1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.
  2. Discord Support team Facebook l2central.info

[Guide Contest Winner] Ghost Sentinel

Discussion in 'Guides' started by -V77-, May 17, 2018.

  1. Leckter

    Leckter User

    Joined:
    30.11.11
    Messages:
    281
    Likes Received:
    145
    Mind explaining the logic behind the enchant routes break vs focus ?
    No matter how I think about it, it feels wrong to me. I could get full break or full focus, but why splitting the aoes ? Or maybe there is something I don't know behind.
    Thanks
     
  2. Palkah

    Palkah User

    Joined:
    25.08.15
    Messages:
    502
    Likes Received:
    295
    I'm not the OP but the logic is that single target skills have Break, and AOEs have Focus.

    1. In Olympiad Break is much better because you don't have a lot of debuffs to give to the enemy, so that P.Def ignore is brilliant.
    2. Single target skills either crit or they don't, so a skill +20 Focus deals the same damage as +0 when it crits, meaning it's a wasted enchant, while Break gives you extra damage in all situations.
    3. Why are AOEs not on Break then? They deal a lot of damage by themselves, and for AOEs it's different. When you hit 20 people with an AOE, crit is calculated for each victim, meaning you can crit on 0-20 people. Increasing Skill Crt Rate here increases the number of people you crit on, and greatly increases your total damage from that AOE if you look at all the people you hit and not just single targets. There is one more reason why these are better than Break - on full skills AOEs often one-shot or two-shot weaker players in sieges when they crit, so critting on more people means more chances to oneshot and more chances to win the PvP faster.
     
    SnoWyz likes this.
  3. Leckter

    Leckter User

    Joined:
    30.11.11
    Messages:
    281
    Likes Received:
    145
    Well, I understand what you wrote, and yet it still feels as wrong as before. Probably lack of numbers.
    Let me write down my reasoning :

    From the very beginning there are only 2 scenari in my book :
    - crit chance >= 100%
    - crit chance < 100%

    Nowadays, with all the crit multipliers everywhere, I guess the dmg from normal hits are irrelevant and you want to crit as much as possible (or full str build would still be a thing, and it's not, right ?).
    Illustration with fictive numbers :
    I would rather hit with focus 90k 90k -> target dead, than with break 110k 40k -> target alive. I would gladly lose break than my double dmg + crit multipliers. It's nowhere comparable.

    Still, it depends on the effect of the focus enchant, if it brings 1% effective chance... then there is no discussion. But the same applies for aoe then. Thus my wondering.

    Also, what's the "normal" crit rate for an archer ? Is it close to 100% (I don't know serious question here, lost track of actual "standards") ? If it is, no point for the focus aoes. If it's not, better get closer to it even on single target (IF the increase is actually comparable to other boosts, see question already above).

    To me, on crit builds, you cap crit, and THEN only you add other bonuses (break) on top of that to scale the multipliers. No point to hit harder base if you don't crit.

    I guess it's all down to numbers, which I don't have anymore (way too many items and multipliers I can't have, but still can tend to it). Understanding the finality helps building and prioritizing.

    Also, if someone has the figures and calculations, can pm me if he doesn't want to "publish", I'm not any threat for anyone, just a random dude liking to "understand" :) (not looking for lessons though, I can very well think by myself :p, I don't want your choices, I want the reasons for those)

    Thanks !
     
  4. Palkah

    Palkah User

    Joined:
    25.08.15
    Messages:
    502
    Likes Received:
    295
    I'd love to test around every skill but it's not possible.

    There is no 100% crt chance and never will be. Even with POM and 101 I don't expect you to reach more than 60%, but for the exact values you need an archer player (which I'm not). Archers take DEX up to 90 and then all goes to STR. As for the Focus on single target skills - it helps, yeah, but you won't feel it that much difference in crit on a single target skill, while break will help you always. The idea for an Archer rn is that when u actually crit, it's tons of damage, and your chances to crit two times in a row are relatively high anyway.

