Evasion Equation - Updated

13»

Comments

  • Kyzarin - Raging Tide
    Kyzarin - Raging Tide Posts: 11 Arc User
    edited September 2011

    The Evade% is rounded up to 2 decimal places (remember 0.67 = 67%). The accuracy and evasion rates of monsters can be found from http://www.pwdatabase.com/pwi/. A few mobs were found to have a 100% hit chance when their evasion was 1, hence the offset value displayed in red.

    I have seen my experience bar say 100.0% before I got to the next level, and after checking the exp numbers, I found that I was around 20 exp points short of the next level. It is likely that the 100% hit chance displayed was really 99.5% or some such (i.e. not quite 100%, but close enough to 100% that it rounds up and 100% is displayed). That leaves open the likelihood that the equation is, in fact, Evade% = Target.Evasion / (2* Attacker.Accuracy + Target.Evasion).
  • AIethea - Sanctuary
    AIethea - Sanctuary Posts: 35 Arc User
    edited September 2011
    I have seen my experience bar say 100.0% before I got to the next level, and after checking the exp numbers, I found that I was around 20 exp points short of the next level. It is likely that the 100% hit chance displayed was really 99.5% or some such (i.e. not quite 100%, but close enough to 100% that it rounds up and 100% is displayed). That leaves open the likelihood that the equation is, in fact, Evade% = Target.Evasion / (2* Attacker.Accuracy + Target.Evasion).

    The two are quite different. However, this was already checked for with Accuracy high enough to get 99.92% using the equation and it still displays 99% (Bluepaw Lemur has an evasion of 2). Only mobs with an evasion of 1 displayed 100%.
  • AIethea - Sanctuary
    AIethea - Sanctuary Posts: 35 Arc User
    edited September 2011
    Or, from a different point of view: evasion can keep your winged shell from popping.

    Or, put differently, if you could evade most attacks and you did not have enough health to survive one attack, winged shell might let you survive that one hit.

    This is what I meant by it will feel like it lasts longer but really its functionality is still the same. The real benefit of it lasting longer is that the longer it lasts the more mana regen you get from it as opposed to if it pops early. However, this is trivial in terms of the mana regen and survivability so in essence it is still doing exactly what it always has done; been a buffer to damage.

    You can make the claim that the damage it absorbs is static but increased evasion reduces the net damage you take thereby having the shield absorb a greater percent of the actual damage sustained. Defense will do the exact same thing though and it is still unclear to me which will provide more survivability.

    Basically, I don't see any unique trait or interaction from the two that warrants mentioning. Any shield (or damage buffer) and evasion provide the same functionality just because of what they are.
  • Fleuri - Sanctuary
    Fleuri - Sanctuary Posts: 1,763 Arc User
    edited September 2011
    You can make the claim that the damage it absorbs is static but increased evasion reduces the net damage you take thereby having the shield absorb a greater percent of the actual damage sustained. Defense will do the exact same thing though and it is still unclear to me which will provide more survivability.

    So, ok, maybe I am missing something?

    But for example: if your defense is high enough to keep your winged shell from being popped in one hit, when you increase your defense you are actually reducing the amount of damage being absorbed by winged shell.

    But this does not ever happen with evasion, does it?
  • AIethea - Sanctuary
    AIethea - Sanctuary Posts: 35 Arc User
    edited September 2011
    But for example: if your defense is high enough to keep your winged shell from being popped in one hit, when you increase your defense you are actually reducing the amount of damage being absorbed by winged shell.

    I think an example will best explain my reasons. Lets say a monster does 100 attacks against you before it dies each doing 1,000 damage per hit. For the sake of simplicity lets say you have 0 defense and 0 evade initially thereby taking all of the damage, 100,000 total. Now, you put on some gear and increase either your defense or evasion to 1%

    --1% Evasion--
    1000 per atk
    99 hit
    1 miss
    99,000 damage taken

    --1% Damage Reduction--
    990 per atk
    100 hit
    0 miss
    99,000 damage taken

    Net result (on average) is the same decrease in damage. Throwing a shield into the mix doesn't really change things between the two situations. Its going to absorb damage, you are going to still dodge and damage will still be reduced. Now, depending on if defense affects the shield or not (I do not know) it would actually favor getting defense.

