So I’ve been updating my old Belgothian Infiltrator and the subject of Amarasta’s Quick Cut came up in our Discord. I decided to give it some proper tests and made some dummy kills.
Felt like a tiny teeny WPS gain but might be wrong here. Even if it is a tiny DPS gain versus a single-target, second Belgothian Slicer wins here and it’s not even close: 1) It buffs AOE WPS and has identical single-target damage 2) It gives physical resist (from Dual Blades bonus) and armor.
Conclusion
So Amarasta’s Quick Cut is still useless on Belgothian Nighblade. And Belgothian Blademaster doesn’t even consider taking it because it has steroided Markovian’s Advantage with 25% proc chance. I don’t know if the animation is still slow or is it the low weapon damage that doesn’t give flat damage proper multipliers but something has to be done with it.
And if new patch is going to increase its proc chance at higher ranks without fixing it it’s going to be even more useless.
Nice tests! I have conclude AQC is rarely worth it, especially if you don’t have the skill points to max it. Deathmark perhaps can use it, on Blademasters should be DPS gain, but elsewhere either DPS loss or the same.
Check out the part where I tested Derp Slicer that gives 15% weapon damage to this wps. It’s not a conclusive test since different sword is used but I expected that it at least makes it a meaningful investment for single target dps. But it doesn’t. I will do Deathmarked tests tomorrow probably if I have time.
If i understand correctly, slow animation of this WPS causes the effect: your supercharged WPS, Markovian and Belgo, start procing less often per second when the pool is burdened by 20% chance of another attack with slow-ish animation. Since the dmg increase on AQC itself is minor compared to regular attack, it is negated by fewer (per second) procs of stronger attacks.
Tests for the builds without WD buffs to other attacks may give slightly different results. Regardless of that, the skill is very, very weak. It’s been known for at least a year (since i joined). Hopefully next patch will change things.
The only meaningful usage of this skill i’ve found so far is the increased rate of Cadence charges.
I have just tested my Deathmarked Trickster, but i’ve put on the set helm instead of Korba to have the skill buffed. 3 dummy kills in a row without using any skills except Pneumatic Burst to have a clean test:
I know it’s a sizeable dps gain on my pierce savagery trickster… but it is able to get 18/8 with like 3 skill point and also has derp slicer mod. Ive noticed the last 3-4 skill levels are pretty huge on scaling the flat bonus.
Perhaps a slight buff plus a change on how the flat is distributed across skill levels?
If you compare the list WD% and flat dmg at 8/8, AQC supposedly does more damage than shears and whirling death, plus it has crit. I wouldn’t describe this as “minor compared to a regular attack”. To be exact regular attack is only 150% WD, whereas 8/8 AQC is 270% WD and 204 flat pierce, yes?
The problem must still be animation speed (after all the description even says “strike thrice in the space of one attack”), or else it really does need even more damage to be useful.
As a comparison though, how much does any one other wps contribute to dummy kill time? say whirling death?
Yes, it seems that animation is the issue. And it’s more prominent on weaker rigs.
Regular attack when dual-welding has a 50% chance to hit with both hands at a pace of one hit so it’s 1.5x1.5 = 225% WD per attack compared to 0.82 * 3 * 1.5 = 369% WD + flat on 5/8 AQC. On paper it looks like a dps gain and when i slowly hit the dummy with dmg display i see the numbers i’m ok with: 3 hits with dmg almost matching regular attacks. It should increase single-target dps. But it doesn’t.
In my opinion, basic stuff like this should not be tested and discovered by regular players. It’s a very simple test and the issue is obvious.
Attacks like Whirling Death and Belgothian Shears are multi-target and can not be used in single-target dps test.
So what? AoE skills contribute to dps. The point was to try and gauge what dps boost AQC should actually be.
Patch 1.1.7 has already confirmed wps adjustments. Hopefully AQC will be re-examined, but what is the best solution for this problem?
Whilst most would agree AQC should be a dps gain (even at 5/8), belgo/melee pierce hardly needs a dps boost does it? I think this consequence may be why AQC already got an animation boost (which apparently only upgraded the skill to be less underwhelming).
I will be very interested to see how 1.1.7 handles balance given the wps adjustments.
Maybe my wording wasn’t correct. I meant, the value of multi-target skills cannot be measured by single-target tests. But yes, both attack are, too, a little slower than regular attacks.
A basic attack animation is 18 frames before various modifiers. Belgothian’s Shears and Amarasta’s Quick Cut are both 28 frames before modifiers.
The player itself has a 1.25 multiplier on animation speed. In addition, Belgothian’s Shears and Amarasta’s Quick Cut have a 1.20 multiplier on top of that.
So…
Basic attack: 14.4 frames
BS and AQC: 18.7 frames
A second is 30 frames.
