tc-subtlety

[coarse laurel]

Since whitelists bypass the generic modifier disabling rules

I’ll have to double check the whitelisted spells

Not sure about SecTec though

Probably some scripting errors on their part trying to get the modifiers to work have changed how they work now that they aren’t school modifiers. Similar concept just involving scripting.

[rugged solar]

somewhat weird that we even with this bugs don't end up higher on wcl

because they should have a noticeable impact on output

[past fable]

Tbf most people swapped away from these talents (besides DD)

[rugged solar]

fair

would potentially also make DB a more competitive choice

[past fable]

Sepsis doesnt seem to be increased by DB nor DD, perhaps its not actually being converted to shadow like im expecting

Not sure about simc behaviour, just testing ingame

[mental dirge]

any effect on shadow evis considering it works in a somewhat similar way to ShB

[past fable]

No

As far as i could tell there is no other spell that is benefitting in a way like we wouldnt expect, besides what i just said about sepsis

But also, sepsis+db seems like a very cursed build

[rugged solar]

it seems like

they fucked up a lot with the deeper dagger/etc whitelisting

and we missed that completely

nice finding it 👍

[past fable]

Ye sepsis seems to be benefitting from DD in sims and not ingame too, seems like they somehow managed to remove sepsis from the whitelist?

[rugged solar]

it still works with shb?

[past fable]

Oh nvm it doesnt seem to be working with DD in sims either, its only about ~10% more DPE with dark brew

It turns it into shadow but somehow doesnt get the DD amp, probably forgot to whitelist it, but it seems to be working correctly in sims, disregard this

[past fable]

Ye thats why i started investigating to begin with, as many people have started saying DB is better than flag. And when i tried it it felt like they were right so i had to check if flag was somehow oversimming or db was somehow undersimming

[rugged solar]

i did only see it in m+

[past fable]

Ye i was talking about m+, but my singletarget in m+ felt surprisingly high

[rugged solar]

db iirc can sim higher for single target in m+

[past fable]

Well do some double checking with sims once theyre in

[rugged solar]

so it isn't that unusual

yes ofc

[past fable]

Your sheet sims timed out just at the right time

[rugged solar]

well i did re-sim assassination just 2 weeks ago

but i think its possible that all 3 specs need re-simming

also probably already for 10.1.5?

i am not sure how outlaw is in terms of implementation status for that tho

given that people reported that their tier set is fixed, but also i haven't tested and did hear multiple opinions

[remote sparrow]

i tried both builds a bit on dummies, and it seems like the pure st build with flag is still quite a bit better than the old dm db build. but db keys build does do quite a bit more damage than the sims. nice find Eleem

[mental dirge]

old dm build doesnt take pe or inev

you'd have to drop improved dance points

[remote sparrow]

yeah, i tried taking stiletto over improve sht since blades is a bit stronger than we thought, but its not that great

idk if dropping dance to 6 seconds is it but ill try it out

[mental dirge]

you have to pretty much lol

and then enter every dance with a couple cp so you can evis on first global to still get 3 finishers in each dance

its the only viable way to still have PE + inev in the build

[past fable]

Should be the opposite, imp sht is much more potent as you need way more energy since gloomblades are more important than before

[remote sparrow]

idk tfd is way stronger than swift i dont think thats the play

[remote sparrow]

yeah stilleto wasnt it

aright ill try this out looks kinda awful but yeah...

wait

gb over inev ig

forgot lol

[mental dirge]

not taking inev is a mistake

[remote sparrow]

what do u drop for it tho

[mental dirge]

you play staccato and no vt or sht

or you could equip 2p+2p lol

[remote sparrow]

hmm ill try them both

[past fable]

Napkin math puts regular build about the same as my rotten with this change, perhaps its time to hang up my rotten hat

Well have to see whenever the changes gets put into simc though

[rugged solar]

also need to look at potential apl chanegs

*changes

maybe it is worth to maintain GB longer due to the lingering interaction

[coarse laurel]

So to explain a bit more of why this might happen.

Shadow Blades damage is a residual effect, but has the Ignore Player Damage Modifiers flag to avoid double dipping in theory.

This includes things like generic damage increases (such as the older Shadow amps) and Vers

However whitelisted damage modifiers do not respect this flag. They always work.

Unfortunately they seem to have accidentally put the Shadow Blades spell in the whitelist for DD/Veiltouched/DB

Which causes it to double dip as per the above mechanics when it previously didn’t. This is almost certainly unintended.

Shadow Blades damage spell (278043) should likely not have any family flags since there’s no reason for it to be modified by anything.

[ivory kestrel]

and quite interesting

[coarse laurel]

I had manually coded Shadow Blades to have no modifiers at all given the previous mechanics and to avoid any accidental stuff being applied.

But I’ll have to wire it up to just support this stuff normally and double check no core stuff is working with it.

However we cleaned up a lot of the core for these flags in DF

So it should be OK

[ivory kestrel]

as far as we know even things like agi buffs may double dip

[coarse laurel]

Since this is not the only place they made this mistake

[ivory kestrel]

ah true

[coarse laurel]

We cleaned up some core logic for the ignore damage modifiers and damage taken modifiers flags

So I think I can just remove my hard-coded stuff and it’ll work. I’ll test tonight.

[ivory kestrel]

well guess thil will make stiletto a bit better

and others

[coarse laurel]

SecTec is probably a similar logical issue just with manual scripting that was trying to translate the generic modifiers across when pets weren’t affected by them.

Now that it’s whitelisted they probably are. Which makes them potentially double dip in a similar fashion. Just a slightly different mechanic and more opaque. Since that’s not spell data.

Hard to say what mechanic is in effect there. It could just be applying the modifier manually.

I’ll probably have to hack it something for that one.

[balmy linden]

What talent build are you guys using for the DB?

[rugged solar]

what about lingering shadows?

probably also falls into the same category

[coarse laurel]

It seems likely but I didn't put in any special logic for it so it probably already works

[rugged solar]

the sim/logs eleem posted initially showed it to not work in simc

[remote sparrow]

wasnt lingering doing about 134% of gb damage? that seems fine / isnt double dipping no?

[rugged solar]

no lingering and blades don't double dip, but seemingly don't get the benefits from shadow damage amps in the sim

[remote sparrow]

oh i see, yes youre right!

[coarse laurel]

I need to look at a bit more data about the SecTec clone stuff to know what is going on specifically there

[rugged solar]

if you need something let us know

[coarse laurel]

I think this is a flaw with the whitelisting and Dark Brew in general

I don't think they added "conditional" entries to the other magic damage whitelists to account for the dynamic school changing

Dark Brew itself just mods everything that it changes so that one works

Veiltouched works also since it doesn't care about school

But the Deeper Daggers whitelist probably doesn't affect any of the Dark Brew stuff.

Sims should do this automatically

Probably was a bit of a non-obvious damage loss from the interaction change

[remote sparrow]

so, the current build sims about 125.5 for my character, the old db dm sims about 124.5. the current build should probably be about 2.6% higher becuase of dd (about 4.6% of damage is lingering, 9% of sectech is the shadow part, and 2.4% is blades) putting this build at ~128.8
whereas the old build going from 124.5, goes up by about 34% instead of 16. the breakdown is 7% lingering, 9% sectech shadow part, 3.6% bldaes, flat buffing these by 34% is about a 6.66% increase, putting the build at 132.8 which is significantly higher

let me know if this napkin math makes sense

[modest quartz]

Are some of these changes already pushed to newest Raidbots build?

[rugged solar]

the whitelisting bugs are pushed

https://github.com/simulationcraft/simc/commit/9fe55518018fcbbf867b75dc7b7fd84905a186fd

the secret technique bug is not

[modest quartz]

getting a 1.3% increase from nightly to latest so far.