    The main reason for determining these equations is because obtaining that 1% is not equal. They are also nonlinear with diminishing returns which means that after a certain amount one will be better to get than the other.
  • Fleuri - Sanctuary
    Fleuri - Sanctuary Posts: 1,763 Arc User
    edited September 2011
    I think an example will best explain my reasons. Lets say a monster does 100 attacks against you before it dies each doing 1,000 damage per hit. For the sake of simplicity lets say you have 0 defense and 0 evade initially thereby taking all of the damage, 100,000 total. Now, you put on some gear and increase either your defense or evasion to 1%

    --1% Evasion--
    1000 per atk
    99 hit
    1 miss
    99,000 damage taken

    --1% Damage Reduction--
    990 per atk
    100 hit
    0 miss
    99,000 damage taken

    Net result (on average) is the same decrease in damage. Throwing a shield into the mix doesn't really change things between the two situations. Its going to absorb damage, you are going to still dodge and damage will still be reduced. Now, depending on if defense affects the shield or not (I do not know) it would actually favor getting defense.

    The main reason for determining these equations is because obtaining that 1% is not equal. They are also nonlinear with diminishing returns which means that after a certain amount one will be better to get than the other.

    So.. ok... I am going to presume that we are still talking about my original assertion:
    I should also point out something that has been getting no discussion that I have seen:

    The archer skill "Winged Shell" interacts nicely with evasion, in the context of short battles. In essence, it gives you an extra chance for evasion to mean something. (And, of course, Deaden Nerves fills a similar role for Assassins.)

    ...but I do not think your example represents a short battle.

    First, your damage is way too low for a short battle. But, also, I do not think 100 hits would be "short".

    Nevertheless, you can still see hints of the issue I was speaking of, here: For example, if you remove one hit from that sequence because of winged shell, with 1% defense, your winged shell reduces damage taken by 990 hp and with 1% evasion your winged shell reduces damage taken by 1000 hp -- in other words, the effectiveness of winged shell against that hit would have increased by approximately 1% from using evasion instead of defense for your total damage reduction. Granted, that's something like 0.01% of total damage received, but even wizards have a 5% chance to evade.

    Anyways, if you are going to use an example involving 1% differences and then also claim that 1% is not worth discussing, I think you are arguing against your own example.

    So, let's try another example:

    Let us consider a battle sequence involving a constant 5000 damage for 5 hits, and let us contract a 20% evasion chance with 20% damage reduction. And, again, we are going to artificially assume one miss from evasion, much like in your setup. And let's also include the fact that winged shell only reduces damage by 80%:

    So now we have: (with 20% damage reduction):

    800
    4000
    4000
    4000
    4000

    Total damage taken: 16800

    And, with 20% evasion:
    1000
    0
    5000
    5000
    5000

    Total damage taken: 16000

    So, you wind up avoiding 800 damage, on average. So, for example, if you had 13k health to start out with, you would take two hits after your charm ticks for the damage reduction case where you would take one hit after your charm ticks for the evasion case. (Of course, this charm tick thing involves luck: if we changed the battle slightly both would still take two hits after the charm tick so the "high defense" case comes out roughly 20% ahead in terms of damage taken after the tick. But a full discussion of the issues involving health charms would be too much for this thread.)

    Or, you might want to think about 10k damage with 50% evasion and 50% damage reduction on an archer with 5k health...
  • AIethea - Sanctuary
    AIethea - Sanctuary Posts: 35 Arc User
    edited September 2011
    First, your damage is way too low for a short battle. But, also, I do not think 100 hits would be "short".

    I will go more in depth when I have time, but in regards to this...it was just using easy numbers. That is not short but we are talking about an average which makes more sense to discuss something of random chance. Most people find it harder to dismiss low probability chances in a large set whereas in 5 hits people tend to think...well what if you don't dodge? So, 100 hits, 5 hits, the example looks the same to me so we can use yours if you like.

    However, one very important issue is that Winged Shell doesn't just provide you with a blanket 80% reduction for one hit. It provides a specific amount total that can be shielded. The 2nd thing which I was touching on is that I don't know defense affects how much damage your shield absorbs. The shield might absorb raw damage or reduced damage and I do not actually know, do you?