Leaving extraneous bonuses aside…
BS at rank 8 is 128% weapon damage and 55-86 Physical damage and strikes with both weapons
AQC at rank 8 is 90% weapon damage and 68 Pierce damage and strikes 3 times with MH, OH and MH again.
So in terms of damage per hit, with two weapons that do 100 damage with additional no modifiers, you would do…
Basic Attack (assuming 50% hit with both weapons and 50% L or R hand hits): 150 average
BS: 397
AQC: 474
DPS, assuming every hit was of that type:
Basic Attack : 312.5
BS: 636.9
AQC: 760.4
This of course becomes more complicated once you include armor, flat damage bonuses, critting, item skill modifiers.
If there is some impact from lower performing machines on animation, then that is something I cannot parse with just math as it would take some deep digging into why that might be on the engine end, assuming such a claim can be backed up with evidence.
I play on everything low. i7-9750h RTX 2060 win10. My test was done in the less demanding area for me - Fort Ikon.
100 dmg per weapon? Common now.
I don’t know what’s the issue either because when i look at the dmg numbers popping up i’m ok with the result and it corresponds with the math - each AQC hit is just a little lower than one regular hit (10% WD of a strong melee build is still often higher than AQC bonus). The game has so many variables, it’s easier just to test it live on an actual build.
I’ve tried DM Trickster with AQC fully converted and buffed and without it and dummy kill time didn’t change. This is actual result. OP has also tested his actual build and had similar results.
Impartial observer because I basically never play melee, but thoughts on this investigation start:
Animation playing slower should be an easy find if you’re already recording all these runs. Grab one of the videos and go frame by frame for each WPS animation to capture if they are taking the same time. Looks like BS and AQC are meant to be the same frames, so can just test across those two. Couple people testing in here so will give multiple PC viewpoints if there is some weird animation bug happening ala Zantai’s comment
For the tests themselves, some mild concerns. Relatively short TTL on dummy which is going to make crits have more impact. Can’t see the other runs so not sure what the variance is with each setup. Also not sure where skills were pulled from to get to 5/8 (only 2, so probably not much) and 8/8, but hoping those weren’t from other dps impacting skills. The derp weapon causes other dps skills to change so without running the math it makes an apples to apples comparison harder.
Also appears you don’t start attacking the dummy at the exact same time in each video, which on these small margins is also going to have an impact. I’m not a statistician so I’m not sure how many runs you’d want to capture to get a good data set to cut out the variability. If going for dummy kill time I’d probably want to mod in something that’s going to take 5+ minutes of hacking away to remove some of those peaks/valleys
You’ve already got what you need for hypothesis 1 so that’s the road I’d go down if I were looking at this
On paper at least AQC should hit similarly hard to Execution and pre-nerf notched bone (especially with a slicer). Do you remember the crit multipliers?
AFAIK AQC used to proc cadence, so it’s three basic attacks strung together really quickly(?); is it possible that it has 3x separate OA rolls, and isn’t critting on all 3 hits consistently?
@Superfluff care to weigh in on yet another AQC feedback discussion?
So then using the real numbers to try explain the infiltrator video. It will be a situation of AQC vs basic attack as wps pool is not saturated.
High weapon damage per hand on the build deals ~546 pierce damage ((15,203/27.8) minimum damage).
Basic attack per hit: 819
AQC: 1678
AQC adjusted down by extra frames: 1191
DPS, Same assumption as Z (assuming every hit was of that type):
Basic DPS: 1706
AQC DPS: 2691
Note AQC % DPS increase is much lower than Z example (as AQC flat pierce is diluted by huge weapon damage stacking),~58% vs ~143%
However we all know the assumption is silly (and these DPS calculations not very useful) as only 1/5 hits are AQC, and on average this means there is less than 1 AQC per second procing at max attack speed.
The damage difference is therefore reduced further. Final damage increase per second gained from using 8/8 AQC over basic attack after accounting extra frame and procs per second: (1191 - 819) x 0.8) = 297 flat pierce per second.
Observation: 297 flat damage per second (scaled up by +% pierce and cunning is only ~8250/s) , in the grand scheme of overall dps, this is very low. I think this is the reason for the videos being “unnoticeable”. Given the low dps boost many experiments would be needed to truly detect this positive change.
However, it is clear that the damage boost is far greater the lower your weapon damage, so AQC is really a levelling skill, or advantageous for those with lots of free skill boost, item mod, low weapon damage or literally nothing else to invest in.
Actually i did two more series of tests of 5 dummy kills in a row with Trickster Deathmarked - with and without AQC - and had -6% time with AQC both times. So it is in fact a dps gain.
10/8 AQC, fully converted and with flat bonus, one of set signature, single-target skill gives 6% bonus dmg against single target. Is that enough?
Edit: the tests were done without LAA, SS and Wind Devils for clean results. LAA increases WD, SS and Wind Devils aren’t affected by AQC, so in actual gameplay total dmg gain of AQC will be even lower, less than 5%. And all that against one target only.
