Talk:Damage calculation

Archives: oldest on top, newest on bottom.
 * /Archive -- User:Cloak of Letters's big list of questions and their responses. Most have been answered (and the articles explain them now).  The only question that seems to be unresolved is whether Necrotic/Tormenter take extra damage from light damage.
 * /Archive2
 * /Archive3
 * /Archive4 -- newest archive

Holy on undead: Double or -40?
I'll leave other ppl to figure out how to test it, I'm just asking the questions to make our knowledge more complete d-:

One interesting thing I noticed about holy damage on undead is that it can't crit. It will always do double the normal value, never critting.--Dragos 13:21, 22 November 2005 (UTC)

BTW, when minions take damage, can the necro see how much damage was taken? If not this might have to be tested in PvP )-: -PanSola 15:19, 5 December 2005 (UTC)

Again: Needed Testing on Damage Level Attribute Threshold
Characters level 2-7, 9, 10, 12, 13, 17. Swing 10 criticals on a Whiptail Devourer (note, they're right outside Ascalon) and report your damage as well as the exact stats of your weapon/ what your attribute level is. We need to verify the diminishing returns formula holds at every level.

Judge's Insight vs Necrotic testing
Since we got two ppl whith necro around, I figure might as well ask here:


 * Damage done with JI from Doppleganger with AL60 standard Armor
 * Damage done without JI from Doppleganger with AL70 Necrotic/Tormentor's Armor (Drok or 15k)
 * Damage done with JI from Doppleganger with AL70 Necrotic/Tormentor's Armor (Drok or 15k)


 * Damage done with JI from Doppleganger with AL30 standard Armor from Lion's Arch
 * Damage done without JI from Doppleganger with AL40 Necrotic Armor from Lion's Arch
 * Damage done with JI from Doppleganer with AL40 Necrotic Armor from Lion's Arch

Unfortunately with necro armor and monk skill this means there's no room to fit wildblow in there )-: So please take around 20~30 70~100 values for each case to account for the weapon's damage range.

If there is a way to set up PvP dueling we can do even more controlled testing. -PanSola 15:16, 5 December 2005 (UTC)


 * I need to test more. The dopple is using a bow and the hits go all over.  Since there is no necrotic head gear the damage varies a lot more than it should.  But as of right now my results are:
 * 44-67 damage with AL60 (full scar pattern set)
 * 42-87 damage with AL70 (full necrotic set - im not too sure on that 42, I need to retest)
 * 27-72 damage with only legs as necrotic (I was shocked to say the least when I saw 27)
 * So far it looks the same as the testing with banish ... +20 damage for 4 pieces of necrotic armor. I'm sure we can get a group of 8 people together to test this in teams (2 teams of 4 in international).  I'm available in the evening generally from 4-9pm (3-7pm Fri/Sat). | Chuiu 21:32, 5 December 2005 (UTC)
 * To account for the lack of headgear, keep statics on how often eachnumber pops up. That way we can filter out the 12.5% oddity, given enough numbers. -PanSola 21:50, 5 December 2005 (UTC)
 * Remember the control case of damage w/o JI also
 * Remember to specify timezone ^^" also if we do get ppl to test in PvP, I want to see if necrotic take extra from Light damage (from monk regular weapons) as well.  and while we are at it, summon minions put shilelding hands on them and have someone with JI do wildblow to see if it's double damage or -40AL ^^" -PanSola 21:45, 5 December 2005 (UTC)


 * Those times I posted are central. And I'm 50% sure (made up statistic?) that necrotic DOES take extra from light.  This is just what i've seen in PvP when monks are attacking me while they aren't healing allies. | Chuiu 22:01, 5 December 2005 (UTC)
 * Could be customized weapon though... hard to say -PanSola 09:26, 8 December 2005 (UTC)

Damage Groups...
It is useful to classify damage into damage groups (especially for PvP).