    A final note, you have just assumed the best case scenario, you get hit while your shield is active. If you start getting high evasion it becomes a very realistic problem that you may / will get dodge everything while it is on and get that one unlucky hit while it is off. This fits within your short battle frame I suppose but I am pointing that out.
  • Fleuri - Sanctuary
    Fleuri - Sanctuary Posts: 1,763 Arc User
    edited September 2011
    However, one very important issue is that Winged Shell doesn't just provide you with a blanket 80% reduction for one hit. It provides a specific amount total that can be shielded.

    So, ok, we might need to discuss the mechanics of winged shell:

    And I believe you are interpreting a bad translation from chinese to be the actual mechanics of the skill.

    In practice, though, that specific amount is the health of your shell. Until your shell absorbs that amount of damage, or until it times out, it stays up. However, on a hit when that health is exceeded, it gives you its full 80% damage reduction, regardless of how much damage you take on that hit. Of course, once it's gone, it no longer protects you.
  • prof
    prof Posts: 1,111
    edited September 2011
    don't believe pwdatase's accuracy thing, every hh boss has 100k accuracy lol

    in my experience as a warrior, i usually tried to get about 1/5th more accuracy than archers could get dodge, that ensured i'd hit at least 9/10 hits, usually not missing any. if i had to estimate, if your accuracy = your opponents dodge, your chance to hit is somewhere between 60~80%
  • AIethea - Sanctuary
    AIethea - Sanctuary Posts: 35 Arc User
    edited September 2011
    However, on a hit when that health is exceeded, it gives you its full 80% damage reduction, regardless of how much damage you take on that hit. Of course, once it's gone, it no longer protects you.

    This has not been my exp when using it but I have not really looked into it. I shall put it on a to do list though. If that is indeed the case it changes things a bit esp when going against very large attacks. It has nothing to do with translations though, the 80% reduction nor this behavior is mentioned anywhere.
  • AIethea - Sanctuary
    AIethea - Sanctuary Posts: 35 Arc User
    edited September 2011
    prof wrote: »
    if i had to estimate, if your accuracy = your opponents dodge, your chance to hit is somewhere between 60~80%

    You know, if you actually tried the equation you'd see that it gives 66% hit if your accuracy = their dodge so you are giving credence to my theory.
  • prof
    prof Posts: 1,111
    edited September 2011
    fleuri is right, but i'm not sure about the 80% thing; i've turned 20k+ mage hits into 800s lol

    as for your formula, i'm not disagreeing with it, but i also can't say it's 100% credible. i can say this, because i know accuracy has a lot to do with luck as well. where i've used characters with 10k accuracy and missed several hits in a row on ~300 dodge mages, and i've used mages that can slash archers with over 6k dodge with under 100 accuracy, i have reason to believe your % to hit or miss is capped somewhere; maybe 90-95%.
  • AIethea - Sanctuary
    AIethea - Sanctuary Posts: 35 Arc User
    edited September 2011
    prof wrote: »
    i have reason to believe your % to hit or miss is capped somewhere; maybe 90-95%.

    Asterelle has been alluding to this as well when scanning with her belt but hasn't really come out and confirmed it yet (5% min evade). This was also specific to PvP not PvE as far as I've understood.
  • prof
    prof Posts: 1,111
    edited September 2011
    95% sounds more logical, based on in-game testing. I know very well that I don't miss 10% of my hits on mages; nor does my mage slash others 1/10th of the time.

    crit caps at 95% as well. pdef, mdef, def level, atk level, and cast all cap at 90%, though.
  • AIethea - Sanctuary
    AIethea - Sanctuary Posts: 35 Arc User
    edited September 2011
    prof wrote: »
    95% sounds more logical, based on in-game testing. I know very well that I don't miss 10% of my hits on mages; nor does my mage slash others 1/10th of the time.

    crit caps at 95% as well. pdef, mdef, def level, atk level, and cast all cap at 90%, though.

    I would love to know where you get this info, if you have a source / link. Heresay doesn't mean much =/.
  • truekossy
    truekossy Posts: 7,021 Arc User
    edited September 2011
    I would love to know where you get this info, if you have a source / link. Heresay doesn't mean much =/.

    The one on crit seems accurate based on what shows in character menu when people use 100% crit rate skills

    The rest aren't, though.
  • Fleuri - Sanctuary
    Fleuri - Sanctuary Posts: 1,763 Arc User
    edited September 2011
    prof wrote: »
    i'm not sure about the 80% thing; i've turned 20k+ mage hits into 800s lol

    I do not remember where I got 80% from, and I have not tested that recently, and I could be wrong about the damage reduction percent you get from winged shell.