[rugged solar]

its only in if you use "latest"

[modest quartz]

yeah, that's why I was wondering if some were pushed because I had run sims just about an hour ago, maybe two, and it had gone up since then.

[remote sparrow]

ah maybe thats why the 2 builds were so close. theyll probably end up almost identical if we include secTech bug

[coarse laurel]

Regarding SecTec, I dunno how comprehensive the testing above was in terms of reversing out which modifiers are likely to be double applied. If it's just the school ones that we complained about and Realz scripted in, or if it's all the whitelisted stuff (like the spec aura for example)

[rugged solar]

did a sim with the changes, rotten is still noticeable ahead, even tho the secret bug will add some benefits.
https://www.raidbots.com/simbot/report/12jELiL1rCZdL3GBoQEiZc

[turbid pulsar]

the numbers in that sim are lower than the ones from your sheet

atleast for the normal build

[rugged solar]

ill add it manually

moment

[remote sparrow]

^ rotten doesnt seem to have gotten any significant changes from the new update. whreas the other builds should get 2 - 5% roughly

[rugged solar]

rotten gets like 300-400 dps

[remote sparrow]

the ones I just tested were vt, dd, db. they all double dipped for sec tech.

[rugged solar]

adding more multipliers than vt/dd/db to it would go over the estimated value, so i would assume its only these 3

but if you have a idea how to reliably test, just let us know

[rugged solar]

thats a consequence of the multi-stage sim i need to add both talents as load-outs to work around this.

https://www.raidbots.com/simbot/report/qorUJwdEAwzcagfaoUGrGv

is the diffrence

so dm + flag is still ahead, but it is likely that th secret bug closes the gap

[remote sparrow]

dm db gets roughly 1.44% more from sectech bug than dm flag so its probably gonna be ahead

[rugged solar]

https://www.raidbots.com/simbot/report/mTyxTADxvJgaLwVxLbnfKR/simc

secret attack is ~9% of the damage

[remote sparrow]

but also, am i missing something, or is the new dm + flag basically basically the exact same as before? you had it in your sheet at 130.747

now its 130.8

doesnt seem right

and yeah 9% -> dm flag gets another 16% on this. so about 1.44%, and dm db gets 34% on it, so 3.06%. so dm db is gonna be ahead now

[rugged solar]

assuming the amps work, we would see a more noticable shadow close attack

https://www.raidbots.com/simbot/report/cQVRaf8WBtXa5BuqtRSvdT

nightly

https://www.raidbots.com/simbot/report/j6bzLXCnnCTxt1cLf3kUZX

latest

roughly 1% gain

[remote sparrow]

hmm interesting, fair enough!

I think dm db is gonna be fairly close to rotten after secTech fix

[rugged solar]

dark brew nightly
https://www.raidbots.com/simbot/report/4Y19mxEm9Pk3hjzR55UVav
vs. latest
https://www.raidbots.com/simbot/report/4tEpm88UrAxyuQ442yo4hE

dark brew should be ahead with secret benefits

because it is already a high amount of % of your total

amping it by another 30% is 1-2% dps

assuming its exactly 30%, it would mean you get ~2,5% more damage

would put it on 133k

[remote sparrow]

yeah, dm db seems to be the correct one.

btw this seems slightly higher than your dm/db

https://www.raidbots.com/simbot/report/f25a3P6PdqR8EnGhi1Pffo

[rugged solar]

rotten could be still ahead, but rly depends

[remote sparrow]

could be margin though not sure

[rugged solar]

interesstingly does sim worse, need to check if the updated apl has any conflicts

[remote sparrow]

which one seems worse?

[rugged solar]

the talent combo you linked

[remote sparrow]

it does? interesting

could be the old rupture optimization?

can you check if it automatically includes dd without using an evis? doubt it changes anything though, if thats the case

not sure if some of the sectech changes have been pushed to latest but botg dmdb (fuu's version) and (dm, flag) sim around 132 now

https://www.raidbots.com/simbot/report/kVqZeszzQGET81ApvLm55x

[rugged solar]

only lingering change

oh i see

you use corrupting rage

thats why you get diffrent results

[remote sparrow]

ah makes sense

[rugged solar]

i assumed it was a conflict of the rotten apl, but comparing it to baseline:
https://www.raidbots.com/simbot/report/apytAeZVwiEqgrtWRQWg5F

it has no impact

[remote sparrow]

mhm makes sense, yeah im using corrupting at 85% uptime.

[rugged solar]

what sim are you referencing?

should be 50% yes

unless gloomblade is reduced by armor

but don't think it is going from spell data

so not sure what happens, i don't see a proc that would increase damage

[golden bay]

Is the ST double dipping currently in latest?

[rugged solar]

@coarse laurel might know, he is more familiar with the implementation

[rugged solar]

no

[golden bay]

For my understanding, this means these talents have double effectiveness right?

[golden bay]

These basicly

[rugged solar]

no

you basically get an additonal ~30% damage from secret

[golden bay]

For ST only is what i mean

[rugged solar]

but only from the shadow part

so even less that that overall

also depends on your talent choices

[golden bay]

But this would bring us closer to the old season build, which might make 2p2p very close to just running 4p

Maybe worth investigating

[rugged solar]

the reaosn 4p is higher is gear levels, so you gain more agi and other stats

they also make secret/etc hit harder

secet is not part of the old 2p, so not sure why it woul suddenly become stronger

[golden bay]

True, this would be more for the 4p than 2p2p, you're right.

[rugged solar]

Eviscerate and Black Powde increase the damage and critical strike chance of your next Backstab, Shadowstrike, or Shuriken Storm by 2% per Combo Point spent.

is the old 2p, i also think this is better discussed in #subtlety to keep the chat in here on topic

[past fable]

Why would rotten be getting any DPS from this? Its not using any of the talents with the interaction?

Also am i reading it correctly that the sectech stuff is not in simc yet and only the blades+lingering whitelisting issue is?

[rugged solar]

yes

if you use latest, you get the whitelist bug

[rugged solar]

you are correct, it shouldn't impact it

given that you don't use dd/vt/db

[coarse laurel]

You have Deeper Daggers?

But you should be able to tell what modifiers are applied in the debug output. I can look a bit later. Setting the DD mod to 0 will just turn off all snapshots anyway so that’s just gonna kill everything that is supposed to work.

It's because the spell data has an AP coefficient of 0.1

[remote sparrow]

@rugged solar when you have time can you double check this one more time? using your setup the old dmdb build sims within the margin with the current flag build. (should easily pull ahead given sectech bug)
https://www.raidbots.com/simbot/report/j7TBW8sjK3ZhTTvMFDmCFn

[rugged solar]

@remote sparrow this is because the nightly version has the secret bug implemented

nightly/latest

[remote sparrow]

ah okay makes sense. the dbdm was simming higher with my own gear actually

[remote sparrow]

does the current apl for db play around finality rupture? I think stealthi found out it was better to never consume finality rupture right?

[rugged solar]

that was a consequence of a bug in simc

and quickly got fixed and we moved back to the default from 10.0.x

[remote sparrow]

i thought it was because of the 2 pc extending rupture that the buff would time out

i see, okay so 10.0 default. thans!

[rugged solar]

simc had a core bug

that did remove the finality amp when rupture was extended

this was why playing around finality was a loss with 2p

[remote sparrow]

oh youre right, i remember that! good memory

[past fable]

Very minor thing that txt pointed out that ive now tested fairly extensively and come to the conclusion that he is right about is instant poison ALSO double dips on dark brew, just like sectech clones, but not on the other amps we have.
DD doesnt seem to affect instant at all (just like it didnt affect sepsis) though
https://www.warcraftlogs.com/reports/Ncr4BXWx96HtdGAg/#type=damage-done&source=1&boss=0&difficulty=0 Logs
https://i.gyazo.com/8c21db54e3d3d311b9bd6274965ced32.png Pic of my haphazardly scuffed spreadsheet with the data

And as you can see by my scuffed pics sims (reasonably) assume it will only affect them once

[coarse laurel]

Not sure about the double dip here, it might be explained via double whitelisting or something. I can check

But based on all the things I've seen in the spell data and kinda what I noted the other day, I would expect:

1. Deeper Daggers probably doesn't work on any school-converted Dark Brew effects--Rupture, Sepsis, poisons, etc.

2. Veiltouched probably doesn't work on any non-magic school-converted Dark Brew effects (Rupture)

[past fable]

Ill check rupture now, how does simc handle it? It assumes it works?