    You don't feel Focus on on single targets, but you feel it on AoE - it's a difference if out of 20 targets, you crit 7 or 11. And the Phoenix Arrow is strong enough to take down multiple people with a single cast - assuming it crits on them
     
  5. Leckter

    Leckter User

    Joined:
    30.11.11
    Messages:
    281
    Likes Received:
    145
    60% full skills ? I would have thought it could be higher. That's the kind of numbers I'm looking for though.
    As for your 20 targets, that's not helping. If from 7 crits you jump to 11, that's about a 50% increase rate. Also, if from 20 pinpoint, 11 crit with focus vs 7 without, well... I'd seriously consider focus :D
    Or maybe focus doesn't give same values on all skills (or even they could have different crit base ?).

    So, again... numbers ?
     
  6. Takanodan

    Takanodan User

    Joined:
    22.08.13
    Messages:
    860
    Likes Received:
    357
    Last time I tested on gainak, it felt like it was closer to 80-90% on full skills on dragon bow, but I could be wrong, since we didn't test enough (only around 30-40 hits, instead of 300-500)..

    About break vs focus, most ppl I see are going for break. Normally you either go for the one-shot (maximize damage) or try to boost the damage per sec (90k + 90k is better than 130k + 30k), but yea, you'd need to see the math to check which one gives more dps..
     
    Leckter likes this.
  7. Palkah

    Palkah User

    Joined:
    25.08.15
    Messages:
    502
    Likes Received:
    295
    Can't provide these. I don't think anyone can, judging by how the only test out there is spamming at least 1k skills in each buff state to see what changes. It's not really 50% increase, it's a modifier of 20% in skill crit rate formula which - before you ask - is unknown. All that is known is that Physical Skills in theory all have the same chance to crit (or 0 if they are non-crittable), and it all depends on your DEX, PoM, 101 skill (if your 101 skill gives skill crt rate), and passives, and Abundance buff (and I could've missed other items).

    I can give you very basic maths though

    1. When you crit, the damage is not double, but more, when you realize all the passives, new crit rings, 101, PoM etc - and all that scales through skill power and P.Atk.
    2. A long time ago, 1 STR increase gave 1% P. Skill. Crt Dmg. increase, but with so many changes to the base stats over the years (some hidden) I cannot honestly say if it still applies

    Meaning that when you crit with all that, it's at least 3x or 4x the damage of a regular hit (a good archer can tell you the exact number)

    Now let's assume you use a Phoenix Arrow on 20 people, and crit 5 of them for easier calculation. Normal hit 20k, crit 80k.
    Your total damage is 5*80k + 15*20k = 700k damage.

    If you increase your skill power by 20% (+20 skill), then we can take a basic model where it increases the damage directly by 20%. (I'm talking SP instead of Break because its much easier to calculate anything, but you can assume that +20 Break with 13% P.Def ignore can safely boost your damage by that sweet 20%). For that scenario it would increase your total damage to 840k, with 24k normals and 96k crits.

    Now lets make a simple calculation, where you take +20 Focus.

    We'll take the old values of damage: 80k and 20k, and make number of crits a variable X.

    Our total damage, depending on the number of people we crit, would be:

    X*80k + (20-X)*20k = 80kX + 400K - 20kX = 400k + 60kX.

    Now let's solve an inequality of when this damage is more than the 20% Damage version.

    400k + 60kX > 840k
    60kX > 440k
    6X > 44
    X > 44/6
    X> 7 + 1/3

    from where we conclude that if X belongs to Natural numbers, X= ceil(7+1/3)=8.

    Observation: the more damage your crits do compared to your non-crits ,the lesser that number is (you need a less 'people to crit on' to deal more total damage from AOE).


    Meaning that if from 5 people on +0, the skill is now critting at least 8 people on +20 (which is probable), it deals more damage to the whole 20 people group overall.

    I cannot generalize accurately without knowing the Skill Crt Rate formula, but we can do something like this model below.

    'Damage equation' when you boost damage by 20% (X still being number of ppl you crit on):

    (1). (X*80k + (20-X)*20k)*1.2

    'Damage equation' when you boost skill crt rate instead of damage, where we denominate the increase of crt rate as parameter P (P>=1, so if you crit on 10% more people it equals 1.1, 20% more it's 1.2 etc):

    (2). X*80k*P + (20-P*X)*20k


    Inequality of (2)>(1) doesn't give a nice result, so let's just plot these with some P values.