    However, I also believe that damage reduction from winged shell stacks with other damage reduction mechanisms. And, I believe I tested that. I just wish I remember when and where I tested that.
  • prof
    prof Posts: 1,111
    edited September 2011
    I would love to know where you get this info, if you have a source / link. Heresay doesn't mean much =/.

    i use to decompose the pw server+client files(i'm sure there's some not-so-legal forums you could browse to find similar variables), then translate them into legible english some time in the past; i lost all my previous documented saves after a hdd crash, and useless mediafire..

    can you believe mediafire deletes what you upload if it goes 'inactive' for 90 days? what's the point of online storage if it's just gana be removed anyway. b:sweat

    anyway. i know defenses are capped at 90% reduction, i know crit is also capped at 95%, and you can see 90% cast in-game using that eye thing. these are all long confirmed(interest fact: cast is capped at 95% in myen - weird that it's 90% in pwi).

    somewhere on the forums, either in psy or sin forum, there was long discussions about atk/def levels; some still seem to believe it caps at 120, but that's illogical based on previous caps.

    as i cannot provide evidence, i'll speak from experience: i know a sin ingame, full rank 9, 4s set full of +1 atk stones. he's holy. he has about 160 atk levels with chill of the deep.

    ..yet he doesn't hit any higher than a sin with 98 atk levels.

    this was done in the raging tides server.

    this is my reason to believe it caps at 90. i'd confirm it, but as i've lost all the decompilers i wrote for pw, i cannot
    truekossy wrote: »
    The one on crit seems accurate based on what shows in character menu when people use 100% crit rate skills

    The rest aren't, though.

    disprove any of them, lol
    I do not remember where I got 80% from, and I have not tested that recently, and I could be wrong about the damage reduction percent you get from winged shell.

    However, I also believe that damage reduction from winged shell stacks with other damage reduction mechanisms. And, I believe I tested that. I just wish I remember when and where I tested that.

    i'm not sure if the shell has a total reduction of damage or not - i've never looked into it. if i had to wager, based on the the skills level(the dark melee skill that gives you a level 5 barrier vs. level 11), both offer the same reduction when hit with something that would break it in 1 hit. this leads me to believe you will always take a maximum damage, but i know this also isn't true.

    b:shutup $5 says the skill is glitched
  • Quilue - Sanctuary
    Quilue - Sanctuary Posts: 5,787 Arc User
    edited September 2011
    If 90% is the most reduction you can get from defense, it explains why BMs have died to sins with Buddha's Guard active.

    Some believed that it was condensed thorn, but I've always doubted that you could be killed from being struck by a faction of a weapon damage...
    Elena Costel: I wash my hands of this affair.
    Legerity: *drags you back* *stains your hands with said affair*
    Elena Costel: Noooo... I don't want to have a dirty affair with Lady Legerity...
    Qui: b:dirty
  • prof
    prof Posts: 1,111
    edited September 2011
    +12 striking dragons on the physical immune boss in fb99 abbadon with fist ultimate(demon version, +135% patk in fire vs. +50% patk in water), you hit less than 200 lol

    really doubt that buff is helping sins kill anyone whose resisting 90% of the phys atk
  • jlstewarts
    jlstewarts Posts: 1 Arc User
    edited November 2011
    For every 15 evasion points you have on your gear equals 1% on your evasion plus base evasion ,so if u wat to be un-hittable in pve you need 680 evasion sharded armor,using all peices u have ,680=41.33333 evasion percentage added making you untouchaable.
    ignorethe formulas this is the correct equation
    thank you
  • choco972
    choco972 Posts: 0 Arc User
    edited December 2011
    Hi, i'm lvl 84 and i have make some test and i have 5300 evasion :

    when i select a huggy hare i have 42% evade, when i put the genie buff who give me extra 10% (482 more evasion) the evasion grow up tu 44%, when the buff end or i remove a piece of armor who give enter 300 and 400 evasion, the bonus down to 40% (for 300-400 evasion added/removed u earn/lose 2% evasion luck)

    But i go on the tree lvl 21, my evasion is 70% but when i remove 300~400 evasion i lose 2% for every gears i remove (68% => 66%, etc), but surprise if i add 482 evasion, i earn only 1 %... maybe evasion is nerfed after 70% no ?