[coarse laurel]

SimC should do as noted above

[past fable]

Aight, ill double check ingame behaviour to see if it matches

[rugged solar]

rupture does not work with dd, i checked that earlier

[coarse laurel]

Replicating Shadow shadow versions of Rupture should work fine, but the base Rupture damage is not affected by any Veiltouched or Deeper Daggers whitelists

[past fable]

Yep it does indeed not get any benefit from veiltouched nor dd

[rugged solar]

its not very intuitive

[coarse laurel]

Well it makes sense, it would require them to have conditional auras for Dark Brew in those other talents as additional effects

[rugged solar]

yes, it seems logical technically

[coarse laurel]

I can't say I see any specific reason Instant Poison would be affected twice by Dark Brew

[rugged solar]

my point is more that the baseline assumption would be that shadow damage gets amped and only testing or reading spell data shows that this isn't the case

[coarse laurel]

Sure, I mean to players this is super unintuitive and just an odd side-effect of the whitelisting effort

If I were Blizzard I'd probably look at adding support for school masks on the whitelist modifiers so that they could add all class spells to whitelist but also add a school mask

But they may not have the support for doing that at this moment

[rugged solar]

sounds like additional engineering effort

[rugged solar]

btw. i love spreadsheets, good job

[past fable]

Neither do i, which is why i was skeptic when i heard it, but figured i was gonna check it out anyway and it seems to be the case, for whatever reason

Blizzard devs cooking up that good carbonara 🍝

[rugged solar]

could be a similar problem to the secret one

[coarse laurel]

I mean that one kinda makes sense in terms of the whole pet issue and them having to script things up

But there would have never really been a reason to massage anything with IP that I can think of

[rugged solar]

there was the poison bug with NS, maybe they moved to a script to solve shenanigans like this

[ivory kestrel]

they also have different poisons for each spec kinda

after they brung them back

[remote sparrow]

for the 6 second dance DM + DB build, I was thinking if it was worth it to enter dance with 2 CP when we dont have vanish or blades up to guarantee 3 finishers in dance. So dance end up dance evis strike st strike strike evis
and when we have vanish or blades up we can get 3 finishers much more reliably so we enter with 0 cp.

[rugged solar]

builds typically use tfd atm, so entering with low cp should be the default

[past fable]

What exact do you mean by shd > 1 cp

Just evis regardless as the last gcd if you are at 4 cp?

Nice finds regardless

It also gets a gloomblade with full lingering stacks 100%

[rugged solar]

the change is probably more specific to dd

https://www.raidbots.com/simbot/report/tyyoGij2im39pqd9vjDExh

the reference line is your line

https://www.raidbots.com/simbot/report/wAKywLZbRUpePUEx3ciMdC

i see 0.2% by just adding it to the dance finish conditions

this is diffrent to the condition

because it does not enforce EVI cast but calls the finisher line

simply finishing last gcd seems like a loss quick

dd is 8 sec, so should not matter

so you are correct

could be due to gb with lingering stacks

but its a loss for rotten

rotten does not play 2 lingering points

[balmy linden]

@past fable This seems to be fixed now

Which sectech double dipping

[remote sparrow]

what about lingering and blades?

awesome thanks

[balmy linden]

@polar palm the log of me hitting the dummy isnt work

https://prnt.sc/MC05fk5lOCO4

https://prnt.sc/fEKooQGwqUBL

[remote sparrow]

are we sure the sectech clones are not double dipping anymore?

[rugged solar]

it is very easy testable

just use secret with and without dark brew talented

but it worked fine when i tested around a week ago

nice

[rugged solar]

#Implementation status:

• Kojiyama implemented the first round of 10.2 talents
• Tier set is still not 100% corrctly implemented.
• Apl is actively changed, is stable but has still some room for experimentation.

How to sim for 10.2

• Raidbots advanced simulation
• ptr=1 on top of the simulation
• Latest profile / apl we provide in this channel

Timeline / What to expect

• Cleanup of the current Apl for 10.2
• Iterative process in an attempt to optimize for the new talents and layout change.

How to contribute

• Highlight problems in sample sequences
• Share ideas or concepts
• Contribute Simulations with comparisons

IMPORTANT

-- PTR is experimental, please keep the discussion about theorycrafting. Tuning of specs relative to each other is irrelevant --

[rugged solar]

unpinned everything from 10.1.x

[low breach]

Nice that we got this but compared to the Outlaw one this is sadly not working for me 😭

[rugged solar]

works just fine

https://www.raidbots.com/simbot/report/3VNAJJumr69wmHgF3j5U4U

is a unoptimized version

i am in the process of doing cleanup/etc

this is my current stand in terms of dps

but expect it to go up more

[low breach]

Ok yeah I got the same error

[rugged solar]

well the sim still works, so no concern ^^

i ask dm'd koji and asked if something is missing setup

[low breach]

Ok yeah sorry 😞

[rugged solar]

all fine, don't orry

i just wanted to make it clear that i actively work on cleaning up the apl

[low breach]

For outlaw ngl it was quite easy and with current gear and talents it was a 9-10% dps increase.

[low breach]

Yes that is great eager to dive in

[rugged solar]

but putting out a incomplete change is not that great

but will put the apl out once it is somewhat okay

[coarse laurel]

Error is ignorable

[rugged solar]

currently highest simming talent combo for me

BUQAAAAAAAAAAAAAAAAAAAAAAAAEEAAAAAASLJJlECpAJiEJJJJRi0iEJaJJIBikSKBAAAgEkA
or
BUQAAAAAAAAAAAAAAAAAAAAAAAAEEAAAAAASLJJlEQKQiIRSSSSkItIJJaJJIBikSKBAAAgEkA

[remote sparrow]

could you also post the links with these so we could look into the report?

[rugged solar]

Dust seems dominant, highest simming dust currently

BUQAAAAAAAAAAAAAAAAAAAAAAAAEEAAAAAASLJJlEQKkkISERSSkItIJJaJJIBikSKBAAAgEEA

[valid robin]

Can you link a picture or something

No clue what these builds are with just a string

[rugged solar]

ye will do in a sec

Dust build

BUQAAAAAAAAAAAAAAAAAAAAAAAAEEAAAAAASLJJlEQKkkISERSSkItIJJaJJIBikSKBAAAgEEA```NOTE: one point finality -> Goremaws might be worth it

Non Dust
BUQAAAAAAAAAAAAAAAAAAAAAAAAEEAAAAAASLJJlEQKQiIRSSSSkItIJJaJJIBikSKBAAAgEkA

rotten is worth it, similar to above GB likely gains value if you optimize it more

[remote sparrow]

im very very very surprised that rotten PV doesnt take sepsis or goremaw

kinda breaks my mind

[rugged solar]

rotten is only worth it due to tier

just saying

[remote sparrow]

oh this is with current tier, maybe we can try with normal 447 non tier items to get a better feel of the actual talents.

the futue tier doesnt really lean into any builds, so this might be a more accurate representation

[rugged solar]

i did dissable tier too and compared

the one thing tier does is make rotten worth it, not much else

[remote sparrow]

interesting, I thought tier would further emphasize shadowcraft and inev too

[rugged solar]

inev is good with swift

but taking swift is slightly worse than this pathing

[remote sparrow]

could you please try this : BUQAdi03WEe/OgykedwA/EwYEBAEEAAAAAASJJJlERSKFHQICRSSSISJSSiWSCSgIpkSAAAAIBB

i think this would require quite a bit of optimization, but im curious how it does with the baseline api

also another thing is lethality with the sepsis dust build, curious if its worth it over tea

[past fable]

Its kinda relevant that the refund mechanic on shadowcraft isnt in simc yet

I assume that would certainly up the value of the talent (and talents that synergize with it)

[remote sparrow]

oh ofcourse, I just meant more sod -> the better sod focused talent

surprised it wasnt as big

[rugged solar]

inev could get higher value once that works

[remote sparrow]

for sure, @past fable btw, I tested your suggestion yesterday a bit more

[rugged solar]

https://www.raidbots.com/simbot/report/fD4RCdLLptM7nXW1q6tM8a

[remote sparrow]

this is so far whats working best for me.

