+ Reply to Thread
Page 3 of 5 FirstFirst 1 2 3 4 5 LastLast
Results 21 to 30 of 45

Thread: [Guide] Defence Penetration vs Critical Damage (Also 2H vs DW)

  1. #21
    Quote Originally Posted by saltminer View Post
    I guess I need to do this in two parts.

    Part one, the resilience formula comes from the Koreans that did the theorycrafting, confirmed by at least a couple people on these forums. That's the same theorycrafting that produced the x/(x+7900) for defense mitigation, that you used, and 8000(?) for toughness mitigation. So I don't get how you're not running with the Korean info for your resilience calculation, yet you are running with it for the next update? Spamming the word 'casual' and speculating incorrectly is lazy, and the BS asymmetry principal comes to mind (it takes an order of magnitude more effort to disprove BS than it takes to spout it), so I'll just leave that at that.

    Part two, where you suggest a 'test,' makes me wonder. It's rather difficult to setup a test scenario where the character receiving the damage manages to have no other stats in play that affect the accuracy of a single round of tests, let alone when more resilience is added for further rounds of testing, since that virtually requires adding gemmed gear, which at least will have a defense value. On top of that, it takes a lot of averaging to handle the outputs of even a single round of testing, as the damage comes out in a distribution of values unless you've hit the crit damage floor.

    That said, if it's 14% per 1k, and you use 1.4k, then it's 19.6% damage reduction, because basic multiplication. While I give you the benefit of the doubt that you rounded, the way you've responded to me, as well as all the rounding you did in your original post, suggests that you just used the formulas you thought were right to produce results from initial values, and didn't do the actual testing in-game. It's especially obnoxious because I've done the actual testing in-game. If you'd done any non-casual testing in-game (since your post was pure theorycraft), you'd notice that some skills output random damage values (flamebolt, for instance) even if it's several hits of the same type (crit) against an unchanged target, so you end up having to average all possible outputs to 'check' the damage. You also have to account for fixed damage reduction, and whatever else on your resilience gear that interferes with testing, as well as staying above the damage floor so results aren't artificially clipped.

    After all that, you wouldn't be saying 'always' or 'exactly' a rounded number like 14%. But I know you didn't do all that. That said, I'm fine with saying x/(x+14000), because that same formula, which I used as an educated guess on my first set of testing how much reduction resilience did(one gem versus no gem in the shirt, all other stats accounted for), turned out to fit the remainder of my tests to 99% accuracy when I did 2,3,4,5,6, and then 7 resilience gems in the shirt and ran averages. All you did was theorycraft from a few good assumptions, and a bad one. I'm just saying, fix the bad one.


    do not spread misinformation. Oh, and thanks for the source.

    EDIT: On a side note, if you ran this game and wanted to make a quick change to improve the effect of resilience, you might just try changing that constant in the denominator. The lower that constant, the more effective resilience would be. And if you lowered that constant from 18000 to 14000, some folks might check that ratio and notice that 18000 is about 1.3x as large as 14000, and if you felt like you overshot it and made resilience too good, you might even try raising it to 15000 instead. 18000 is 1.2x as large as 15000. And if you think most of the people that would read all that would be annoyed, you might simplify it, and say "we're going from 1.0, to 1.3, to 1.2, roughly speaking," just like your link says.

    Or you could handwave "illuminati" at me. Meh.
    1) The Damage Mitigation Formula of Defence, is not something "Korean", you just look at which number you reach 50% Reduction and you simply use the Formula that 99% of MMORPG/RPG use for Diminishing Returns... Value/(Value + X)... You reach 50% at 7900 Defence, so you don't need neither a degree in rocket scienze nor a Korean to know that the formula is, indeed Value/(Value+7900).

    2) Other Mitigation Stats do not matter, because the Damage get reduced by a fixed % amount, so that you are doing 100, 1000, 2000 the % of Damage reduction is always the Same. The Final Damage is still the Final Damage, and when you Crit it get Multiplied normally by your Critical Damage, so if you do 20, 30, 40... It does not matter, if you have 50% Critical Damage it will do 30, 45 and 60. It's basic math and how Multiplications works.