This seems to imply every damage belongs to (at least) one group. Is that the case? -PanSola 23:55, 6 December 2005 (UTC)

It is useful to classify damage into damage groups (especially for PvP).

Can someone word that better, I'm not sure that I can figure out what it is trying to say. --Rainith 16:17, 12 February 2006 (CST)
 * I'm changing that section into a "see also" since no one can seem to explain it. -PanSola 04:13, 26 February 2006 (CST)

TeX formatting
Does this wiki support it? I believe I tried previewing a $$formula$$ but it didn't show up. 69.124.143.230 03:20, 25 February 2006 (CST)
 * My guess is guildwiki doesn't support TeX formatting. -PanSola 04:14, 26 February 2006 (CST)

Illustrative Examples
--Savio 06:34, 27 February 2006 (CST)

I think we want a few examples of how the formula works, so people can follow the example and truely understand how to use the formula. As opposed to some black-and-white curves comparing warrior skill damages at different levels. People who were confused by the formula will remain confused when looking at those graphs. -PanSola 04:25, 26 February 2006 (CST)


 * Sorry to be so critical, but the graphs are very hard to read. Take this from someone who makes her living reading graphs all day. If the raw data were posted somewhere, I can make better looking plots out of them. &mdash; Stabber 04:30, 26 February 2006 (CST)
 * I agree with yoru point too, but my primary point is that those graphs don't seem to suit the need of the "Illustrative Examples" section of this particular's article d-: -PanSola 04:36, 26 February 2006 (CST)

Also, the comparaison between BP and Cleave is a bit ridiculous, Cleave is a powered up ES, if you need to swap out something to take Cleave, this is what you should remove, not BP. Also, considering that you'll be doing Cleave more often than BP, it's doubtfull that the small difference in dmg per hit is really going to be significant. --theeth 06:27, 26 February 2006 (CST)


 * The work itself has been refuted by Ensign in that thread as well as others. --Karlos 19:15, 26 February 2006 (CST)


 * I'm not critisizing the data, obviously, great work and care has gone through that, I'm debating the interpretation and conclusion thereof. Basicly, I feel like the article is saying that one skill is fondamently better than the other while only comparing single-hit damage and not addrenaline charging time nor even DPS output. --theeth 22:18, 26 February 2006 (CST)


 * But the data is wrong, and it really shouldn't have been left up when the creator knew it was wrong. The +damage should fall under DShift, not in AE. In other words, +damage ignores armor. If it wasn't removed already I would have taken it out because it was misleading. -Savio 06:34, 27 February 2006 (CST)
 * Yeah it looked fishy to me, but I didn't have the original numbers nor did I know the thread where it first came up, so I figure I'll just argue they don't belong here first, and if somehow they are staying then I'd worry about checking their correctness d-: -PanSola 06:38, 27 February 2006 (CST)

stub status
While I don't feel insulted over this article being stubbed, personally I think this article is sufficiently complete to be removed from stub status. Consider this an invitation for comments. If I don't see reasonable opposition I will remove the stub from this article in a few days. -SolaPan 09:24, 20 March 2006 (CST)


 * Agreed. JoDiamonds 07:00, 22 March 2006 (CST)

Damage equation issues
moved from User talk:Savio

Splitting this off, there are currently three issues I have with the damage equation listed. The first is more of an issue with Knight's/Ascalon armor and absorption rather than the damage equation, but the equation should be able to take them into account. Knight's/Ascalon ignores most but not all damage; you can see this happening with Giant Stomp which deals 25 damage, and Mursaat Tower which deals 20. However, shield absorption will reduce them just fine, so either Knight's/Ascalon has a category of skills that it doesn't reduce damage from, or the absorptions apply at different times.

That Giant Stomp dilemma is somewhat related to the more important second point. Where does armor-ignoring damage come into play in the equation? It obviously does not bypass all of it, as Obsidian Flame will deal double damage to a Frenzy user, and Smite still does more damage to Undead. It would make much more sense to change ([BD × DEnhance × AE × DScale] + DShift) in the equation to something like [Actualized Damage] (which I took from Ensign's Guide) which then can come in two forms: AD as damage from attacks, which is equal to the aforementioned formula, or AD as damage from armor-ignoring skills.

