag def => full dmg
when abs dmg < mag def => little reduce
when abs dmg << mag def => dmg/10

Still no one giving the balance formula?
I went to test again at lvl 22 (without adding stats points)
then re-tried after adding whole 3 points into str

from these tests I can find out the accurate multiplier for my char and notice the changes that happen to it with adding str or int , give me the formula so it become easier than posting test results and wait 1~2 days every time to give tomiator time to see, calcluate and post , this will save much time

char now is lvl 23 without added the 3 points again ready to do new test at this lvl but I don't wanna do till finish calculating the previous lvl that still need the balance formula

any help ?

maxstat = 28 + lvl * 4
int_balance = 100 * int / maxstat
str_balance = 100 - ( 100 * 2/3 * ( maxstat - str ) / maxstat )

K I will contribute some data too
Lv. 60
Mag. atk 1347 - 1530
Mag. balance 114%
Fire 51 Light 51 Ice 51 Bicheon 30
Things to count in: 5% damage from hallowen suit no other buffs
(5% applies to final damage and is not somewhere else in formula I also tested without suit and my damage was 1.05 lower )

lv.6 Fire wave arrow 198 - 318 (250%)
damage on mangyang no imbue used: 10727
damage on CTF flag no imbue used: 10873

lv.9 Shock Lion Shout 106 - 196 (100%)
damage on mangyang no imbue used: 4006
damage on CTF flag no imbue used: 4062

All max damage

U are using pure mag attack (nuke) , dun no if it works da same.
Using tomiotar's multiplier:
on mangyang : (1530 + 318)*1.51 - 10) * 1.15 * 2.5 * 1.05 * 1.287004542 = 10 802 (0.7% error)
on flag : (1530 + 318)*1.51 - 0) * 1.15 * 2.5 * 1.05 * 1.287004542 = 10841 (0.3% error)

shout/mangyang: (1530 + 196)*1.51 - 10) * 1.15 * 1.0 * 1.05 * 1.287004542 = 4034 (0.7% error)
shout/flag: (1530 + 196)*1.51 - 0) * 1.15 * 1.0 * 1.05 * 1.287004542 = 4050 (0.3% error)

The right multiplier:
mangyang: 1.27800148
ctf flag: 1.29075354

But im aware that it wont give the correct result since they are different objects and you got different build than tomiotar's which is hybrid.
In another word, a 10 mag def reduce 0.01275206 off the mag multiplier. And since your are pure int, your mag balance is higher than tomiotar's, it makes sense when your mag multiplier is higher than his.

But atm, finding the multiplier is not the point, but understand how it changes with the build and mob defense. Till now, we only got random tests, no one could give me a statistic table int/str/multiplier from 20 int/str to 100 int/str yet.


Edit: testes the attack rate/ parry rate thingy. It indeed doesn't increase the dmg. But i notice that only 40 rate in difference, it gives a lot more chance to deal max dmg (~ 4x more chance). So, dont under estimated the Concentration skill of Lightning (as well as pacheon skill).

Why did u use 1.15

And what exactly is 1,287004542 ?

damn, i thought it's 115% balance

Can you tell us how many ints points do you have or use this formula:

maxstat = 28 + lvl * 4
int_balance = 100 * int / maxstat

to have less error on the calculation?

The 1,287004542 value is that we call the "multiplier", and to find the true nature of multiplier is last step to get the final formula. At the moment we just find it by testing but we dont have the true expression.

Sry I am bit late ...

So if we use the balance formula we get that your mag balance is 114,179, so now we applied the formula again:

on mangyang : ((1530 + 318)*1.51 - 10) * 1.14179 * 2.5 * 1.05 * 1.287004542 = 10 725,4
on flag : (1530 + 318)*1.51 - 0) * 1.14179 * 2.5 * 1.05 * 1.287004542 = 10 764,0

shout/mangyang: ((1530 + 196)*1.51 - 10) * 1.14179 * 1.0 * 1.05 * 1.287004542 = 4005,9
shout/flag: ((1530 + 196)*1.51 - 0) * 1.14179 * 1.0 * 1.05 * 1.287004542 = 4021,4


So the formula works really good on the mangyan (10725,4 predicted vs 10727 real and 4005,9 predicted vs 4006 real), but it fail considerably againts the **** pole. I have no idea how it works damage on the flag pole or if we have to take into account just mag def= 0 or something else, but I will not worry much about the flag pole considering that is inmortal from the beginning .