[remote sparrow]

thanks, this is closer than I expected tbh.

btw are you syncing flag and blades together?

[rugged solar]

Dust Rotten

BUQAAAAAAAAAAAAAAAAAAAAAAAAEEAAAAAASLJJlEJJpUEBSEkkESSLSSiWSCSgIpkSAAAAIBB```

turns out swift ins out even

with current tier

eleem found out

but i would to a degree return to actual working on the apl more

because while its fun to poke around with talent combos

it is more relevant once things are better optimized

[remote sparrow]

makes sense, probably more important to make sure we dont dance on cd, have 2 for cds.

I expect these to be big changes

[rugged solar]

the apl does not dance on cd

[past fable]

Yes it does

[rugged solar]

so thats not a big issue, but there might be some optimizations in changing cooldown handling

flag is synced

goremaws has a initial decent condition

still a lot needs to be tested

[remote sparrow]

ah I see, interesting, ill go over the html report and see whats going on

[rugged solar]

probably a lot of weird stuff ^^

because i spend some hours now running talent sims instead of looking into apl

[past fable]

I would expect non-dust to be closer to dust if you swap to using the embersoul trinket

[remote sparrow]

for sure

[past fable]

Now its only using bomb and beacon, which dont really care about damage spread

[remote sparrow]

this is kinda alarming to me.

the peaks are way too low

this is also very weird

[mystic sequoia]

Because those are averages over all the sims

You can theoretically have a totally flat damage per second graph, even if each induvidual pull has insane peaks and valleys

A flatter dps graphs more often than not indicates that cd's are used at different timings between the iterations

[past fable]

His point is that the peaks arent just low, theyre very low. At 2 minute your burst not even reaching 200k

Means that the APL still requires a lot of work

[mystic sequoia]

If you run just one iteration, it'll look different

[past fable]

As ingame you can consistently get it to like 350k

sometimes to like 400k+

Albeit with an on-use trinket, but still thats only like 20-30%, not doubling

[mystic sequoia]

Like say one iteration you have 400k burst at 50-60 sec. Then 100k 60-70 sec.
And the next you have 100k at 50-60 before you burst at 60-70 for 400k.
Then the graph would list the burst window between 50-70 as 250k

[past fable]

But its gonna use shadowblades at the same time

Every pull

So thats not an issue

[remote sparrow]

^ its also the cps coming in, the begining is very spiky with the pump and dump of blades cps. which makes sense, it happens again at roughly 100 seconds, which is the 2nd blades

but then you never see the cp spikes, on 3rd and 4th blades

[mystic sequoia]

If you SB even 1 second difference

The timing of you having max cp, or 0 cp will cancel eachother out and equalize

Between two runs

And shadow blades will only be used at 0 cp according to apl. And with flag up, and snd refresh condition. Which means that any variation in cp influx will offset it. Which is why your cp graph evens out

And the dips towards the end is because some simulations finish earlier due to fight length variance

[remote sparrow]

hmm maybe then it would make sense to add a smoothing factor to it rather than litearlly adding things up per second.

[mystic sequoia]

It's because it's an aggregate of several iterations

[remote sparrow]

oh I know, i meant you can add a 10 second smoothing factor then average that, so the spikes can be more pronounced

[mystic sequoia]

That's just adding another average to it :p

[remote sparrow]

it is, but it should help with the cp income looking flat during blades

just a quick look at the dust apl -> misses 1 sod use (doesnt sod before first vanish due to tier and then misses it

doesnt tornado before vanish, ends up desyncing it with the otehr blades

[past fable]

Yes it needs a lot of work

[remote sparrow]

double vanishes in open, sends again at 1 min then ends up desyncing things even more

[mystic sequoia]

To check that, it's a lot easier to just check what uptime it has in the sim, and divide that on the theoretical maximum uptime

[remote sparrow]

dance and sod are not synced, really,

doesnt have 2 dances for blades

thats first 2 minutes, im expecting rest to be the same

[past fable]

Yes, having dances pretty strictly tied to symbols and/or blades is probably ideal, its currently doing stuff like sending a dance at 2:00 and then using blades at 2:06 and then not dancing throughout the whole blades. Which seems kinda far from ideal

[rugged solar]

i haven't rly tried to sync things more

[past fable]

Its not a jab towards you, i know you havent had time to fix all of it, or even close

[rugged solar]

i know just wanted to emphasize that it is a area i haven't looked at in detail

[past fable]

Ye i just suspect its an area where we'd find a lot of upgrades, seeing as the non-opener blade peaks shouldnt be that far below the opener ones theres probably lots of gains to be made

[rugged solar]

its amazing, we now have a lot of room to experiment

[azure drift]

[mental dirge]

Yea the one at 2 mins should not be below 200k lol

Or even close to that

Somethings going cataclysmically wrong there

All of the cd peaks are relatively bad

[remote sparrow]

it doesnt have dances for them up

[rugged solar]

yes

also, even if we end up not aligning everything, i will make sure to find a way to allow it

so worse case we have something similar to prio_rotation

[mental dirge]

Explains why sepsis simming well

[rugged solar]

sepsis is probably in general not bad

[low breach]

Thats st? Althougth with BP baseline we have already enough aoe

[rugged solar]

yes, all of the linked is single target

[coarse laurel]

I mean this is just meant generically for resources over time. There really isn't any way to time-align something like CP over multiple generations

It's just gonna skew towards the average no matter what you do

It's only relevant for things that would temporarily have a major skew factor (let's say a buff that gave infinite combo points for 15 seconds or something)

But since there's really nothing like that at play here, there's not really anything to add any skew to the graph

Similar to this, mid-iteration damage peaks are going to be suppressed somewhat unless there is very strict timing alignment. Anything where it's +/- some timing window is going to pull down the peaks on the graph. Doesn't mean that the iterations themselves aren't hitting higher peaks, it just means at that given timestamp that's the average across iterations.

[remote sparrow]

this is what kind of tipped me off about it initially, since blades does kind of give you very very large amount of combo points for 20 seconds straight, and its most of the time should be cast around the same time give or take like 5 seconds. so I was expecting much higher input there, which is not what it was showing. but I think it just seemed to be an issue with APL.

[coarse laurel]

If you run a single iteration sim, you'll see all the interim peaks actually do break 200k, but their timing is not strictly aligned since they are minor CDs

So across the entire sim, this will average down any specific peak at a specific timestamp

[remote sparrow]

yeah that does make sense, if the peaks are bit seperated.

[coarse laurel]

They'll really only show the "true" peaks in cases with really strict CDs (something like DK CDs, or Demo Warlock CDs are a good example)

That always end up landing on roughly the same timestamps

Just kinda a weakness of SimC datacollection

[past fable]

Why would shadowblades not be such a strict cd? (disregarding dust builds)

[remote sparrow]

btw koji idk where else to say it I guess, but I'd be happy to help a bit with the codebase if you need any help. Im mostly working in python nowadays (since all of ML community is basically in python) but I used to mostly work with C++ (not a lot of C though) for quite a few years before, so should be able to help a bit when works gets too much

[remote sparrow]

I think it will be, once we fix the APL. right now, blades is not really benig treated as an actual cd because not having dances up for it basically devalues it heavily as a CD.