    3) You use Basic Attacks, you do not use Skills to test it, Basic attack have a variance of 1% (They changed Weapon Damage Variance from 10% to 1%), so you just keep attacking, you see your maximum and minimum damage, and same for the Crit... After that you just use a simple:
    Non Crit:100=Crit:? (Again, basic math)

    3) No Korean did ever spoke about that Formula, it all comes from that Thread (http://forums.archeagegame.com/showt...-and-Toughness), and the guy who did that awesome test is this Russian (https://www.youtube.com/watch?v=CTlporSMN6k).
    After that NO ONE ever spoke about that formula again.
    So please avoid making up things to try evaluate what you are saying with the classical "The Korean said/say/did/do" since people normally can't confirm or deny it, because you probably encountered one of the few person that lurk the Korean AA Board every single day since AA Launch.

    4) Resilience does not have a Formula with Diminishing Returns like Toughness and Defence, for a SIMPLE reason, Critical Damage CAN GET COMPLETELY NEGATED... You can even test it by yourself, Just go with a low amount of Critical Damage Bonus (Like 60%) against a Target with even only 8k Resilience... You will see that your Attacks will never exceed that 50% Base Critical Damage Bonus Mark.

    If it had a X/(X+Resilience) Formula, your Critical Damage would be reduced in % and would never be at Base 50%, so if for Example you had 60%(+50% Base = 110%) Critical Damage and your Target 8k Resilience, the Resilience will reduce it by 30% (Using your Resilience/(Resilience+18000) = 8000/(8000+18000)=30% Reduction), you would still deal 92% Critical Damage (60x0.7=42 42+50=92%), but it's not the case and you deal only 50% Base Critical Damage.

    Here we go, simple as that.

    5) I'm sorry but I always test everything myself before speaking and you would probably agree if you at least tried once and saw it by yourself... I'm still open to meet you in Marianople on EU Jakar if you have time


    Being the classical Internet Hipster that want to think differently even when is in the wrong, don't help your cause for sure, but I suppose it would not be a classic Forum Thread without some pointless and baseless opposition. But feel free to stay on your opinion, no one is forcing you to do otherwise.

  2. #22
    Senior Member
    Join Date
    Apr 2018
    Posts
    214
    Quote Originally Posted by Harveaux View Post
    1) The Damage Mitigation Formula of Defence, is not something "Korean", you just look at which number you reach 50% Reduction and you simply use the Formula that 99% of MMORPG/RPG use for Diminishing Returns... Value/(Value + X)... You reach 50% at 7900 Defence, so you don't need neither a degree in rocket scienze nor a Korean to know that the formula is indeed Value/(Value+7900).

    2) Other Mitigation Stats do not matter, because the Damage get reduced by a fixed % amount, so that you are doing 100, 1000, 2000 the % of Damage reduction is always the Same. The Final Damage is still the Final Damage, and when you Crit it get Multiplied normally by your Critical Damage, so if you do 20, 30, 40... It does not matter, if you have 50% Critical Damage it will do 30, 45 and 60. It's basic math and how Multiplication works.

    3) You use Basic Attacks, you do not use Skills to test it, Basic attack have a variance of 1% (They changed Weapon Damage Variance from 10% to 1%), so you just keep attacking, you see your maximum and minimum damage, and same for the Crit... After that you just use a simple:
    Non Crit:100=Crit:? (Again, basic math)

    3) No Korean did ever spoke about that Formula, it all comes from that Thread (http://forums.archeagegame.com/showt...-and-Toughness), and the guy who did that awesome test is this Russian (https://www.youtube.com/watch?v=CTlporSMN6k).
    After that NO ONE ever spoke about that formula again.
    So please avoid making up things to try evaluate what you are saying with the classical "The Korean said/say/did/do" since people normally can't confirm or deny it, because you probably encounter the only person that lurk the Korean AA Board every single day since AA Launch.