Ok here we go


Now trying at lvl 22 before adding stats :

Soul spear move(lvl 3) : 62(+250%) ===> 824 damge
Phys = (246.58228+ 62 * 1.19 - 7) * 0,77011 * 2,5 * 1.06 * 1.289092 = 824.3

Dancing demon spear (lvl 5) : 62(200%) ====> 659 damage
Phys = (246.58228+ 62 * 1.19 - 7) * 0,77011 * 2,0 * 1.06 * 1.289092 = 659.5

Ghost spear petal (lvl 1) : 65(220%) ====> 733 damage
Phys = (246.58228+ 65 * 1.19 - 7) * 0,77011 * 2,2 * 1.06 * 1.289092 = 733.7

wolf bite spear (lvl : 65(160%) ====> 533 damage
Phys = (246.58228+ 65 * 1.19 - 7) * 0,77011 * 1.6 * 1.06 * 1.289092 = 533.6

Still working well , now let's add 3 str points and see :

Soul spear move(lvl 3) : 62(+250%) ===> 849 damge
Phys = (249.03487+ 62 * 1.19 - 7) * 0.78735* 2,5 * 1.06 * 1.289092 = 849.4

Dancing demon spear (lvl 5) : 62(200%) ====> 679 damage
Phys = (249.03487+ 62 * 1.19 - 7) * 0.78735* 2,0 * 1.06 * 1.289092 = 879.5

Ghost spear petal (lvl 1) : 65(220%) ====> 755 damage
Phys = (249.03487+ 65 * 1.19 - 7) * 0.78735* 2,2 * 1.06 * 1.289092 = 755.9

wolf bite spear (lvl : 65(160%) ====> 549 damage
Phys = (249.03487+ 65 * 1.19 - 7) * 0.78735* 1.6 * 1.06 * 1.289092 = 549.7

Multiplier still working

uhm, k
Let's do another test when you level up then. I don't believe it's just a constant works with all char.


Hawk dmg: (hawk_power * (1.pacheon_mastery) - mob_def) * 0.7752 * 1.276772606
where 0.7752 is hawk's attack multiplier (approximation)

Tested the mangyang multiplier on my bow char, doesnt work

here we go
lvl 25 , str 82 (9 stat points left)

Soul spear move(lvl 3) : 62(+250%) ===> 826 damge
Phys = (251.48746+ 62 * 1.19 - 7) * 0.7604166* 2,5 * 1.06 * 1.289092 = 826.7

Dancing demon spear (lvl 5) : 62(200%) ====> 661 damage
Phys = (251.48746+ 62 * 1.19 - 7) * 0.7604166* 2,0 * 1.06 * 1.289092 = 661.4

Ghost spear petal (lvl 1) : 65(220%) ====> 735 damage
Phys = (251.48746+ 65 * 1.19 - 7) * 0.7604166* 2,2 * 1.06 * 1.289092 = 735.7

wolf bite spear (lvl : 65(160%) ====> 535 damage
Phys = (251.48746+ 65 * 1.19 - 7) * 0.7604166* 1.6 * 1.06 * 1.289092 = 535

Still working
Now let's add more 3 str stats : str 85

Soul spear move(lvl 3) : 62(+250%) ===> 850 damge
Phys = (253.94005+ 62 * 1.19 - 7) * 0.776041667* 2,5 * 1.06 * 1.289092 = 850.2

Dancing demon spear (lvl 5) : 62(200%) ====> 680 damage
Phys = (253.94005+ 62 * 1.19 - 7) * 0.776041667* 2,0 * 1.06 * 1.289092 = 680.2

Ghost spear petal (lvl 1) : 65(220%) ====> 756 damage
Phys = (253.94005+ 65 * 1.19 - 7) * 0.776041667* 2,2 * 1.06 * 1.289092 = 756.5

wolf bite spear (lvl : 65(160%) ====> 550 damage
Phys = (253.94005+ 65 * 1.19 - 7) * 0.776041667* 1.6 * 1.06 * 1.289092 = 550.2

works too , there were mag part of test but i feel lazy to post

ok, thx for sharing, that's enough.
If int/str don't involved in that multiplier, what could it be then? Weapon
You don't have the same multiplier as tomiotar's.

a few days ago, I went to a math professor who is also game designer and I asked him how to solve the formula to calculate the damage done in a game.

he told me I'd have to use PDE (partial differential equations, no idea what it was), and that I'd have to solve a differential equation something like f(x,y,z...) = dx + dy + dz ... for all the variables that would affect the damage. He said if we can like figure out how the change in damage changes with respect to all the dynamic variables etc, then we should get a working formula that is accurate AND universal that works for every mobs and every class, chinese and euro....