The third problem is that DEnhance and DScale do the same thing, except one comes from weapons and one comes from skills. An important note to add is that all separate +/-xx%'s multiply together, so you're actually getting 138% damage from a 15^50 customized weapon. Also important is that they all do not affect skill damage, only base damage. Pet evolution would fall under DEnhance/DScale/whatever you want to call it, as Jenosavel and Epinephrine have disproven the myth that pet evolutions add/subtract a fixed amount of damage, here. -Savio 16:21, 12 April 2006 (CDT)


 * Frenzy and light/holy damage vs. undead causes a DScale or DMult of 2. I think the other points you note are well covered by the fact that these damage multipliers are used with "BD". &mdash; Stabber (talk) 16:34, 12 April 2006 (CDT)


 * I'm gong to respond to your issues in the reverse order, easiest ones first.
 * First of all, DEnhance and DScale are DIFFERENT, in the way they stack. The end of the Talk:Damage/Archive3 section chronicles how I came upon the discovery (originally DEnhance and DScale were part of the same term), with skills and numbers clearly outlined so you can reproduce the test.  The fact that they only modify Base damage and not "skill damage" is explicitly stated in the Damage article.  If it is not clear enough, I can look into rewriting it to be even more explicit.
 * 1.5. As for pet evolution, the test conducted by Jenosavel and Epinephrine failed to take into account critical hits affecting damage range distribution, thus it did NOT disprove the fixed amount of add/subtract theory. I have communicated this issue with Epinephrine, and we found that it requires a coordinated PvP match to research the data necessary (impossible to test with mobs), and it's hard for Epinephrine to find enough people to set up a coordinated match for research purposes.  Unless pet damage range has been extensively established with critical hit damage explicitly taking into consideration, AND evolution damage study being re-done based on the result of the new base damage range study, the interpretation of the numbers Jenosavel and Epinephrine collect remain questionable.  Not to say I think the fixed amount theory/myth is absolutely right, but before a better study is done, I opt to go with the easier-to-apply rule which hasn't proven wrong yet, unlike what you believe.
 * Second, when does armor-ignoring damage come into play? It comes into play when calculating the AE.  Frenzy is NOT related to armor (it is DMult), so your example of Obsidian Flame to Frenzy is moot (and if you can't figure out why Frenzy isn't party of armor, then even using Ensign's formulation you still would have the exact same problem).  Undead creatures don't have -40 AL vs holy, they take double damage (also DMult), so there is no issue with Smite being armor-ignoring and dealing double damage to undead (and once again, using Ensign's formulation will make no difference.  Before you take another shot, Aura of the Lich is also DMult (anything that doubles/halves are DMult).  DMult are not related to armor, and thus act business-usual even when encountering armor-ignoring damage.  Besides, what you are proposing should only be setting [BD × DEnhance × AE × DScale] to [Actualized Damage], since DShift will still take full effect against Obsidian Flame and Smite.  Considering DEnhance and DScale only affect weaponry damage (though they stack differently), and AE's the part that gets ignored for armor ignoring damage, when calculating armor ignoring damage you will end up having 1 = DEnhance = AE = DScale.  So I do not see the point in replacing that part by [Actualized Damage].
 * Finally, absorption. I am inclined to agree that "Knight's/Ascalon has a category of skills that it doesn't reduce damage from".  A better study should be conducted over it.  Do they only (and always) not absorb damage from monster skills?  Is it by damage type and/or skill type?  These info would be crucial in revising the formulation of absorption properties. -PanSola 18:29, 12 April 2006 (CDT)
 * 1. your tests here at the bottom show the exact opposite of what is shown now, that weapon % multiplies and skill % adds. So I don't know who proved what, and I really don't care as it'd just be easy to run out and retest it against a 60 AL target.
 * Ebon Hand Axe of Fortitude, Slashing Dmg: 6-28, 15^50, customized. If it adds, it's 28 * sqrt (2) * 1.149 * (1.35) = 61.42 ~ 61. If it multiplies, it's 28 * sqrt (2) * 1.149 * 1.15 * 1.20 = 62.78 ~ 63. It's 63. Weapon boosts multiply.
 * Candy Cane Hammer, Blunt Dmg: 15-15. At 12 Hammer Mastery/0 Strength, normal hit always 15, critical 21. Life Attunement and Flurry: if they add, it's 21 * .45 = 9.45 ~ 9. If they multiply, 21 * .525 = 11.025 ~ 11. Crits are showing 11. Skill boosts multiply, but since people seem to hate CC stuff, retest.
 * Ebon Hand Axe of Fortitude, Slashing Dmg: 6-28, 15^50, customized. At 16 Axe Mastery/0 Strength, critical hit 63. If they multiply, 63 * .525 = 33.075 ~ 33. If they add, 63 * .45 = 28.35 ~ 28. Tests show 34. (Probably some rounding error earlier on. Grr.) Skill boosts multiply.
 * 1.5. As far as Pet damage goes, I haven't really tested it myself and thought that they had proven it enough. I don't recall any tests having results for fixed pet damage, so if's just an unfounded theory I'd rather it was left out until we know for certain. People have this habit of reading an old page, looking at misinformation, and not checking back later.
 * 2. For Actualized Damage, I wanted to clarify certain things with the DEnhance/DShift. DScale claims that it causes "x% more/less damage from skills" which is false, so I'd be fine with the equation if just DScale and DEnhance were fixed and merged. Possibly it could be called DAttack since all it affects are attacks. For other sources of damage it should state that DAttack is 1.
 * 3. Absorption: It could either be skills that have a fixed, non-changing damage, or Monster Skills. Since all Monster Skills have fixed damage (i.e. Monster Stomp does 25 regardless of level, armor, or attribute), and no player-available skill does fixed damage, I wouldn't know how to prove the difference between the two. -Savio 23:45, 12 April 2006 (CDT)