[coarse laurel]

I assume Shadowblades is being aligned with Dance/SoD/etc. conditions which causes it to drift somewhat instead of being used exactly every

But for the CP chart it's actually also just fighting sampling windows

If you consider a builder/spender model, if you have infinite CP generation you actually start to skew towards the mean anyway

Because you have 1 generator second with max CP

Then 1 spender second with 0 (+ refund) CP

As far as SimC sampling is concerned this is just gonna descend into a mean pretty quickly

CP isn't a great thing to track this way unfortunately

[past fable]

Ye and this is something were likely to remove/scale back as shadowblades is THE cooldown now, it seems similar in strength to some of the other classes CDs youve mentioned.

[coarse laurel]

Just becomes a random distribution at given timestamps that average down to the mean

Like this is a good example

Once these high gen spikes get overlayed on N iterations

It's just gonna flatten out because enough will be offset +/- 1 GCD

To pull it to the mean

Might be a smarter way to do this for charting, but would require core SimC changes

[remote sparrow]

that makes sense, I thougt about maybe adding a weighted smoothing over it per run, but if the windows shift, over N itterations, itll flatten out again as long as the cds are not exactly aligned.

[coarse laurel]

Thanks for the offer! For now, I think the bulk of it should be handled but certainly always helpful from a testing and suggestion POV, or if you notice anything when reviewing the code or testing sims. More eyes on this would be great. I'm not really playing much now so the coding time is less of a problem than the testing and knowledge of all the wonky behaviors/bugs that come with Rogue. 🤣

[coarse laurel]

I'm actually curious how much this actually matters tbh. Feels like with the way it's been described to work, that this actually isn't that big of a deal? I'm not sure I've seen any sims where the Shadow Techniques stacks ever overflow so it seems they are all being consumed either way? Maybe some subtle timing shifts into CD windows I guess? But I am not sure if it will make a big difference or not.

(Not saying it won't, just an observation based on looking at the stack counts and stuff last night while testing.)

[past fable]

During shadowblades youre not using any of your sht stacks to get more cp (naturally as you get full cp on every builder) so it builds up quite quickly. And everytime it refunds you get to send a finisher instead of a builder which is a fairly big deal during all of your cooldowns (blades+flag+trinkets etc).

Its also very possible we end up pooling in some way ahead of our regular dance windows to make use of the refund, but yes the default apl doesnt get a ton of stacks

[coarse laurel]

Seems like the main gain from Shadowcraft is just the large jump in generation to begin with (lowering the swing target and giving 2x stack instead of 1x) but seems like the mechanics on builders still already can consume those stacks pretty quickly unless Shadow Blades is up. So it'll definitely help with CD windows in that regard. (Although curious if you still end up wanting to use some builders anyway due to other mechanics. Not sure.)

[past fable]

next week theres supposedly a fix that makes it so the refund will interact with the tier bonus aswell. So whenever the 2p procs of a eviscerate shadowcraft can spend 5 cp to fill up your combo points instead of 7 (since you already got 2 from the set)

[coarse laurel]

Right

[past fable]

Which is random for evis/rupture but sectech always gives you 4cp, so you just need 3 stacks for that

But even now you can get 14 or 15 finishers in your shadow blades window (with 20s symbols from dust)

Rather than the 10/10 you'd get without shadowcraft

Although im not sure that will outvalue the value you get from the extra procs, but its atleast a noticeable difference

[coarse laurel]

As someone who hasn't really tested on PTR much, Shadowcraft's tooltip is still maybe the #1 thing that made me triple-take to figure out what in the world it was actually trying to say 😄

[past fable]

You are definitely not alone

[remote sparrow]

this matters in 2 ways, 1 ) during dance you can go from 4 builders 4 finishers to 3 builder 5 finishers (since it essentially switches the place of a builder with a finisher), which is a noticable gain. 2) during the entire blades, you can go from 10 builders 10 finishers, to about 6-7 builders 13-14 finishers. currently I can do 8-12 on dummies, with on average 16 4pc cps wasted. so should be atleast ni theory get 2 more finishers, making it 6-14 or 7-13

[past fable]

Realz has even apologized for it

[coarse laurel]

haha

Ok, thanks ChatGPT

😄

[remote sparrow]

surprisingly its good!

ChatGPT is very good at making up stuff

[coarse laurel]

I did have to help it slightly, but mostly just gave it the existing tooltips lol

Then had to define what "finishing moves" do

Not too bad all things considered

[remote sparrow]

oh, yeah that makes sense, yeah it does a pretty decent job when given context to just summarize

[coarse laurel]

Can you try again with the additional information that "finishing moves" are defined as something that normally consumes combo points (leaving you with none) when used.

Is basically all I had to tell it beyond the tooltips

[remote sparrow]

I remember asking it to tell me about myself, and it just said oh Armin is this and that, and these are the papers hes published

[coarse laurel]

Not bad haha

[remote sparrow]

and i had not published any of those papers

i asked it to show them to me, and it just linked random things

[coarse laurel]

You heard it here first guys, Subtlety confirmed S-tier

But yeah, I should have this in tonight I think

And some first pass on the set bonuses

[remote sparrow]

dont show this to the 'content creators' pls. i cnat take anymore tierlists

[coarse laurel]

Any anything else you guys notice might be broken

Feel free to ping me if you discover anything

Was already fixing Whispyr's Assassination reports earlier

[remote sparrow]

sounds good, good luck!

[ivory kestrel]

for sub at least might want to wait 1 week to see some thing but i will try to do some work on ER things now that i saw tea was kinda meh

[coarse laurel]

It's funny, since on the surface the storage mechanism seems less annoying for ER than random CP procs

But given how it works, feels like the updated Shadow Techniques mechanics would really make playing ER ultimate pain

[ivory kestrel]

actually it is easier now kinda

but yea until i play it more cant say much

[rugged solar]

It is harder because you get a lot of random cp

And you can't land it during shb

[past fable]

Well theyre less random than before, but its still rough

[rugged solar]

4p procs

Sht is the same

You can similar to before delay builder slightly for sht cp

Or we'll it is different in that you now consider time to sht and could end up delaying builder

When you would prior kind of delay finisher (or rather skip a builder)

[ivory kestrel]

ER is easy tho with 4p you can pop 2 for sure 100%

one on use 3 or 5 then 1 with sec tech on 4

[past fable]

Theres no 2

Theres only 3,4 and 5

[ivory kestrel]

yea i never talked about cp 2

[narrow leaf]

what build was thid

[turbid pulsar]

which one of them?

[remote sparrow]

the apl is all over the place, it doesnt really play the build yet.