    [​IMG]


    The points of intersection with the black function are the points where Crit starts to be better. Now don't be alarmed just yet. The values of 20%, 30% and 40% are relative here - meaning that if you critted on 5 people before, you'd need to crit on 6, 6.5 (xd) or 7 people now. Is that possible? Yes it is. If you crit on 5 people out of 20, thats 25% Crt Rate. If we assume that +20 enchant simply raises that by adding +20 (and it doesn't but it's an OK approximation for our case), we 'get' 45% Crit Rate. 45% Crit rate on 20 people is 9 people (that's close to double more people affected). Even if if was half of that, it's still 35% -> 7 people out of 20, so more than we need.

    Observation: the bigger the ratio of your damage when you crit to the damage when you don't, the faster these Focus lines overscale the Skill Power line.
    Observation: if you divide X from the plot by 20, you get what seems to be a absolute Skill Crt Rate which you can apply to any number of targets hit as a percentage, and the plot will still work.

    You wanted maths. That's the best I can honestly do with how much is hidden about this game, but it more than I've seen anywhere on EU forums. It's only for AOEs though.

    EDIT: I nowhere claim superiority of Focus vs Break. It's all a matter of preference - as we noticed, Break will help you deal more damage with a single crit, while Focus will help you to crit on more people. If you have all the gear in the game though, and as some people say that equals to 80% Crt Rate, it's highly unlikely that 20 focus will be changing that to 100%. So the more stacked you are, the more I'd go towards break, so when you actually crit, it hurts. This guide was made by an archer who doesn't have everything, and on his gear level, where he crits less people, Focus is indeed better on AOE on sieges, where his goal is probably to maximize the damage on a group of people, to help the top carries. Towards the endgame, you want burst on Archer, where it goes down to that 'crit on less people, but kill those that were critted on', and that's where I'd take Break.

    If you are not even close to that full-gear toon, then I'd go for Focus, since your hits from AOE won't 1-shot anyone anyway, but at least you'll crit on more people, causing strain on enemy healers.

    That's kinda all I can say in the topic.
     
    Last edited: Sep 11, 2018
    Hax likes this.
  8. iKepler

    iKepler User

    Joined:
    19.09.17
    Messages:
    37
    Likes Received:
    21
    Just one problem there, Skill Power enchant increases base Skill Power (what shows on skill description) not total Skill Power (base Skill Power + P.Atk). Also for some reason from what I tested this bonus from Skill Power enchant doesn't count towards crit damage, so Break = Skill Power on normal hits but Break >>> Skill Power on crits.
     
  9. Palkah

    Palkah User

    Joined:
    25.08.15
    Messages:
    502
    Likes Received:
    295
    True, but only 30% of P.Atk is added to Skill Power for Damage Calculation step

    It has to, you probably couldn't test properly due to random damage.
    It's
    Skill's Base Power + Enchant Power + 30% of P.Atk + other SP modifiers from buffs
    and then it's taken through the Skill Critical Damage in some way
    Then that's taken to calculate vs enemy P.Def and Skill Crt Dmg reductions
    Then there's a modifier of PvP Damage vs Defense
    Then there's a modifier for attribute difference
    and finally to the last result, Random Damage is applied, which for bows should be +-5% but we have no idea if it applies to skills too.

    One more important information: Break does give more damage than Skill power with high enough P.Atk, so Break is always better than Skill Power the more gear u have
     
    Last edited: Sep 11, 2018
  10. iKepler

    iKepler User

    Joined:
    19.09.17
    Messages:
    37
    Likes Received:
    21
    From what I remember Break and Skill power normal hits damage were around 1% average damage difference from each other (not statistically significant cuz of random damage) while Break crits were around 7% more damage than Skill Power.

    Could be really lucky/unlucky with random damage or damage formula doesn't work like you think.
     
  11. Palkah

    Palkah User

    Joined:
    25.08.15
    Messages:
    502
    Likes Received:
    295
    It does, the problem is that it all depends on your P.Atk.

    The more P.Atk you have, the more Break bonus gives vs Skill Power bonus. And then that difference is SCALED through all the crit modifiers when you crit, so yes, your results could be correct in general, if not for this particular value.

    7% increase is hard to be sure of, where +-5% random means that 100k hit is in range of 95k-105k, while a 107k hit is in range of 101.65k to 112.35k. And that's assuming random for Skills is the same as your weapon's random, and not, let's say double of that. All I'm saying is that due to uncertainty of the measurments, it's hard to give a real number. What we really need as a curve of P.Atk vs Damage of a Skill in both conditions to legally say any numbers except for "Break is better (and it is)".

