[Mechanics] Davlok's Random Testing Shack

Monk
Prev 1 30 31 32 Next
11/28/2018 12:32 AMPosted by StoleOwnCar
AD applies damage to all mobs within a certain radius as long as you're the damage source. It scales with density within a radius, and isn't limited to the 6 mobs that receive the bonus (shouldn't this be 7 though? Too tired to check...). Its damage is effectively only limited by how many mobs you can gather in one area (preferably within 10 yards of each other).


https://www.youtube.com/watch?v=ydMEhQaULu4
Your premise is totally wrong on the statistical end to begin with. 2 Targets within 10 yards of each other actually have a 36% chance to trigger AD at least once, not 20%. Please look up Binomial distribution and its cumulative probability functions. Furthermore the probability of success, as I said, increases as you add more targets. So you actually have 10 targets around you, within 10 yards of each other, and you are calculating the worst case scenario damage of AD, it's 0.73*9*150. Actually it's more than that because I thought Scarbringer was 6 targets. It's actually 7, which is 0.79 as the chance.

So let's calculate things out.
Use this handy calculator:
https://stattrek.com/online-calculator/binomial.aspx

Technically if Wolfram alpha could iterate over a binompdf, this would be easy, but I had to put each number of successes in manually since my graphing calculator from college is dead now ( :( )
1=0.367
2=0.275
3=0.1147
4=0.029
5=0.004
6=0.00036
7=0.000013


By the way the probability of at least one success is 0.7902848. If you add up all of the probabilities I posted above you'll get 0.790073 (due to rounding). It's safe to use these directly in calculation because they can't exceed the base number of at least one success, even if the number of trials approaches infinity (iirc).

So the amount of damage the entire mob of 10 would take, given one success, is (N-1)*1.5=9*1.5=13.5. If you had 2 successes you would do 2*13.5=27, and so on and so forth, because each AD originator can get hit by external procs of AD, so only one in the mob doesn't get hit per proc of AD. So let's distribute that number across 7 trials and add it together:
1*13.5*0.367+
2*13.5*0.275+
3*13.5*0.1147+
4*13.5*0.029+
5*13.5*0.004+
6*13.5*0.0036+
7*13.5*0.000013
=
19.1536785


If you're doing your "per target" damage "modifier", then 19.15/10=1.915 "modifier" per target, I guess. But raw damage wise AD did over 19 times the amount of damage each kick did, when distributed among the mob.

The thing is that you can also clearly see that we can move the N-1 out of the series of additions to the side, to get an idea of how this function scales with density, which the answer is "pretty much linearly". ie f(x)=(x-1)*(1*0.367+2*0.275+3*0.1147+4*0.028+5*0.004+6*0.00036+7*0.000013)=(x-1)*1.395351 for the total raw damage done by AD, where (x-1) is a stand in for the original N-1.

You can view the plot of that here:
https://www.wolframalpha.com/input/?i=(x-1)*1.395351,+2%3E%3Dx%3E%3D40&assumption=%7B%22C%22,+%222%22%7D+-%3E+%7B%22NumberMath%22%7D

It doesn't take many targets at all for the total amount of damage done by AD to greatly outscale the amount of damage done by the original LTK, which is helped by the fact that most of LTK's damage is limited to just 7 targets. The thing that limits this most is mob density, just like with many AD centric builds. The closer the mob density the more probability that they all proc AD and get hit by it simultaneously. That is why mob types like swarms are greatly values for greater rift progression. It's also noteworthy that even if they don't get hit by the bulk of LTK's damage due to being over scarbringer damage, additional targets can trigger non-Scarbringer AD to still do scrape damage to the entire mob.

The original damage by just the 7 LTKs is, well, 7, compared to the 19 I did above for the AD procs. So 7*3 for the bracer is 21. At just 10 mobs the original LTK's area damage already almost exceeds the total amount of damage off Binding. At 20 mobs, it does much more.

11/28/2018 03:10 AMPosted by Balbero

https://www.youtube.com/watch?v=ydMEhQaULu4


Obviously the above post is not in response to you, but I'll note that I simply read a forum thread about it here, just in case:
https://us.battle.net/forums/en/d3/topic/12032383478

It works about as I expected it to.