    4) Resilience does not have a Formula with Diminishing Returns like Toughness and Defence, for a SIMPLE reason, Critical Damage CAN GET COMPLETELY NEGATED... You can even test it by yourself, Just go with a low amount of Critical Damage Bonus (50%?60%?) against a Target with even only 7/8k Resilience... You will see that your Attacks will never exceed that 50% Base Critical Damage Bonus Mark.
    If it had a X/(X+Resilience) Formula, your Critical Damage would be reduced in % and would never be at Base 50%, so if for Example you had 60%(+50% Base = 110%) Critical Damage and the Resilience reduced by 50% you would still deal 30% more Damage (Aka 80% Critical Damage), but it's not the case.

    5) I'm sorry but I always test everything myself before speaking and you would probably agree if you at least tried once and saw it by yourself... I'm still open to meet you in Marianople on EU Jakar if you have time


    Being the classical Internet Hipster that want to think differently even when is in the wrong, don't help your cause for sure. But feel free to stay on your opinion, no one is forcing you to do otherwise.
    1) I didn't say the formula is Korean, just saying, those guys commented on it first, as far as I can tell, and later, people on these forums confirmed them. You misread me.

    2)Other mitigation stats matter for in-game testing, but not for your contrived theorycraft. Fixed damage reduction can potentially shave off the last few points on lower-damage attacks (the archer lament of "1 1 1 1 1," for example)and obfuscate the results of damage testing. One of your examples on the calculator had 52 fixed damage reduction. If you do 20, 30, or 40 to that guy on a non-crit, on a crit, you'd do 56, 71, or 86, for example. Again, you'd know this if your tests in-game were rigorous.

    3) Cool, you haven't found it, so it's wrong and you'll keep handwaving? They came up with x/(x+33000) as an estimate for crit rate reduction too, is that in either of your links? I'm gonna bet no.

    4)I agree with you that the floor is 50%. Ah, I see the part you're not getting. The formula reduces the total damage of a critical hit. So, take your example ("60%(+50% Base = 110%) critical damage. That's 210% of a normal hit. The formula, at 7k resilience, reduces it by the fraction of 7000/21000, or 1/3. So it would be 2/3 of 210%, or 140%. But the floor is 150% if you're describing the floor that way. So the damage dealt on target would be 150%, or dealing 50% more damage than a normal hit.

    5)I don't have an EU character or the time.

  3. #23
    Quote Originally Posted by saltminer View Post
    1) I didn't say the formula is Korean, just saying, those guys commented on it first, as far as I can tell, and later, people on these forums confirmed them. You misread me.

    2)Other mitigation stats matter for in-game testing, but not for your contrived theorycraft. Fixed damage reduction can potentially shave off the last few points on lower-damage attacks (the archer lament of "1 1 1 1 1," for example)and obfuscate the results of damage testing. One of your examples on the calculator had 52 fixed damage reduction. If you do 20, 30, or 40 to that guy on a non-crit, on a crit, you'd do 56, 71, or 86, for example. Again, you'd know this if your tests in-game were rigorous.

    3) Cool, you haven't found it, so it's wrong and you'll keep handwaving? They came up with x/(x+33000) as an estimate for crit rate reduction too, is that in either of your links? I'm gonna bet no.

    4)I agree with you that the floor is 50%. Ah, I see the part you're not getting. The formula reduces the total damage of a critical hit. So, take your example ("60%(+50% Base = 110%) critical damage. That's 210% of a normal hit. The formula, at 7k resilience, reduces it by the fraction of 7000/21000, or 1/3. So it would be 2/3 of 210%, or 140%. But the floor is 150% if you're describing the floor that way. So the damage dealt on target would be 150%, or dealing 50% more damage than a normal hit.

    5)I don't have an EU character or the time.