[narrow leaf]

@rugged solar 12300

closet i got was 11900

[turbid pulsar]

BUQAAAAAAAAAAAAAAAAAAAAAAAAEEAAAAAASLJJlEQKSSiIRcAJJRi0ikkolkgSgIpkSAAAAIBB

[narrow leaf]

with e one dust and goremaw

ty

oh yeah

I'm very new to advanced sims and just trying to learn. Is there a way to break down how ability's are used like on regular sim or is that not possible on advanced

[turbid pulsar]

ýeah

html report

action priority list
sample sequence table

[narrow leaf]

looks like it's using gear i dont have equipped as well like becon etc

[turbid pulsar]

look at the breakdown

[narrow leaf]

wait i think i have it

wasn't inputting my gear

Thanks for the help

[rugged solar]

The attention is my current iteration of the 10.2 apl.
Please keep in mind that a lot of things are likely not optimal

for profile, i typically just use the one form !fuu

!fuu

[gilded crest]

Rogue Spreadsheet - Light Version / Dark Version

[rugged solar]

so you can fairly simple do sth like this for testing:

[rugged solar]

^ given that i got some dms about this

this is the entire profile, you can just copy/paste in advanced

[past fable]

Fuu i might've found a reason as to why dust sims so much higher than the left side
https://mimiron.raidbots.com/simbot/report/3xibKHCgKeSTQrxedvaF9Y

[rugged solar]

what did you find

[past fable]

DD doesnt do anything

0

[rugged solar]

trinket comparison

trinkets if you want to skip beacon:

trinket2=,id=204211,ilevel=457

[rugged solar]

oh rip

[mental dirge]

[past fable]

DB seems to work though

[coarse laurel]

So... long ago in beta, DD's buff value was on the buff

Then it was changed after launch to have it on the talent, with the value in the buff as 0

For some reason in this PTR it was changed to have the value on the talent as 0 and the value on the buff again... smh

[mental dirge]

[rugged solar]

yikes

[past fable]

indeed

[rugged solar]

[mental dirge]

Theyre just trying to keep you on your toes koji

[past fable]

Is VT similar

[mental dirge]

Its like a boxing match

It probably has something to do with DD existing as a conduit before

[rugged solar]

vt wasn't changed

[mental dirge]

So i assume vt is not similar

[lilac crescent]

I suppose this may mean they are moving away from server side scripts

[rugged solar]

dd changed form 2 points to 1

[past fable]

Ye VT works, just thought to check all the amps as theyve been funky in the past

[rugged solar]

dd, shadowed finisher, lingering

[past fable]

I was looking at a breakdown of the best sim you posted fuu and was wondering how 1.5% from lingering could possibly be better than my napkin math of ~3% from DD

[rugged solar]

basically all the talents that got changed

[rugged solar]

great find

[coarse laurel]

I assume it was related to it going to 1 point talent, but I don't think it actually needed to be changed.. oh well

[past fable]

good stuff

[coarse laurel]

Checked in

[past fable]

(if anybody is wondering the reason the number is so low is because the talents are completely scuffed to only showcase the DD stuff)

[rugged solar]

only dd tho

but lets see, maybe it changes talent setups

[coarse laurel]

Might be some limitation they have on the server side for single rank talents not inparting any rank modifiers or something idk

Fun times

[rugged solar]

probably needs som minutes before raidbots rebuilds

[past fable]

It should atleast make DD better than lingering, but not convinced it will change anything else drastically

[rugged solar]

clos the gap between db and rotten a bit

moment, ill check if the sim is updated

think we have a new top build tho

need to run some more combinations

[rotund flint]

^_^ show

[rugged solar]

BUQAAAAAAAAAAAAAAAAAAAAAAAAEEAAAAAASLJJlEQKlEJiEhkkEJSLSSiWSCSgIpkSAAAAIBB

best dark brew
BUQAAAAAAAAAAAAAAAAAAAAAAAAEEAAAAAASLJJlECpAJiEJJJJRi0ikkolkgEISKpEAAAASQA

interessting that gb is skipped

in both, let me quick check without current tier

[rotund flint]

thats interesting that GB with DB is stronger than inev

or goremaw

[rugged solar]

you mean weaker

it is due to tier too

so moment

[remote sparrow]

gb is roughly 2% if you have db and dd

its mostly good because of FW positional with backstab only

[past fable]

But thats bugged and doesnt work so its fine

[rotund flint]

yeah but beat goeramw and inev pv

[past fable]

Backstab applies fw from the front too

[remote sparrow]

yeah exactly, lets hope it remains bugged

[rotund flint]

fuu does your apl use Gloomblade in dance? for dm stack ?

like when you have shadowblades up

[rugged solar]

interessting

i think it stays the same

without tier

but this is the diffrence

BUQAAAAAAAAAAAAAAAAAAAAAAAAEEAAAAAASLJJlEQKkkISkEJJRi0ikkolkgEISKpEAAAASAA

[remote sparrow]

im having a hard time seeing why is better than Sepsis PV tbh. (instead of inevn dd)

[rugged solar]

best dark brew

i ned to check if the middle tier talents also change maybe

[remote sparrow]

I think you always want PE. its like the stronges ttalent in the middle

[past fable]

Are you including swift in your combinations or no

[remote sparrow]

^^

[rugged solar]

haven't yet

[remote sparrow]

swift provides same symbols uptime as inev roughly

[rugged solar]

its why i mentioned that i need to run combinations

[ivory kestrel]

sepsis and PV/rotten syncs are not in etc

so it also messes up the vanish

but we will get to it eventually LUL

[rugged solar]

yes, ofc there is still optimization potential

[ivory kestrel]

first let's try to get whole blades working

so it does not krangle it

and does weird things

[remote sparrow]

oh also another thing with shadowcraft optimization (maybe for the future)

its usually possible to get 5 finishers 3 builders in dnace

100% possible for every blades dance

[ivory kestrel]

yes entering dance with around 4 sht

[remote sparrow]

but also possible with tier fix on non blades dances

[ivory kestrel]

also yea tier fix is a big ? for me at least

cant really math it out easily

like we can know the value of cps but the moment they come in is also important you could get lots of gcd value

[remote sparrow]

also blades can be a 12-8 right now. 12 finishers 8 builders, maybe 7-13 with tier fix

[remote sparrow]

tier premed fix is like 6-8 cps per dance

so one finisher

[ivory kestrel]

it is also the diff if you get to 5sth and proc tier to get one more finisher in dance

instead of 7

and well sec tech saves 4 too apart from premed

[remote sparrow]

yeah, hopefully once these are fixed, it'll be good

[rugged solar]

swift sims worse for me

[remote sparrow]

could you show the swif builds you tested

[rugged solar]

shadowcraft is quite good

we probably won't skip on it

[remote sparrow]

i think it was swift = inev not shadowcraft

like inev gives you roughly 3-4 seconds extra sod.

oh and this is my main issue with non swift double dust builds right now.

but after your first blades and vanish, youre gonna overcap on dance waiting for sod to come back up

like a good 5-6 seconds of dance over cap

[ivory kestrel]

i dont think i would play dust without swift tbh

it unkrangles many things

[rugged solar]

think shadowcraft still does not expell cp

so it will potentially get marginal stronger

[past fable]

Even without the cp refund part it still seems to be like a 9% talent

So its absolutely cracked

[rugged solar]

its rly good

but its not as cracked if you remove tier

[past fable]

fair enough

[remote sparrow]

what do you mean? not sure i understand

[rugged solar]

the sim implementation of shaocraft is iirc still not finalized

[remote sparrow]

ah yes ofcourse

i just meant for future optimizations

little things we could play around that could potentially be impactful

[rugged solar]

reminded to myself to try to keep the pinned version up to date

[remote sparrow]

these are a little weird.
the dust build in the pinned -> 1 point of finality, inev, and goremaw all sim within 0.5% with the current APL
but if you take that point and put it in gb its now 10% weaker. not sure if this makes sense, dont think 1 point of finality, inev, or goremaw are anywhere close to 10%

[mental dirge]

[rugged solar]

could be a apl problem

[remote sparrow]

maybe its forcing gbs in dance?

but if you dont have gb talented it skips backstab?

[rugged solar]

don't think there is a finality specific optimization tho

[remote sparrow]

its not finality specific. 1 point of finality is like 128.300, inev is like 128900, gb is like 128000, gb is 119000

[rugged solar]

inev is worth a lot

because of tier

set_bonus="tier30_4pc=0"
set_bonus="tier30_2pc=0"

if you want to sim without tier

[remote sparrow]

thanks!

so... without tier all are basically the same

[balmy linden]

Is this for live? Echo and bomb are only .3% worse than bomb and beacon I assume?