It takes a very high multiplier to outscale AD. Chantodo was an example, but it didn't really "outscale" anything, it was just useful for speed runs. LoN Doc is the best example, and I don't really even understand how that spec worked, to this day. It was just messed up.

I'm not 100% sure how the ad calculations work with WoL, but if WoL has a higher multiplier on its normal attack, it will greatly outdamage LTK considering the number of attacks that can proc AD on a mob are essentially unlimited. The number of trials goes to like 40 or 50, and all of them do the same damage.
11/28/2018 07:21 AMPosted by StoleOwnCar
Your premise is totally wrong on the statistical end to begin with.

No

11/28/2018 07:21 AMPosted by StoleOwnCar
2 Targets within 10 yards of each other actually have a 36% chance to trigger AD at least once, not 20%.


Right... but, the different approach is, that you re calculating the total AD and i m calculating the single target potion of AD

concerning the 36%:

chance to proc:
T2 ......... T1
0,2 ....... 0,8 --> 0,16 AD for Target 1
0,2 ....... 0,2 --> 0,04 AD for Target 1 AND 2
0,8 ....... 0,2 --> 0,16 AD for Target 2
0,8 ....... 0,8 --> 0,64 no AD Proc

the chance for at least one AD proc is 36%
and the chance for Target 1 is 20% and for Target 2 20%

so you can go on with 3,4,5...

both calculations are valid depending, what you want to calc
Total AD or the single target modifier
for me the latter one is more interesting

11/28/2018 07:21 AMPosted by StoleOwnCar
If you're doing your "per target" damage "modifier", then 19.15/10=1.915 "modifier" per target, I guess.


not correct! you re underestimating the single target part:
Per Target > Total AD/Number of Targets :)

11/28/2018 07:21 AMPosted by StoleOwnCar
I thought Scarbringer was 6 targets. It's actually 7, which is 0.79 as the chance.

Right, but the result is: 6 Scarbringer-Buffed AD Procs

11/28/2018 07:21 AMPosted by StoleOwnCar
At just 10 mobs the original LTK's area damage already almost exceeds the total amount of damage off Binding. At 20 mobs, it does much more.


much more is relative, due to the massive drop when hitting the Scarbringer Limit
from 7 to 10 is around 4%
Okay, I have no idea wtf you're talking about anymore, and frankly I don't care. I gave you pretty concrete math relying on statistical distributions, proving how, in mediocre-case scenario mob sizes, AD easily beats both the original LTK damage and the proc damage scarbringer would have.

I don't know if you're trying to treat LTK as some sort of ST elite killer or RGK build, but it never has been that. It sucks at that. There's no point setting it up like that. Without a significant amount of AD bouncing around inside of decent mob sizes, you're losing a lot of damage. If the clone can't proc AD, it gets easily outscaled. Like period. There's no way to argue this, unless you're just engaging a low amount of targets on purpose. In which case, you're probably not pushing anything worth a crap.
I cleared a 125 on Inna's woL (RS no BOTLG)

BOTLG = no Tzo krins, no crudest boots, or no pintos, Tzo Krins would be the only thing I'd even consider dropping and I did manage to clear a 122 or so with botlg no tzo krins using a melee lightning pillar build but it was much harder to play trying to use assimilation to outweigh the additive damage column in favor of tzo krins + zeis, which turns out to actually be a higher damage multiplier.. and part of that might be that

so the clone does not proc area damage? ...

K so ^ clone doesnt proc area damage, area damage was still super worth having a decent amount of for s11 Lon LTK BotLG even tho the clone did do a higher amount of initial damage.

Guys are arguing like area damage is so cut and dry... the situation you are in specific to any moment will change how the area damage will act because mobs are constantly moving, depends on the attack used, its arc, depends on proximity of the mobs, hit box sizes, mob type.

Also any stat needs to be argued in terms of its strength versus other affixes and each individual affix worth in total damage % is what is important but any stats value of % damage is lessened per unit as you gain more of that stat.