 * Re point 1: both you and PanSola seem to show that weapon bonuses multiply. The damage article seems, to me, to contain a transcription error (I'm being charitable here) and the stacking mode of DEnhance and DScale were switched. Anyway, it's fixed now. &mdash; Stabber (talk) 00:00, 13 April 2006 (CDT)


 * I just went out and tested with Doppleganger. To my surprise, I indeed get a result inconsistent with that recorded in archive 3.  I can only surmise that between November and now Anet has changed how the modifiers stack up.  And because now they stack the same way, it does make sense to merge them into one modifier.  As for "x% more/less damage from skills", my wording was probably not clear.  I meant the damage reduction effect were from skills, I didn't mean it reduce damage from skills.  Anyways, that is moot now that we are merging the two.-PanSola 00:46, 13 April 2006 (CDT)


 * Referring to the above, the damage done by pets: The difficulty was in separating the critical hits (which are variable, and thus difficult to determine) from the actual hits.  Is a low frequency 29 damage hit a low value critical hit or is it simply an infrequent high damage hit.  One might argue that where the drop in porbabilities occurs the critical hits start, but it isn't known with any certainty that that is the case.  PanSola suggested a great idea for testing the actual damage range, and the effect of critical hits, but it has proven beyond my abilities to arrange, as it required using scheduled GvG and willing participants, which I was unable to get enough support for.  The upcoming Factions release is said to have 1v1 guild scrimmage, which is a boon to testing - now we will be able to easily test the damage range of pets at least, and come up with (hopefully) conclusive evidence to back up either the +X% damage or +X damage theories.  Given that pet criticals seem to behave nothing like ordinary criticals, one wonders whether the use of balanced stance will be enough, but one can hope that it will be sufficient to unravel the mystery.  --Epinephrine 13:58, 24 April 2006 (CDT)Given that the

Excact Damage Equation, with 0 Strenght and 12 Axe
I have noticed this has not been discovered yet, or at least is not present on your page. I have from own research, found an equation for damage. The only information discovered by others I based my research on, is that you deal 100% damage with 12 axe, and that with 60 AL, you take 100% damage. I used a candy cane axe, to make my damage a constant (10).