He also gave me a PDF copy of a PDE book so we can learn how to do it. But the subject seems too difficult for me. So I've attached the book so hopefully there will be a smart whiz here who can figure out and solve the formula for dmg.

PDE Adobe Acrobat PDF book: http://www.yousendit.com/download/Y2ovU ... NitGa1E9PQ

i dont believe we are currently on the right track. the damage formula isn't as simple as


i'd don't think we'd get a universal damage formula from whacking mangyangs. We might get a formula that will work for mangyangs, but it won't be universal and works for all mobs and people. It will work only for mangyangs. do you guys see what im seeing? we can start by studying existing formula already created... NuclearSilo do u happen to have the SP chart formula that d2a created? ifwe can reverse that and see how he did it, then perhaps we will be on the right track.

anyways heres the book again. u need adobe acrobat to open it http://www.yousendit.com/download/Y2ovU ... NitGa1E9PQ

heres the def of PDE from wikipedia



now who has the time and motivation and a brain to read the attached pdf book and finally cracking the secret to the dmg formula?

i'm not pro math, that's your task
Even what variable affect the dmg, we don't know, what can you do?
Don't talk about D2A, he used reverse engineering technic to get all the codes in sro.

I have my engineer title and my PhD so I know what that professor is talking about, but isnt as easy as you said apply partial diffential equation when you have some many variables and no idea of the effect of each variable nor the starting points. You are doubting about this expression to be too simple:



but I already did lots of test and I know that the expression is perfectly accurated on how they work on every factor with the only exception of the multiplier and maybe the defense value. Morover, I do know that this expression (with the right multiplier) it works for every monster from mangyan to rockys, because I DID the tests and I check it changing skills and imbues always with great results. My only 2 doubts are if I have to take defense values form where it is now and include it on multiplier or not. Im not playing that often as I use to do it, thats why Im not finishing working on the formula, but Im sure that if we have to make any change on the formula as we have it now, that change it can be to include def value and maybe NPC accesories of same levl of the monster to find the true expression of multiplier.

With the formula as we have it now and using the graphic and the tables of multiplier that I provide some page ago you can already calculate your damage really accurated for every monster, even when isn't totally finnished.

lol i suck at math also. to find out what variables affect damage, we run some tests by whacking some mobsLOL. i dont think that's something that need to be reverse engineeer but can anyone on this forum capabable of everse engineering also? what is his occupation?

i already gave you all the correct key to unlock the secret door, now it's yall job to wield the key and bring it to the door, insert it in the lock, and open it all the secret is in that book. have u taken a look at it yet?

@tomitar: well it's been 1 year, 1 month since this thread begun and no real progress hopefully this will give a boost to get us on the right path

Ok, we can now make a list of what could affect that multiplier. Since we know that it changes for every char, so it's not purely monster related.

Green = needed to verify, red = doesnt involve:

monster defense (needed to test with warlock/force debuff)
monster attack (sounds stupid, but meh) (need debuff)
monster absorption (unknown value) (notice the ignore defense dmg not the same as mangyang)
hit ratio
parry ratio
int/str
balance (int/str affect balance)
weapon level, degree
weapon attack power, reinforce
weapon type (bow, spea, ...)
character skin, gender (sounds stupid, but meh)
imbue/force type (fire,ice,light)
weapon skill type (bicheon, heuksal, pacheon)
skill type (soul spear, ghost spear, strongbow...)


Go, help me to list all things possible could affect it.


Too bad after that door, is a labyrinth, show me the right path then?

Okay, tomorrow I'll have some free time. I'll be able to plvl 3 characters to 24 (I'll get level 24 weapons with high AR). STR, INT and 1:1. After that I'll submit the excel file that was requested a long time ago. Sorry for the delay. ^^

Edit: I'll write down this...
Something more I should add?