If a build can proc area damage you most likely want a solid ~100% at least, some builds you want to completely max out area damage, except on a spot like neck, where you'd be giving up 10% chc, 100% chd, or 20% elemental, yet there are even some builds where it could be argued 20% area damage is wanted on neck and not just those that purposely negate crit hit chance to gain more of it later (broken promises) or gain it by other means (shi mizu, Znec corpse lance, iceblink, barbskill) an example of this is IMO the best Ammy you can have for Garg Doc is 10% crit 100% crit dam 20% area damage because you are using Mask of Jeram which is additive with elemental as enforcer USED to be, and garg swing in an arc that heavily favors area damage. naturally you will still get 20% on bracers cuz theres no other damage stat but the 220->240% = 3.4 / 3.2 = 6.25% damage which is by no means negligible but that 20% area damage will most likely outweigh it, but then again maybe you struggle on the rg so you decide to go cold over AD.. While an LON spirit Barrage, that has multiple cold % on it because its LoN, so you got frostburns, andy's, swampland waders, bracers, ammy and it still has the additive with pet helm 200% (3x) so Ave damage can even beat it out when you have enough cold % on things despite ave damage only being the same amount as you can get on rings. The fun part of this build is when you start to consider attack speed being incredibly important until you have too much or you dont know how to attack with attack speed over a capped amount because you can only have 4 phantasms out at once and ive gone crosseyed.
11/28/2018 09:54 PMPosted by Darth
Guys are arguing like area damage is so cut and dry... the situation you are in specific to any moment will change how the area damage will act because mobs are constantly moving, depends on the attack used, its arc, depends on proximity of the mobs, hit box sizes, mob type.


first gz to the clear!

maybe cutting is the wrong word, just a "term" to somehow handle the Scarbringer

it acts for the 1st 7Targets, hitting 8 Targets or more, the latter one gets an unbuffed damage and that corresponds to a lower area damage contribution (by 1/7) if u re add more and more targets

the damage is still increasing, but not that fast as without the Scarbringer-Target-Limitation
the 2nd part is, that all the area damage procs re coming from the lower non-mimic-kick

of course we (or in cause of the area damage bonus to a single target) re just calculating perfect conditions
but nevertheless it smears out a bit, if these conditions are not fullfilled

and i agree, more AD, more total damage, just try to figure out the efficiency

edit:

11/28/2018 02:09 PMPosted by StoleOwnCar
Okay, I have no idea wtf you're talking about anymore, and frankly I don't care. I gave you pretty concrete math relying on statistical distributions, proving how, in mediocre-case scenario mob sizes, AD easily beats both the original LTK damage and the proc damage scarbringer would have.


as i mentioned, we re calculating the same from a different point of view and i m exactly using the math from the linked "fred" and the linked youtube video

the only difference is, you re calculation the total area damage and i m just the part for a single target, but all of them are getting that enhanced damage (on average) not just one
maybe here is the bug

if i m interested in the total area damage value, i would use your math

the only thing we both can t do is to calculate easily the single target gain from your math and the total area damage from the single target value from my math

that one can easily see in the 2 target table posted above.
the probability for each target is 20% (my math) but the Total is just 36% (your math)

that s due to the chance that both targets can proc area damage with 4%
16%+4% for each and 16+16+4 in total

hope that it now becomes clear and no one should be pissed
11/28/2018 09:54 PMPosted by Darth
I cleared a 125 on Inna's woL (RS no BOTLG)

