Stricken Question (0.9 ICD)

Demon Hunter
I had discussion with @Ketsuen before about this - but in a thread where it didn't belong.

Common knowledge:
Stricken internal cooldown (ICD) is 1/APS.

My interpretation wat this means 1/actual APS (so breakpoints are included).
Which means: There is no cooldown at all.

Example: With 3 actual APS cooldown would be 1/3= 0,3333 - so you'd get 3 stacks in 1 second- 3 aAPS. You might lose one potential stack if enemies move towards you (so less travel time and getting into the ICD phase). I've seen that it can happen to lose a stack. My explanation would have been less traveltime, but maybe there is another reason.

However - some people did find out through dataming that ICD of stricken is actually 0.9/APS. Can't confirm it though. But seems legit.

What does this mean?
A)Nothing because not actual APS is used for stricken cooldown (which I don't believe, because stack updates need to hit a full frame and screen APS shouldn't be used by the engine IMO), so it is just there to get the 3 stacks.
B) We get new "stricken BPs" - which might be relevant for S6 group meta.
C) Something different and both is wrong.

How could you hit above your own APS or is this even possible ?
Well - not with one skill like impale. But if you could hit with multiple skills simultanously it could work. Stricken has no procc coeficient and works on enemy hit. Don't know if those skills work, but for instance: Shuriken cloud, sentries and vengeance would allow you to hit above your own APS. Or those rockets from MS.

Testing this is pretty hard, as I lack the legacy +min ring in NS and also lack ressources in season to get the right gear.

Anybody an idea how this works?
Sry not that much time right now but maybe those links below give you some answers for your questions: (Quin69)
Yes. ICD scalars. For most procs this is 1.0, but the most commonly used exceptions are Bane of the Stricken with 0.9 and Taeguk with 0.8.
This means that Bane of the Stricken can actually be stacked slightly faster than we originally assumed, because its ICD is ceil(54 frames / APS). This means you reach an ICD of 12 frames at 4.5 APS and an ICD of 11 frames at 4.91 APS for Bane of the Stricken.

Note: Snapshot (dynamic) APS for lower stricken ICD values via Potion don´t work anymore.
I can't say anything to the actual ICD, but these links are outdated.

Please note that this does not include attack speed modifiers for skills or attack speed buffs that do not appear on your character sheet. This means that if you get non-sheet APS or use skills with an attack speed modifier that is greater than 1.0, your hits and the ICD may get out of sync.
This is not true anymore. Stricken scales with individual APS modifiers like Yang or Hunters Wrath.

When they changed that, they may have changed other parts of the internal cooldown mechanic as well ...
Thx Loque, I know those links, but they don't explain how it works.
Don't know much about monk shizzle, though.

If stricken updates at 54 frames and not at 58,0645 frames like impale - this would mean different BPs for stricken (aka scenario B). But if it actually works this way - this is the question here.
this should be fairly simple to test:
1) equip stricken and testing gear (low dmg weapon + min dmg jewellery for same dmg ever shot)
2) get to an insane level of APS where a static ICD would be an issue, I'd aim for 3+ aps
3) start shooting the monster, if:

a) every shot keeps increasing in damage then the ICD is 1/aps, and therefore the ICD can be ignored for impale (still relevant to things like RF/Strafe)

b) certain shots retain the same damage as the prior shot, then the item was on ICD at the time and you didn't gain a damage stack. This means it has a static ICD with no aps scaling. Then a lot more investigating would need to be done to figure out the statc ICD value.

Test like this have been done in the past for things like Buriza. Doing this people were able to determine it has a static ICD and going over 2 aps results in getting ICD locked and wasted shots that don't get pierce.
06/19/2017 12:36 PMPosted by TastySouP
every shot keeps increasing in damage then the ICD is 1/aps, and therefore the ICD can be ignored for impale (still relevant to things like RF/Strafe)

Can't do it this way because I have no min ring so there is damage range, which is huge. So I can't calc damage and compare. And there are vengeance hits mixed in. And impale is kinda mandatory (because it is the slowest skill so you get 2 BP difference if stricken BPs would work), which needs either very high RCR or Karlei + dawn (so more damage range - even if you roll damage off).

But I think I found a test scenario that could work. Just takes 4 mins+ at rakanoth (or so and I have to figure out damagelevel to get this kill time and not run into enrage timer and I'm not sure what it does to stats). Maybe the spiky dude that breaks through walls in cementary (no enrage timer) would be better, but he has to be found first. Higher kill time would be better to see % damage difference.

Just need to find some test gear first. Gonna try if I find the right gear. Just needs to take enough time at Rakanoth to get enough hits, so damage range evens out and stricken bonus matters enough. If stricken BPs exist, ~+14 gear IAS shouldn't be much faster in my scenario with dawn equipped.