    All of that could of course be tested, if @Yumi hired me as an Experimenter xD
     
  12. Leckter

    Leckter User

    Joined:
    30.11.11
    Messages:
    281
    Likes Received:
    145
    It's all nice and fine, but still doesn't bring what I was asking for :D Thanks for the effort though, appreciated the maths, and still highly interesting for the main topic.
    My concern is that, as you said, you don't know the formula, and noone knows, plus you talk only about aoes, when I was questionning why 2 different routes on same char.

    Also, end result is not unique solution, DPS is not necessarily linear with kill speed, so not necessary the better choice, same for burst damage. Every one has it's own balance sweet spot for that, because of tasks, stats or objective. Let me find myself what's the best for me :)

    Taka's elements were actually more interesting to me.

    What's the crit rate without focus nor boosts of an endgame archer ?
    What's the crit rate without focus with full boosts ?
    What's the crit rate with focus without boost ?
    What's the crit rate with focus with full boosts ?

    That's what I need. I don't need formulas, noone knows them. Give me the end game results (with little context like how much dex and what gear for scale evaluation, could be the best), and I'll elaborate my formula suited to my needs :)
     
  13. 21Sav

    21Sav User

    Joined:
    02.04.18
    Messages:
    241
    Likes Received:
    64
    This will all be answered only if @Palkah becomes experimenter, so #Palkah4Experimenter !!!
     
  14. Palkah

    Palkah User

    Joined:
    25.08.15
    Messages:
    502
    Likes Received:
    295
    90 DEX, rest to STR is the common consensus right now, due to softcap on 80 and hardcap on 90 (which should be retested because that's few years old info). The crit rates - nobody can really answer that because that would require doing ~500 hits in each of the 4 states for a sample representing anything, and people aren't that bored. I am, but I don't have a character with skills +20 and a ton of Giant Codexes.

    I know of at least 2 separate good archers who claimed their crit rate is around 80% 'most of the time' without the Focus enchant, on endgame gear. As for how each buff/gear exactly affects the rate - again, stuff to test.

    The whole idea of why top people don't like Focus is that they reach the 'cap' (whatever it is) with other buffs and gear, so all that's left is to increase their damage. The only 'item-related' crit boosters you need are Abundance 1 and 90 DEX, rest is buffs/passives.

    xD
    well not exactly 'becomes an experimenter', but reducing strain from @Yumi by voluntarily (and for no compensation) doing some tests which require little to none 'magic GM commands' on some dusty PTS4 (that nobody even remembers it exists) would be perfect for a start :)
     
  15. BoscoMagico

    BoscoMagico User

    Joined:
    12.03.16
    Messages:
    29
    Likes Received:
    1
    Most important question its not about break or focus. Most important question is or equal geared yul like in this post can do smth vs feoh? Becouse now we see how candy killing everything who move but he have totaly everything so we dont count that people becouse its pointles, probably you net even enought 30k eu to get what people like him have. we count char for ~ 5k eu.
     
  16. Palkah

    Palkah User

    Joined:
    25.08.15
    Messages:
    502
    Likes Received:
    295
    Yul starts lower but scales better into lategame.
     
  17. Hax

    Hax User

    Joined:
    24.11.11
    Messages:
    4,883
    Likes Received:
    841
    Like Palkah said.
    Yul without end game gear will get 1shot by feoh. Yul with end game gear (yes you even need dragon weapon to really be end game) does enough damage for feoh to not be able to keep up
     
  18. BoscoMagico

    BoscoMagico User

    Joined:
    12.03.16
    Messages:
    29
    Likes Received:
    1
    So you say i need spend atleast 15k eu to start do smht? Becouse now i make same build like in this topic and i still pointles to pvp i hit feoh ~ 4k or 5k with boost geting back 15k :D and i even dont talk on pve bcs its sad story :D
     
    Hax likes this.
  19. Palkah

    Palkah User

    Joined:
    25.08.15
    Messages:
    502
    Likes Received:
    295
    PvE on yul is pulling mobs to a train, which Tyrr does better, but there are less Tyrrs available.
     
  20. Hax

    Hax User

    Joined:
    24.11.11
    Messages:
    4,883
    Likes Received:
    841
    It is the sad truth. Archer only is better endgame, not before