First off, my results (Damage/Critical Damage)

Now, as you can see, there is a pattern. Everytime armor is reduced by 20, you do 1.4x the damage. Seemingly, critical hits does 1.4x damage as well. Then, I tried making an exponential function for this problem. As we know, with exponention functions, a^Tn=n, where T in the multiplication by n-constant, and n is any real number. Our n is 1.4. T of 1.4 is -20.


 * a^-20=1.4
 * a=1.4^(-1/20)
 * a=0.983317115

Then b. The function is of the type f(x)=ba^x. Thus


 * y=ba^x
 * b=y/(a^x)

I used 60 AL, since that is where you do the excact damage.


 * b=10/(0.983317115^60)
 * b=27.44

So f(x)=27.44*0.983317115^x where x is AL.

Now, remember, I did 10 damage. But what if I did 20 damage? I would do 2x the damage. So I simply multiplied by damage/10.

Damage=BaseDamage/10*27.44*0.983317115^TargetAL

To further prove this, I attacked the Master of Survival. I did 8 damage, 12 critical, and 12 when she had Weaken Armor on her, 17 critical when she had Weaken Armor on her. But hold on. 1.4x damage with critical. Seemingly not, as 8*1.4=11.2, rounded off 11. But I did 12 damage. And when he had -20 AL I did 12 damage as well. So it seems critical gives the target -20 AL. I used my formula from earlier, a^Tn=n. As we know, -20 is our Tn. And we know a. 0.983317115^-20=1.4

Which is the excact same as I did in reverse earlier. So critical reduces armor by 20.

But back to the ranger. As we know, rangers with best armor has 70 AL vs physical.
 * Damage=10/10*27.44*0.983317115^70=8.4515426=8

With 20 less armor, ie 50.
 * Damage=10/10*27.44*0.983317115^50=11.83216=12

And with my theory on 20 less armor when doing critical, applied to when she had Weaken Armor on, and I did critical, ie 30 AL.
 * Damage=10/10*27.44*0.983317115^30=16.565023=17

As you can see, it fits the bill. So yes, now you have an official damage calculation formula. I tested it in reverse on Mergoyles outside Beetletun on my elementalist. I normally do 105 with Aftershock. This time I did 180. I used CAS to isolate the TargetAL variable.

TargetAL=-59.440269*(ln(Damage/BaseDamage)-1.0094167)

I used this to find their AL.
 * TargetAL=-59.440269*(ln(180/105)-1.0094167)=27.961904=28

So Mergoyles have 28 AL vs earth.

Yes, I permit you to use my formulas for calculating damage or armor level, as long as you credit me. IGN: Zexion Sixtin, Wnx Zexion, Wnx Zexiana Leader of Warrior Nation [WN] WNxZexion on www.warriornation.net

Thanks you for reading.


 * Thank you for your efforts, but we already have our own calculations in the article that are not encomberred by requirements for giving credit. However, even if we chose to use it, you should not that the edit box shows below it: "Please note that all contributions to GuildWiki are considered to be released under the Attribution-NonCommercial-ShareAlike (see Project:Copyrights for details). If you don't want your writing to be edited mercilessly and redistributed at will, then don't submit it here."  Any applicable credit to you would already be provided by the name/datestamp in the postings log.  Technically, by posting it yourself, you've given free reign to use it with no further crediting required. --24.19.168.170 09:55, 23 April 2006 (CDT)