Edit: It is more complicated as I have to compare no vengeance vs vengeance and then hit the same DPS and then there would be a huge 1.4 factor in there. So less stacks which makes it way more complicated. Making it up with paragon and dex is not so easy.

Edit2: Without S2 Vengeance damage plays a to big role. With S2 - killtime is to fast even with !@#$ty weapon. So not that easy to find a good build. Need more gimp gear.
06/19/2017 01:11 PMPosted by Kindergarten
Can't do it this way because I have no min ring so there is damage range, which is huge.
I do :D
3.1 aps, same dmg every hit, and a nice socket for stricken. I'll test it when I get the chance later.
I'm afraid those won't show you all we would want, namely whether it would be 1/actual APS (the BP APS; hence the same delay for impale and stricken ICD), or 0.9/sheet APS (without BP; which could perhaps mean some optimization could be theorized as stricken ICD would be shorter than impale rate), both of those should result in each impale doing more and more damage (as ICD would be shorter or equal to impale rate).

Of course if some hit do the same damage as previous one, it would support the idea of 1/sheet APS ... (with BP bonus APS making attack during ICD)

Same as Kindergarten I started D3 later and so I do not own min damage jewels :(
06/19/2017 05:11 PMPosted by Ketsuen
I'm afraid those won't show you all we would want,
now that you mention it I was reading it as a static ICD of 0.9 seconds, not 0.9/ApS. So yea you're right, my test wouldn't figure that out, but with that said:
the smaller dividend will always produce a smaller number, so that means if strickens ICD is 0.9/aps its cooldown would be shorter than our aps can generate attacks.
if you has 2 aps it'd be

1/2 = 0.5 shots per second
0.9/2 = 0.45s cooldown

it would always be off cooldown before the next attack would strike would it not? That means it doesn't matter if the ICD is 1 or 0.9, both will correlate to a stack every time an attack hits.
Well if stricken is 1/sheet APS ... then it would matter in some cases as BP will give couple more % over your sheet APS.

For instance say you have 25 IAS with 7% weapon (or 2.00625 sheet APS) you hit 29 frame BP for 2.069 actual impale per second, e.g. attacking every 0.483 sec while stricken ICD would be 0.498 ...

I really doubt this is the case, hence the follow up question whether it is:
a) 1/actual APS
b) 0.9/sheet APS

Both those allow you to get 1 stack on each impale, however the later would leave a tiny window between stricken ICD and your actual next impale ... for which Kindergarten wonder if you could somehow manage to get even more stricken stack than your impale attack rate (from other effect such as vengeance rocket and whatnot ..)

While I'm not really interested in that aspect, I would still like to know as much as possible as to how mechanic works (since sure for current impale it might not matter, but perhaps down the road on other build perhaps it will)
I see I once again misread a post for its question. I gotta work on that, I can't count how many times you've corrected me, ketsuen, in the past little while >.< thanks for clearing up the actual question for me.

Well, I have an offer for both you, ketsuen and kindergarten. I know you both play on EU, but I do have a few spare testing rings and weapons on the US servers, and I like the work you two do with testing. I think the items could be quite valuable to you, so if you wanted I could power level you to 70 and give you each a ring.

And another thing you could do for testing: You get this item when you make a new char. it can only produce 2 different dmg numbers, so it makes testing much easier and allows you to verify info without having legacy test gear. You can obv get that on EU, so that could help if you don't want to flip flop servers.

Like I said though, I have a few spare test rings if you'd like. You would just need to use the US servers to test stuff
06/20/2017 10:16 AMPosted by TastySouP
Like I said though, I have a few spare test rings if you'd like. You would just need to use the US servers to test stuff

Appreciate the offer - but on US server I'd have to farm all gear, do bounties to extract stuff, get decent paragon and probably get some ressources for rerolling. I don't even have enough yellows in eu season to reroll gear ;) This would take days of hard work just to get a point to test and then find out it was all a waste.

06/19/2017 05:59 PMPosted by Ketsuen
I really doubt this is the case, hence the follow up question whether it is:
a) 1/actual APS
b) 0.9/sheet APS

Well thats not really what I meant. I have seen in tests that stricken updates with every hit (unless enemies moving towards you). So IMO there is no cooldown at all (for nomal skills, channeling skills are a different thing. Would be the same with 0,9/sheet APS. The difference with sheet APS used would be very small, IMO.

I don't believe in sheet APS used, because the engine should internally always use aAPS. But there was a time where Yang and Hunters didn't work with stricken - so that may speak for sheet APS. 0,9/sheet APS would just be a fix that stricken works. I think this is Ketsuens interpretation.

My thinking was that it could be 0,9/aAPS.
Example with s6: You need 81 gear ias + 7 IAS weapon to reach 3,0 aAPS.