BOTLG = no Tzo krins, no crudest boots, or no pintos, Tzo Krins would be the only thing I'd even consider dropping and I did manage to clear a 122 or so with botlg no tzo krins using a melee lightning pillar build but it was much harder to play trying to use assimilation to outweigh the additive damage column in favor of tzo krins + zeis, which turns out to actually be a higher damage multiplier.. and part of that might be that


would only make sense, if the dmg woul finally be in the orange text -.-
(how long do we need to beg for this ??? ... so much for we are listining)

then you could drop kyoshis for the innas staff, and then wear either:
aughilds helm+shoulders (would like to see here a buff frome 7% on melee/range up to 25% or even 30%)
or you could go for lefebvre's + andarials/eyeofthestorm(forlightning)
@ Darth,

btw. WOL doesn t ve the Target-Limitation and can principally hit an infinite number of targets with the same damage, additionally due to the lack of Bindings, the own Bell has the same "initial " damage as the mimicked bell
Been a while, but AD did work with the Rabid cloned attack (but haven’t checked for a few patches) and keep in mind you need to hit them with CS for BotLG to work.
11/29/2018 09:57 AMPosted by Davlok
Been a while, but AD did work with the Rabid cloned attack (but haven’t checked for a few patches) and keep in mind you need to hit them with CS for BotLG to work.


latest statement i ve found was from Hales:

https://us.battle.net/forums/en/d3/topic/20759277728#post-14

no AD, no Stricken stacks no LOH (ptr 2.6.1)

edit:
some years ago, the statement was, that AD was working

if AD works correctly, would change everything (besides the 7-Target-Problem and the reduced efficiency )
11/29/2018 11:30 AMPosted by Lutzer
latest statement i ve found was from Hales:

https://us.battle.net/forums/en/d3/topic/20759277728#post-14

no AD, no Stricken stacks no LOH (ptr 2.6.1)

He is correct that the cloned attack doesn't proc Striken or LoH (game probably considers the main attack + cloned attack a single attack?), but it does still does proc AD as far as I can tell.

https://i.imgur.com/LojUQPx.jpg
Interesting. If something else wasn't happening (some weird interactions not involving rabid strike but making it seem like it was rabid strike), that actually goes a long way to explaining why people running Rabid with WoL are getting so much mileage out of it. It's practically doubling damage. I wonder if they could drop Tao Krin's and then cube the lost bracers while gaining damage. LTK could run that without dropping its helm, I think.
tzo gives factor 2.5 if optinally rolled. the bracers give your clone.faxtor 3(optimal), which means you get factor 2 compared to rs without bracers
11/30/2018 12:34 AMPosted by Atomfurz
tzo gives factor 2.5 if optinally rolled. the bracers give your clone.faxtor 3(optimal), which means you get factor 2 compared to rs without bracers


Nah, it's not 2.5. It's just a standard WoL increase damage skill modifier. Skill modifiers go into the same bin as other additive increases, so it's actually far from 2.5. It's not even 2.

I just pulled up a random WoL profile I found on d3planner and the damage difference was about 555 vs 952, so depending on how many additive bonus sources you have it only counts as as about a 71.5% damage increase. Or alternatively basically you're doing 58.3% of the damage that you would with it, if you didn't have it on.

So with Rabid, WoL without Bindings of the Lost does basically 2 full WoLs of strength 1.
1+1=2

Without Tzo but with BotL, WoL does 58% of the above on its base attack and then 3x that on cloned attack.
0.58+3*0.58=4*0.58= 2.32

On the other hand you lose Zeis. I'm not sure if Enforcer works with Rabid's duplication, that's something I found in the profile I pulled up. Interesting idea.
11/29/2018 09:57 AMPosted by Davlok
Been a while, but AD did work with the Rabid cloned attack (but haven’t checked for a few patches) and keep in mind you need to hit them with CS for BotLG to work.
+ Follow up

Nice thank you I was getting confused there on the real cause I thought I had tested it myself lol.. so it follows the rules of a pet moreso than a proc.

Pets dont proc on hit effects, stricken, but can do area damage.. i guess im assuming that it procs PE, but prob trying to stay away from that for kyoshiros sake anyway..

Which BTW dav.. seems like that 2.4~ APS kyoshiro bug, from what I can tell it doesnt happen on low ms servers.. I had been wondering for a while cause top HC monk in korea uses Radiance + 2 APS rolls with firebell, when Im on a West Coast server it tends to work fine at higher APS.

As far as BOTLG w/ RS vs TZO.. and ya it'd be nice if Tzo was just a straight multiplier not % on gear which is additive for non pet skills..