 * I have taken the liberty of reformating your post so that it is readable. Unfortunately, I find nothing there that isn't already adequately explained by our damage calculations. Furthermore, your reasoning does not adequately differentiate between a DL bonus of 20, an AL penalty of 20, or a damage multiplier of sqrt(2), which is the biggest unanswered question with regard to critical damage as we understand it. For your theory to become accepted, it must make an accurate prediction that the wiki's damage equation fails to predict. I haven't looked closely, but it seems that you haven't produced such an example. (I might even say that your equation is simply a simplified form of the wiki's damage equation that neglects to mention things such as DL bonuses, armor penetration, etc.) &mdash; Stabber 10:23, 23 April 2006 (CDT)
 * Your experiment with the Master of Survival does not account for the fact that the 8 damage might itself be the result of rounding down, and that the multiplier is not 1.4, but 1.4142... (i.e., sqrt(2)).


 * Thank you for explaining the sqrt(2) part to me. That is really appreciated. I did this mostly for my own fun, and to share it with my clan. But anyways, I have something you might wanna use: You currently use the formula 1/(2^((AL-60)/40)) to describe the % of damage taken. First off, that does not calculate the percentage. If it were to do so, it should have been 100/(2^((AL-60)/40)). Secondly, I have a more user friendly version of the same formula; 2^((60-AL)/40), or as a percentage 100*2^((60-AL)/40). Not that it makes much of a difference, but it is simply more pleasing to the eye. Yes, my previous formula was incorrect, but only due to it being sqrt(2) and not 1.4. Sorry for believing I was correct.
 * Also, I have now corrected my formula, and here is what it looks like. Not that you will need it, but if you could perhaps look for any more mistakes? I know I am still missing damagescalers (this includes weapon attribute) and armor penetration. But as I said, this is my own "for fun" project.


 * But anyways, are these formulas correct?
 * RealDamage=2*BaseDamage*sqrt(2)*2^(-TargetAL/40)
 * CriticalDamage=2*BaseDamage*sqrt(2)*2^((-TargetAL+20)/40)
 * TargetAL=(-20*(2*ln(RealDamage/BaseDamage)+ln(1/8)))/ln(2)


 * 2 * sqrt(2) = 2^(40/40) * 2^(20/40) = 2^(60/40)
 * So your RealDamage formula can be re-written as BaseDamage * 2^[(60-TargetAL)/40] which is more pleasing to the eye. And yes, it is "correct" as long as you realize there are things that are missing.  -PanSola 14:16, 24 April 2006 (CDT)

A fifth damage modifier type
The flat +damage vs creature type bonus on some crafter weapons does not follow the same rules as DShift, instead apparently adding to RD. axe mastery 12, all other warrior attributes at 0 (warrior secondary) LA axe = battle axe (6-25, 7 req, damage +3 vs undead) from Grahame in Lion's Arch collector axe = battle axe (6-25 slashing 8 req +15% stanced) from Crispin Bryllis in Silverwood -- Gordon Ecker 23:11, 3 June 2006 (CDT)

Strange problem with daggers
I have three sets of daggers that when I was testing always deal +15% damage (two that are +15% -5 and one with +15%^50). One has a req of 12 and deals slashing damage. One has a req of 10 and deals piercing damage. The req 12 will deal 32 damage as a critical hit and 39 with 20% sundering. The req 10 will deal 38 damage as a critical hit and 47 with 20% sundering. The third is Ceremonial Daggers and deal 32 on a critical hit. The target is a Suit of 60 Armor in the Isle of the Nameless. And finally my attributes are 16 dagger and 13 critical.

I have no idea how this is happening. It appears my req 10 daggers are dealing 18.75% additional armor penetration always. If this is true then it could mean I've found a pair of daggers that might work like Hornbows. The req 10 ones seem to attack less often, I did 4 tests of 20 seconds each showing 20, 22, 18, and 24 as the number of hits. And 23, 24, 24, and 21 as the number of hits from the green daggers. Can anyone make anything of this or should I test them for longer periods of time? (T/C) 02:22, 5 June 2006 (CDT)