[remote sparrow]

like inev is 124.16

1 point of finality is 124.1

gorema is 123.8

[rugged solar]

yes

it is for 10.2

[remote sparrow]

goremaws with rotten is rughly a 2.4%, finality and inev 2.5%

havent checked to see if it uses goremaw with pv and rotten only though

somehow dd is weaker than i thought

ist 2% instead of 3%

maybe because no gb

VT about 1.2-1.3%

[past fable]

Well were missing the new 2p

Which does a sizeable chunk of shadow damage

[remote sparrow]

true, these are all without 2p.

although that entire 4p is about 5.5-6%

so you can just add a 5% or 8% to that

and get a pretty good estimate

[rugged solar]

tier set def. will make brew a bit more dominant

[remote sparrow]

tier ads about .5% to db, 0.4% to dd, 0.3% to vt

not that much

[past fable]

Tons of cp too

[remote sparrow]

idk how relevant those would be to these specific talents

very hard to tell

given that sepsis is not being used with pv or rotten right now, and it still sims about 2%, might put it ahead of the other options

like this is sepsis damage with db,dd

this is with goremaw pv

I think sepsis is similar to rupture, dd vt dont work on it?

when talented into db that is

[past fable]

yes DD doesnt affect sepsis after its been "converted" unless theyve changed it very recently

[coarse laurel]

I should hopefully have the set bonuses in tonight

[narrow leaf]

says in the combat log it's shadow damage

[past fable]

Yes, but it doesnt benefit from DD (or vt in the case of rupture)

[narrow leaf]

ahhh ok

was kinda the reason i was trying it out

[remote sparrow]

I think highest simming with tier disabled

BUQAAAAAAAAAAAAAAAAAAAAAAAAEEAAAAAASLJJlEQKlEJiEhkkEJSLSSiWSCSgIpkSAAAAIBB

[rotund flint]

since tier set buffing finisher i can see finality being better with it

[remote sparrow]

i cant dude

this is with vt best simming no tier build i have rn

this is with RS

instead of VT

how garbage is our tree

[rotund flint]

dead nodes:))

[remote sparrow]

cant believe
I was right about RS being better in ST than our other talents

[past fable]

Could you link/dm me any of the sims real quick

Just wanna check an interaction

[remote sparrow]

rotten looks quite weak. about 3%

pv looks like a 0% talent btw.

its like 0.2%

[past fable]

Tbf theres probably some room to optimize pv with sepsis/goremaw/er to make it better

But ye were not gonna make it an amazing talent either way

[remote sparrow]

this seems to be the best so far

with the current apl

without tier

the bad news : this builds aoe is really bad. assuming 55% of your aoe damage is in rupture and bp, you lose 15% on bp, 16.5% on rupture, 10% on ST

so your aoe is just going to be really bad if you run it.

good news (ish) : if you want to take 2x finalityt db dd for max aoe : you only lose about 4.5% single target

so not too bad

although idk how big dust is in aoe

[past fable]

Its gonna be a lot closer with the new tier, since it synergizes a lot better with the right side

[remote sparrow]

probably yea

[past fable]

tier is like ~20% more evis and rupture damage

[remote sparrow]

its better with right side yeah

but right now left side is very unoptimized

with regards to both pv and rotten

so the difference is most likely more

oh interesting, goremaws better than inev without tier.

BUQAAAAAAAAAAAAAAAAAAAAAAAAEEAAAAAASLJJlEQKQCSSSSSSkItIJJaJJIBikSKBAAAgEkA

looks like this so far:

[rugged solar]

111.2 st seems low

[past fable]

He is simming without tier

[rugged solar]

oh

[rugged solar]

BUQAAAAAAAAAAAAAAAAAAAAAAAAEEAAAAAASLJJlECpUSkIChkkEJSLSSiWSCSgIpkSAAAAIBB

https://www.raidbots.com/simbot/report/iHyjX6ui5z63Pw5f9mGcG2

this one is 112k st

[past fable]

Well ye thats dust

He has a dust at 115

[rugged solar]

oh, non dust

i assumed the above is the best overall

[rugged solar]

did you already add shadowcraft?

[remote sparrow]

I haven't tested everything yet, I'll look into more builds. Later when I get home, in a bit

[rugged solar]

no hurry ^^

[narrow leaf]

I can get non dust to 110,00 ish I just wished that DD and VT worked with sepsis. It reads like it should

it is what it is though i guess

[past fable]

If you end up implementing the tier sets tonight koji the 2p interacts with our amps in the following ways (unless it changed this most recent patch)
Rupture clone proc only benefits from 4p, so its either 50% or 55%
BP & Evis clones benefit from dB & dd but not vt. Works with 4p. So ~65% total

Pretty sure the procc chance is 33%, could be 30 or 35 too but fairly confident in 33

[remote sparrow]

what about ST physical hit?

im also assuming bp is the same as evis?

st is just 65% then?

similar to evis physical?

[past fable]

Yes its based entirely on the physical part, shadowed finishers have no impact on it

[remote sparrow]

did a quick SF test

its a 0.1% node

[valid robin]

Makes sense since it doesn’t do anything

[remote sparrow]

yeah, surprising a bit but wm is one of our stronest talents

around 2.5%

[valid robin]

Unoptimized talents

[remote sparrow]

eh, i mean compared to things like finality, db, pv, inev, lingering, gb etc

rotten is like 3% unoptimized (ish), db is like 4%, etc.

[mental dirge]

you know whats funny

even if they fix tornado + blades we might just still take it just to not take VT

[remote sparrow]

whats funny is that I was right

about running RS

in single target

[mental dirge]

because you have to take one of those for flag

[valid robin]

I mean rotten and pv are completely reliant on what spells you use for it

[remote sparrow]

yep!

[mental dirge]

okay yea they need to numerically buff like

[valid robin]

And if that’s just random

[mental dirge]

a shit ton of our talents

[valid robin]

It’s not going to be great

[mental dirge]

when its VT vs Tornado to reach flag

and tornado is probably better

[remote sparrow]

its not entirely random, rottens always used with strike right now

[mental dirge]

thats a gg

[valid robin]

Vt is a shit talent

[remote sparrow]

oh you have no idea how shit VT is yea

[valid robin]

Doesn’t make sense

Why does it say all magic dmg

[mental dirge]

bcuz they probably thought it would be too redundant with deeper daggers to just read shadow dmg

so they made it different and its magic

[valid robin]

It only made sense as a talent when it affected trinkets

[mental dirge]

magic also includes nature so

i mean its just a worse DD

and DD isn't good as is rn lol

[valid robin]

Yes

It’s trash

[mental dirge]

VT realistiaclly needs like an extra sentence on it

not just the 5%

[remote sparrow]

its 0.5% weaker than RS in single target LMFAO

[mental dirge]

should turn shadowstrike to shadow dmg tbh

[valid robin]

Kinda broken

[mental dirge]

shit DB turns rupture to shadow

and were not even looking to take that

so xdd

[valid robin]

Ur increasing it’s dmg by like 40%

[mental dirge]

ok and

talents trash right now

make it not trash

but not just by 5x'ing its number

[remote sparrow]

rotten increases it by 50% and gaurantees it crits

its still a shit talent

LMFAO

or actually 2 strikes by 35% each

[valid robin]

Rotten doesn’t do that to all strikes

[remote sparrow]

it does it to half of your strikes though

you get like 4 strikes per dance now

and dance goes with sod so, half of them

[mental dirge]

what would you rather have, VT amp magic damage by 15% or whatever they make it or have it do something at least kinda cooler flavor wise

[valid robin]

It’s probably going to be 10%

And that’s it

[mental dirge]

probably

[remote sparrow]

i hate all this useless passive non interactive 1% amps

[valid robin]

It’s a middle of the tree talent, it won’t do anything cool

[remote sparrow]

anyawys its not really TC related ig

[mental dirge]

outlaw has ace up the sleeve as middle of tree talent

driving a whole build in tandem with crackshot

weird innit

[valid robin]

Ye but do they have 10+ aces

[remote sparrow]

not true, uhuh is kinda middle of the tree, cto is middle of the tree, exsang is kinad middle of the tree, etc

[valid robin]

A few middle tier talents being special is fine

[mental dirge]

uhuh is 3rd tier

[remote sparrow]

yeah but its close

its not like our 3rd tiers do anything either

pv is 0.2%

[valid robin]

“It’s close”

[remote sparrow]

lingerings same

dd is like 2%

[valid robin]

I mean idc about the % of talents

[mental dirge]

what are our special middle talents?