    1) Quoting you: "Part one, the resilience formula comes from the Koreans that did the theorycrafting, confirmed by at least a couple people on these forums. That's the same theorycrafting that produced the x/(x+7900) for defense mitigation"
    You said it was Korean, but the funny thing is Korean do little to no Theorycrafting / Min-Maxing, what they usually do is stacking any Tanky Stats in any Game they play, and that's it... But let's avoid speaking about it since it's not the main Topic.

    2) Shouldn't the Fixed Damage Reduction be Subtracted before the Critical Damage Multiplicator take place? Should be easy to be tested by critting against a naked Target without any Critical Damage Bonus, you will deal X Damage and exact x1.5 when you Crit, but maybe the damage is so small that you don't even notice it.
    Like if you deal 1050 Damage, you remove 50, it become 1000 and then you Multiply it by 1.5, resulting in 1500, but it can work differently, I'm not sure.
    But even if it wasn't the case you just use a Skill that deal a large amount of Damage (The variance is still 1%), and you solve the problem, that is why when I tested the Precision Strike Critical Damage Bonus (8k Base Damage) the results were anyway 150% on high Resilience Targets. So as I said... It still don't matter.

    3) It's not that I haven't find it, is that it does not exist, I just posted the only Thread where that "Formula" was mentioned, and THE WRITER OF THE POST EVEN SAID that it comes from the Video of a Russian Player and he linked that Video, Video that I also posted where that Russian Player try to calculate that Formula... I mean, it's self explanatory.

    4) There is a little fallacy in your theory... I will try to explain it another time, since you are supposing that Tests that are already be done, do not exist.

    10k Resilience: 140% Critical Damage Reduction (On the true formula) or 35% Total Critical Damage Reduction (With your opinion).
    I have 190 Total Critical Damage (140% + 50% Base), following your formula, if we reduce 290% by 35% we will get, 188.5% right? Ok, nearly 40% Damage Bonus from the Base of 50%, this is a huge amount right?
    The problem is that with 190% Critical Damage you deal 50% more Damage against a 10k Resilience Target and NOT 88.5%, this is TESTED, NOT THEORY...
    And you can't even say "Oh is a small number, it can be hidden behind the 1% Variance of the Weapon Damage", it's 40%... You notice a 40% difference even on a small attack that deal 10 Damage. (And my Basic Attacs deal near 2k damage).

    Another thing is that if you use the first example that I suggested you to try "No Critical Damage + Battle Focus" Against someone with 1400Resilience (1420 would be ideal, but I menaged to get exactly 1410 with 1 T1 +0 Gem, Ancestral level 36 and 1 Resilience Lunafrost), you will see that with 70% Critical Damage you will still deal 50% more Damage, if you spend some more point in BattleRage and you get Delyrium Stack, you will see that with Delyrium Stacks the Damage of your Critical Strikes will accordingly increase by 5%, 10%, 15%, 20% and 25% at Max Stack... A thing that is not possible with your Multiplicative Formula.

    Resilience get directly subtracted by a Fixed Amount and right now is 14% every 1k.

    If you are really interested in the argument and are you really in searching for knowledge you have two options: 1) Test it 2) Trust it.
    The third option is related on "I want to be right even if I didn't test anything, And I only read something about it 2 years ago without any real method/calculation/specific" and is not for sure related to know the truth or improve yourself as a person and/or as In-Game Character.

    The first thought that came to my mind, is that you have a strong bias on your Build or simply on your preferences, and you want to be right so your choice/taste can't be challenged by anyone... Or at least, that is what happen in 90% of the case when someone debat something without anything concrete to back it up.

    As I said, I'm still open to meet you in Marianople But I guess is not a problem related on time, but a problem on being wrong.


    5) Indeed you don't have time to test things, I do... So MAYBE and I'm saying maybe, don't want to step on your ego... You are wrong and someone who spent time testing it is right? It's just an option.