If "stricken BPs" would work as above, you'd get to 3,0 stricken aAPS with only 69 Gear IAS. Over a long time and with high enough stricken, 69+vengeance should be very close to 81 DPS/1,4 without vengeance, although 69 gear IAS would only mean 2,73 APS. Because with 500+ hits or so the main damage comes from stricken and the normal hits (2 BPs) faster don't matter much anymore. Which then would mean for stricken builds/Boss killers (S6 group meta, UE nades solo) in high levels that the currently used BPs are "wrong".

So I'm gonna try something like that. Problem is that I have to run it with S6 (so vengeance damage becomes irrelevant) but everything dies to fast with this setup. So I need better gimp gear. Without tripple hit quiver, you also need more RCR.

Test scenerio with min ring: Do 100 or x attacks without using vengeance. Look at the damage of the final hit.
Then do the same with vengeance and look up the damage of the last impale hit.

This wouldn't proof whether it is 0,9 sheet or 0,9 aAPS but at least that there is something. Tricky part there is that you can't use Dawn and Karlei (because you need gimp weapon with super low range so that min ring works). So you'd need huge amount of RCR for perma impale.

In the end it is probably not that relevant - at least for me - I just play publics and have no meta group. But it would be interessting to find out, how it works.

Edit: For testing stricken Rakanoth is bad. He has just 3 adds or so - but he also ports. And you'll gonna hit those adds anyways.

Best Meatpuppet/Testingspot in the game is the spiky dude in cementary who spawns at level 2.
Think he is always there - you just have to check all 3 levels. It does happen sometimes that he hits you and you get throwed back - but most of the time he will attack follower or pet.
I think I thought of a way I could test it, maybe you guys can elaborate more and let me know if I missed something that would flaw this:

Using Rapid Fire, you get 6 ticks per attack cycle, or 6 ticks per 1 aps. If you sit at 2 aps you could expect 2 stacks of stricken per second because of the ICD, but too shot 12 arrows overall. When sitting at 2 aps, if the ICD is 0.9/aps, you would expect the cooldown to last 0.45 ms, which means after 4.5 seconds of channel you would squeeze in an extra stack.

So basically channel for 10 seconds and see what the damage value is. It will either be equal to 20 stacks, or 22 stacks. If its 22 stacks, then that means the ICD is 0.9

I don't think I missed anything?
K here is my test result, not perfect but the best I could do:

Test 1: 225 secs on spiky dude. 91,2 Gear IAS with 6% dawn (and gogok). So "normal" 3,3333 aAPS. Stricken at 91. 158.042 screen DPS. No Vengeance used at all.

Test 2: 223 secs! Same setup but: 77,2 Gear IAS (with Gogok). So "normal" 3,0 aAPS, but theoretical 3,3333 stricken APS. Perma vengeance. Screen DPS should have been 158.042/1,4= 112.887. I had 112.784.

Probable error factors: Damage range of dawn is 135-764. Really low critchance: 19% vs 17,4% so there is some possible deviation here. Maybe I got more lucky on Test 2.
But after so many seconds with 3,0+ APS it should even out, IMO. Thats more than 600 hits.
+- some seconds variation - sure.

My interpretation:
That Test 2 was faster is related to RNG and i played a little bit better. Should have been a few secs slower. But normally there should be a big difference. 2 more BPs with stricken would matter.

If stricken would be based on 1/actual APS Test 1 should be far ahead. If you use stricken total damage formula, Test 1 should be ~ 20% faster or so.
Total damage stricken formula for impale ONLY is sf*(N^2+N)/2+N.
With vengeance: You can subsctract the 0 damage hits hits of Vengeance - but it won't be much after 200secs+ in the end

N = Number of stacks. sf=stricken factor - 0,016 for stricken 80 for instance.
As I used full S6 build (noS2 and no tripple hit quiver though) vengeance damage shouldn't play a big role.

Made vids of both runs - but they will take some time to upload. They are very boring though.

Maybe I'm missing something here or I did something wrong - but it looks like that stricken works with 0,9/aAPS and vengeance applies the additional stacks.
BPs used:

Testsetup - Rapid Fire
272.5 avgDmg (272 - 273)
* 40.62 (3962 Dex)
* 6.85 / 6 (685% Rapid Fire / noRune)
= 12637,05125 (12613,864 - 12660,2385)
= ~12,6k (12,6k - 12,7k)
lvl 100 Stricken (1.8% = *1.8%/100+1 = *1.018)
Rapid Fire on Singletarget captured with Fraps
Stricken Formula
(Stricken% * #ofStrickenStacks / 100 + 1) * BaseDamage
A with 2.0016 APS (1.2 * 66.8%) [4 FPA = 15 aAPS]
after 10sec = 15.1k [11 Stricken Stacks -> (1.8% * 11 / 100 + 1) * 12,6 to 12,7]