[valid robin]

Honestly

[mental dirge]

+2 dance duration?

or the talents we require to function like relentless strikes

i feel very special

[remote sparrow]

its not just %, its that theyre both weak and non interactive.

we have like 3 semi exciting talents in the entire tree

[valid robin]

Ye I am just saying, we had these talents for a long time, the specs dmg is not reliant on % of specific talents

[remote sparrow]

and 2 of them are new

[valid robin]

They are all boring

More importantly

[remote sparrow]

new blades and shadowcraft are exciting to me

flags fun too

everything else sucks

[mental dirge]

Dark Shadow best most sub talent in entire tree

always has been always will be

[valid robin]

Danse macabre is a nice talent

[mental dirge]

yep

[remote sparrow]

DM is nice, and strong

its 11-12% rightnow i think

its interactive, cuz you kinda play around it, but its not too oppressive, it fits the theme, and is strong

[mental dirge]

not every talent needs to morph your entire gameplay or approach to the game but i'd at least like to be excited about specing them

[valid robin]

I think sub lacks a clear identity in the spec tree

It’s still just a mix of things

[narrow leaf]

clones from sec tec have to hit in dance window to get the amp right?

[mental dirge]

ye

[valid robin]

Trying to go for shadow dmg amps, but there is a lack of shadow dmg

[remote sparrow]

well, the shadow damages got nerfed

40% -> 25%

so even less emphasis than before

[mental dirge]

yep we got bent by the 2 point -> 1 point changes

[valid robin]

If gb is not worth taking and lingering is shit, where is the shadow dmg

[narrow leaf]

Just dropped PE for it and got about an extra 400 dps

[mental dirge]

yea idk these talents are really fuckin down bad LMAO

this is absurd

RS and Tornado more powerful ST talents (likely even after tornado fix if it happens) than most of the other middles

[remote sparrow]

idk what you did, but pe is still a few 100 higher than vt for me.

[valid robin]

Lingering can prob become 100% again

[remote sparrow]

^ wouldnt matter

[mental dirge]

negative

do not make lingering potent

[remote sparrow]

and if it did matter

[mental dirge]

for love of god

[valid robin]

It adds shadow dmg

[narrow leaf]

I'm playying that heavy bleed build

[remote sparrow]

itd just make the playstyle like before and worse

[valid robin]

Why

It just makes downtime less shit

[remote sparrow]

there are other ways of making shadow damage matter.

[valid robin]

Doesn’t nerf dance dmg

[remote sparrow]

dance can litearlly turn everything to shadow! ez

[valid robin]

How?

Ok

[mental dirge]

bro just buff DB DD and VT

that as a ripple buffs gb

which then buffs lingering

[valid robin]

They have to be really high numbers

[mental dirge]

ok?

[valid robin]

What % of dmg is shadow

On ptr

[mental dirge]

isnt it like 40 or some shit

[valid robin]

On ptr?

[rugged solar]

Shadow strike could be shadow damage^^

[mental dirge]

idk i recall armin saying its like 40% maybe misremembering

[valid robin]

10% blades, 10% sectech

[remote sparrow]

its about 40% including tier

[valid robin]

Then what

[mental dirge]

shadowed finishers?

hello?

[remote sparrow]

and rupture with db I think

[mental dirge]

fifan assa

[valid robin]

Shadowed finishers was gutted

Is it 5-6% of total?

[mental dirge]

no

bro evis is like 35-40% of your dmg

[valid robin]

Link some evidence then

[mental dirge]

https://www.warcraftlogs.com/reports/9kNhxdzjnXvgm6FT#fight=last&type=damage-done&source=1

[valid robin]

Private

[rugged solar]

The Sim I linked earlier has it around 30-35%

Shadow damage

[mental dirge]

this is 5 min

[rugged solar]

But I am on mobile, so

[mental dirge]

11% of the evis is shadow

40% of the dmg in that is shadow

not including the racial

oh

[golden bay]

was rupture fixed to be shadow dmg with DB?

[valid robin]

Ur playing gloomblade tho

[mental dirge]

and not including the tier

[past fable]

Thats also not including 2p which is also like 6%

[mental dirge]

but that doesn't matter since it doesnt double dip iirc

or wait

its based off phys hit

so it does matter

so yea add the tier as well

[valid robin]

I specifically said without gloomblade tho

[mental dirge]

ok so without gloom its 39% if including tier

[valid robin]

Ye

So is sub a shadow dmg spec or not?

Doesn’t seem like it

40% kinda weak

[mental dirge]

actually if we wanna be precise it was 41%

bro what

so if its not over half ig its not a large portion?

[valid robin]

With a 10% shadow dmg buffer no

As capstone

5% vt

16% dd

[mental dirge]

all youre doing is citing numerically bad talents

DD is 8%

[valid robin]

What??

[remote sparrow]

1 point now

[golden bay]

so was rupture fixed?

[remote sparrow]

so 8%

[mental dirge]

went from 2 pt to 1 pt

[valid robin]

Holy shit

What an awful change

[remote sparrow]

yes

[mental dirge]

yea lot of the 2 to 1 points literally gutted our talents lol

[valid robin]

They could increase all these talents by at least double

[mental dirge]

we get to spec more but they all suck

[turbid pulsar]

db should just turn all of our dmg into shadow

[remote sparrow]

^ during dance

[turbid pulsar]

now thats a capstone

[valid robin]

Full shadow dmg evis

[mental dirge]

that also has the added bonus of making autos not miss

[valid robin]

Dark brew name..

Make it a sub only tea

That gives you 100% shadow dmg

For x seconds

[mental dirge]

but yea all this is some fantasy land hopege

if they functionally change anymore talents for sub ill eat a rock irl

just buff these awful talents lmao

[valid robin]

Ye they wont

[mental dirge]

RS and Tornado should not be front runners of single target

[valid robin]

They could buff them by 100% across the board

Problem is they are still terrible and boring

[remote sparrow]

^^^^ yes

[mental dirge]

yea i mean thats a separate problem that they can tackle in 11.0 if they want

[remote sparrow]

why do i have to talent into 17 different 1-2% essentially verse talents

[valid robin]

Well

Why not

[remote sparrow]

btw im expecting shadowcraft to be like a 3.5-4% talent

without the refund its about 2%

so im guessing another 1.5-2 from the refund

[valid robin]

It kinda feels like the entire sub spec except for dance dmg and dance talents is already a complete spec

And that’s why the talents are so bland

[mental dirge]

yea i mean i am okay with the specs gameplay as long as i can play the talents i want to play

i mean i'd even be okay with rotten if it wasn't also force feeding you dust

[remote sparrow]

well talents are taken usually to change things slightly, enhance little things, make you play a little differently

[mental dirge]

the 10.2 gameplay is fun

[turbid pulsar]

historically thats allways been the case for sub

very strong core, but very bland borrowed power

[valid robin]

Like you only really need dance dmg, dance duration, blades and relentless. And its a playable spec

[valid robin]

So much of what makes sub, sub is just baseline

Prob has the most baseline stuff

Out of any spec

[past fable]

Replicating feels like its a strong part of the aoe core nowadays despite being a talent

[remote sparrow]

^ you can find ways of enhancing things.

[valid robin]

Make rs baseline

Make relentless