So if Im gonna look at the total damage vs total damage for innas..

typically in HC, prob both there's 15% on my boots, 40% from Mystic Ally, 150% I cant think of anything else that goes into that additive column I used or would normally use..

so its 1 + 1.5 +.15 +.4 = 3.05 in column A if not using assimilation,blind, moment, deter ...

3.05 x ( mine + clone ) = 3.05x + 3.05 (aimed less specifically [this makes a lot less difference when you use firebell, which is one of the reasons I went with fire instead of lightning]) so ~6.1x then you get zei's too

or With Innas helm WoL rolled on helm its 1 + .15 + .15 + .4 = 1.7*(mine) + 1.7*(3)*(mine as clone) = 6.8x, but the majority of the damage is now not only not aimed it requires you to have cycloned the mob recently, which isnt bad for area damage necessarily, but it hinders Zei's and to get the additive multiplier up you prob would want to use assimilation since there are skills that are free so it does make this harder to play, also you have to remember that when you cyclone in epiphany the clone will cyclone outside dragging stuff away,, the mob must be hit by YOUR cyclone strike in order to receive BOTLG damage buff, not the clones cyclone.. of course it can be hit by both I'm just saying it gets messier in thicker mob types. No Zeis either and what to make up for zeis,, I still am not 100% sure lightning never bugs out at higher APS (if u inflated with pain enhancer) must but it just felt like an overall less fluid playstyle.. but ya I might favor lightning for this over fire and this would also more likely be my multiplayer set up if doing a highest possible in Group with WOL because pillar cant be reflected, as you go higher more time to set up , more assimilation stacks = favors this build more.

{Edit: note about ^ for a group scenario, first off WOL wont be the highest damage dealer possible, so I was more speaking of just for science or fun, but you would note that most buffs from other players are additive as well}

Just one of those things that I initially tried the latter and fond that what other people were using (tzo krins) was better.
11/30/2018 11:55 AMPosted by Darth
No Zeis either and what to make up for zeis


Like I said, does enforcer work? I saw it in a random profile but never got to test it. Perhaps PE.
Hi,

Rabid definititely procs AD with the enhanced Bindings+CS Damage

checked that, just to be sure

edit: Enforcer affects the Clown Damage as predicted
12/01/2018 09:32 AMPosted by StoleOwnCar
11/30/2018 11:55 AMPosted by Darth
No Zeis either and what to make up for zeis


Like I said, does enforcer work? I saw it in a random profile but never got to test it. Perhaps PE.


as Lutz had mentioned yeesssir.. PE wont artificially inflate his damage tho as if its an ally. IAS for allies does.. but not for clone. would be OP AF

if anyone can test IAS w/ Kyo Blade on inna's wol that'd be chill... trying diff servers east/west
12/02/2018 01:23 PMPosted by Lutzer

edit: Enforcer affects the Clown Damage as predicted


Well I guess Enforcer opens up some options for melee WoL variants without Tzo's. Zeis might have a higher top end damage, but it's also trickier to keep at its highest multiplier.
11/30/2018 10:09 AMPosted by StoleOwnCar
11/30/2018 12:34 AMPosted by Atomfurz
tzo gives factor 2.5 if optinally rolled. the bracers give your clone.faxtor 3(optimal), which means you get factor 2 compared to rs without bracers


Nah, it's not 2.5. It's just a standard WoL increase damage skill modifier. Skill modifiers go into the same bin as other additive increases, so it's actually far from 2.5. It's not even 2.

Well, I think I should have been more precise:
there is the WOL bonus dmg, with optimal rolls you have 15% boots and the 150% from Tzo, so all together 165% since it is additive, which is a factor of 2.65. Taking Tzo out of equip brings a bonus of 15%, so the factor is 1.15, which means the bonus dmg of Tzo compared to a setup without it is 2.65/1.15 ~ 2.3, so the increased dmg would be round about 130%. Factor is >2 if the calculation is right, maybe you know more than me, but d3planner gives similar result. It gave
With Tzo bonus dmg 134,106,878,711
without 58,197,324,723

Join the Conversation

Return to Forum