    As said multiple time on that Forum, I always tend to speak with Math and Facts, I don't have a bias in anything since I couldn't care less if Defence Penetration is better or if Critical Damage is better, I will end up anyway on "CC + Slash Slash + CC + Slash Slash" in any case.

    If I don't know something, I usually ask about it, stay quite or say I'm not sure, since I don't like looking ridicolous on a Topic where I don't know anything about, but that's only me.

    In any case, keep your opinion, not really my problem. If you don't believe it, test it yourself, if you don't care to test it, why should I care for you.

  4. #24
    Junior Member
    Join Date
    Aug 2018
    Posts
    16
    hey thanks for the testing and all, as a archer i like these info topics.

    dont iunderstand why players keep bringing Korea into the forum because our game is difrent from theres and even if they get new stuff does not mean we get it, so testing things on these servers is a good thing.

    got one question

    how wil a penetration build work in pve and by how much wil i see loss in damage ?

    this i ask because i'am not fully gemmed and not good in pvp yet but befoor i fully gem my self i dont want to have it all taken out and recloak them and lose alot pf gold and time with it.

  5. #25
    Quote Originally Posted by Eagon View Post
    hey thanks for the testing and all, as a archer i like these info topics.

    dont iunderstand why players keep bringing Korea into the forum because our game is difrent from theres and even if they get new stuff does not mean we get it, so testing things on these servers is a good thing.

    got one question

    how wil a penetration build work in pve and by how much wil i see loss in damage ?

    this i ask because i'am not fully gemmed and not good in pvp yet but befoor i fully gem my self i dont want to have it all taken out and recloak them and lose alot pf gold and time with it.
    We will eventually get everything that is on Korea, but the problem is not Korea or Koreans... The problem is when people bring them up and make up things or create false statements saying "Koreans said that!"/"Koreans did that!" when it is completely not true and they only want to add a Non-Existent Autoritarian Source to support what they are saying... That's the only problem


    Anyway, you deal zero damage in PvE with a Defence Penetration Build compared to a Critical Damage Build, so yeah, it will be hard if you do a lot of PvE/Dungeon/Grinding, but if you have really good Weapon (Like a Mythic T4 Hiram) you can still sustain yourself, but nothing compared to a Critical Damage Build.

    For a more precise answer you will lose exactly the % of your Critical Damage So if with a Critical Damage Build you have a total of 250% Critical Damage, you will deal 200% less Damage.

  6. #26
    Senior Member
    Join Date
    Apr 2018
    Posts
    214
    Quote Originally Posted by Harveaux View Post
    wall o' text
    1) The ideas stand on their own. A Korean finding them first doesn't make them Korean ideas, just ideas. I tested those formulas, except for the crit rate one, and got the same results, so they seemed sound. I tested the resilience after the first change ("september change" in the Korean link you posted), and it appeared to be in line with x/(x+14000), so while there may have been an undocumented change to it since then, I really really doubt the same two companies, that didn't have the competency to successfully localize the LANGUAGE of the Korean updates that they patched to NA, ninja-adjusted resilience.

    2) You keep appealing to "just try it," when the resilience you talk about is wrong after I've "just tried it." I don't feel like just trying it, though I checked my files, and thankfully labeled my spreadsheet of values I got from "just trying" resilience before. I'd be willing to wager my wallet that you don't have any numbers compiled from your tests. The randomness of the dmg values practically required a spreadsheet to make the comparisons/averages.

    And again, you're just 'telling' me that fixed damage reduction is applied first, meanwhile, there's word from a worker at Trion who says that flat reduction comes last in the damage formula, but I'm supposed to take your word for it AND assume that my testing was messed up and gave me the same conclusion as a guy that worked for the game.

    3) The ideas that come from those posts stand, regardless of who said them first/second or in Korean/English/Esperanto.