B with 1.20 APS [8 FPA = 7,5 aAPS]
after 10sec = 15.1k [11 Stricken Stacks -> (1.8% * 11 / 100 + 1) * 12,6 to 12,7]

Note Can´t explain this. It might work different with channeling skills, idk. Or maybe it works different (or generally not) with the BP-Chart for Rapid Fire on d3planner but not sure about that at all.
Testsetup - Impale
272.5 avgDmg (272 - 273)
* 40.62 (3962 Dex)
* 7.5 (750% Impale / noRune)
= 83017,125 (82864,8 - 83169,45)
= ~83,0k (82,9k - 83.2k)
lvl 100 Stricken (1.8% = *1.8%/100+1 = *1.018)
Impale on Singletarget captured with Fraps
A with 2.04 APS (1.2 * 70%) [29 FPA = 2,0690 aAPS]
after 10sec = 114k [21 Stricken Stacks -> (1.8% * 21 / 100 + 1) * 82,9 to 83,2]

B with 1.20 APS [49 FPA = 1,2245 aAPS]
after 10sec = 102k [13 Stricken Stacks -> (1.8% * 13 / 100 + 1) * 82,9 to 83,2]
Testsetup - Impale (with Vengeance/Dark Heart)
272.5 avgDmg (272 - 273)
* 40.62 (3962 Dex)
* 7.5 (750% Impale / noRune)
* 1.4 (40% Vengeance)
= 116223,975 (116010,72 - 116437,23)
= ~116,2k (116,0k - 116,4k)
lvl 100 Stricken (1.8% = *1.8%/100+1 = *1.018)
Impale on Singletarget captured with Fraps
A with 2.04 APS (1.2 * 70%) [29 FPA = 2,0690 aAPS]
after 10sec = 160k [21 Stricken Stacks -> (1.8% * 21 / 100 + 1) * 116,0 to 116,4]
Sidenote 160k / 1.4 = 114,3k
B with 1.20 APS [49 FPA = 1,2245 aAPS]
after 10sec = 144k [13 Stricken Stacks -> (1.8% * 13 / 100 + 1) * 116,0 to 116,4]
Sidenote 144k / 1.4 = 102,9k

Note Since the number of stricken stacks are exactly the same without and with vengeance available, it pretty much tends towards that vengeance hits through Side Guns and Homing Rockets does not add any additional stacks to the stricken gem but they are affected by the stricken multiplier (1.8%)!
Interesting discussion. :)

So the idea is: If the ICD is 0.9/aps, the Vengeance should add more stacks?

Here is my first lazy contribution to that. I did not checked any breakpoints. Just fired some EA (using a cubed Krider) at a Zombie. After some time with vengeance (using cubed Dawn and Zodiac), the damage is still in sync with the damage without vengeance.

I will repeat the test later without the fault that the zombie moves towards me in one part; and with Impale as well. ;-) // Edit: even if the movement is on the no-vengeance side, so it would favor the result "vengeance do add more stacks"

But it seems to me that additional vengeance hits do not add any stacks ...
Ok, another one.

My idea: If the internal cooldown depends on the sheet APS, then the results should differ, when you change your sheet-APS, but not your actual APS. Using a level 1 hand crossbow with 1.2 APS, you stay within the same breakpoint from 39% IAS to 42% IAS (source: Dawgs calculator, verified in the video).

So I fought 4 zombies for some time using 39% vs. 42% IAS, without vengeance and with vengeance. I did not manage to have the same duration of Vengeance in both cases, but Vengeance was up for ~100 seconds in each case.

After that, the damage numbers in all 4 cases are the same - in all 4 cases I had applied the exact same amount of Stricken stacks at the end.

Even if the internal CD (using the formular 0.9/sheetAPS) in that scenario varies only from 0.5395 to 0.5281 seconds (difference of ~0.0114sec) - after roughly 200 attacks there should be a difference of at least one or two stacks. But there is none ...
+1 for the video evidence Hebalon, unfortunally i can´t upload stuff with my !@#$ty 0.5Mbps, but the video perfectly matches my observations and shared informations above. Thanks alot for sharing!
Yes good test video from hebalon (also nicely with the syncing 2 vids) - so we can ignore this 0.9 ICD thingy and stricken works simple: There is no cooldown at all for normal skills. I was also sceptical with this 0.9 ICD thingy - good we can ignore it.
But also great contribution from the others answering!

Maybe it exists for channeling skills - but not sure how they work (but here are old threads that explain them afaik and show number of ticks).

Still wondering though why in my testruns both end up with the same time, although 1 has 3.0 APS and the other 3.33 APs where stricken should matter (~20% more damage at same time) . Maybe it's the vengeance damage itself (boosted by stricken) - so it is not just "0". But also sounds a strange as an explanation.

Don't think it was RNG after 660+ hits.

Join the Conversation

Return to Forum