    4) Your numbers, with my 'opinion' and 10k resilience, was done incorrectly. x/(x+14000), not 18000, x = 10000, => 10/24 = 5/12 reduction, not 35%. 290% goes down to about 169%, not 188.5%. But anyways, restating I'm wrong again, and telling me to test again, when you're repeatedly doing bad math, (and I have tested the underlying value of resilience,) is bizarre, and contriving examples on paper that you apply your bad assumption to, in order to 'prove me wrong,' won't ever go anywhere.

    Anyhow, according to your guess, 170% crit dmg gets brought down to 150%, but according to the formula, it gets brought down to around 154.5%. I've already found one error in your testing (you didn't know how flat damage reduction works), which might account for the discrepancy on its own, but your math has been sketchy so far too, and I don't know what numbers you're guessing from or what methodology you applied to get them, so I don't know.


    Might want to take your own advice though.

  7. #27
    Quote Originally Posted by Ollolol View Post
    That doesn't seem correct to me. Maybe it is but that would mean that all the lower level people that hit my lvl 39 have sufficiently worse gear than the high lvl people that hit me.

    It's really noticeable when higher lvl people are hitting me. When I check the people who are hurting me they are always high lvl. Correlation is not causation so it's possible that they are high lvl because they hit hard

    I'm not particularly abnormally geared. Mostly T3 Epic Hiram plate with 2 pieces T4 Mythic (Pants + chest with gems). No shield (dual wield).
    Yeah, at the beginning I was worried too about that... But it's only an impression, probably the people with Higher level also have higher gear score (or simply better damage) and made you thought that Ancestral mattered in some regards, but don't worry it's not like that!

  8. #28
    Junior Member
    Join Date
    May 2016
    Posts
    6
    i never heard of that formula for calculating res,when i asked in game in the past i always got the answer of -1000res reduce 10.5%/11%- and i know that recently they made res stronger dont know how much
    what i know for sure is that i have with two chest pieces one plate with spirit and one plate with intelligence both divine but the one with intelligence is full gemmed and the one with spirit have zero gems i have like 140 crit dam and 2800 dpen on my archer alt and i deal the sam exact damage with both chests shouldnt be different if it decrease the crit dam in %?in total i have 9800 res with int and 7100 with spirit
    edit> i just spammed endless for like 10 min the min damage was 581 and the max 604 with crit 867 min and 890 max these min and max appeared multiple times in both chest pieces is it normal? if i look at it the combat log the number do repeat and are the same for both chests

  9. #29
    Junior Member Silencium's Avatar
    Join Date
    Jan 2019
    Posts
    23
    The crit rate reduction from Resilence was calculated a long ago on russian servers by some players with all formulas. I took a little time to translate just a quick graph to reveal the crit dmg and crit rate reduction from fixed amount of resilence.

    They also shared all formulas and even some Excel tables to put, for example, some amount of Def pen or Crit dmg and get the final amount of dmg from some skill. I'm a bit lazy to translate it also but maybe i'll share it after.

    Also few words about Def pen vs Crit dmg. Without getting into some formulas and hard math. If you are using 2h, def pen has more profits for you. Mainly because you can't really reach the required minimum of crit dmg with 2h. Of course you can go with Ayanad set for melee, for example, make an eternal weapon for additional gem slots but those people already know what they want from their character.
    Then you get some defence reduction debuffs (critical discord which gonna be buffed also, charms with songs, melee passives) which can change the game too.

    At least till the update with buffed 2h gems i'd recommend to go def pen with 2h. Even knowing it gonna be boosted after. In reality the dmg difference between def pen and crit dmg is low against leather set or cloth with a shield.
    Silencium | Jakar
    Introverts Party

  10. #30
    Senior Member MiroH's Avatar
    Join Date
    Nov 2018
    Posts
    139
    I need to see numbers ingame with staff 2h magic def pen alot of people on my server think crit damage is better for a staff user but the math tells me other wise. Ill make a staff and test it Im sure some people in here have tested it ingame if you have please comment.

+ Reply to Thread
Page 3 of 5 FirstFirst 1 2 3 4 5 LastLast

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts