VSK & Time of Day

By BigStupidJellyfish | 2022-May-15

Skill Details^#

Double Alert ICD 5s/CD 8s

When this skill is activated, ~~transfer~~ [add] 20% of accuracy to damage for 5 seconds. The first 3 attacks after this skill is activated deal two instances of damage.

It's rather important to understand a doll's skill before analyzing them. However, we run into an issue straight away: how VSK's accuracy-to-damage conversion works isn't clear.

The in-game TL says it "transfers" 20% of her accuracy. But other translations say she just "gains" damage equal to 20% of her accuracy. So, does she debuff her accuracy?

Also, what value of her accuracy does it take? Is it her base stat? With fairy aura/tiles/buff skills? Is that increase affected by FP buffs, or is it a flat amount? How about at night?

(There are a few even more minor questions you could ask but those start getting super difficult to test for.)

Accuracy transfer

VSK's skill says it "transfers 20% of [her] accuracy to damage" when active. However, unofficial translations say nothing about a transfer. Further testing indicates her accuracy remains the same in-skill. It is likely her EN skill description was mistranslated (not like that hasn't happened before).

Video source. Note also the lack of debuff arrows.

We can reject the "debuffed" case with p<0.01.

FP buffing

Note: in this section, some numbers are slightly rounded/simplified. Damage values and ranges may be off by ~1. Should still be precise enough to determine VSK's math.

If VSK is completely on her own, there's only really one way her damage could be calculated. We expect her in-skill FP to be 140+(0.2*83)=156.6, or 136-183 damage after variance.

Checks out.

Okay. Here's another question. Is this damage boosted by crits? We now have two possibilities: [140+(0.2*83)]*1.75≈274 [235-318], or 140*1.75+(0.2*83)≈261 [224-303]. With a VFL/cape/Five-Seven, we now see:

So her acc-to-damage buff is applied before the crit multiplier.

Next question. Does the same apply to damage buffs? Looking at her non-crits and adding a 5-star Damage II artillery (140 FP without aura, 217 with), there are three options: [217+(0.2*83)]*1.15=268.64 [231-311] (acc damage is increased by skills/talents, but not aura), 217*1.15+(0.2*83)=226.15 [229-309] (acc damage is not affected by FP skills/talents), or [140+(0.2*83)]*1.55*1.15=279.14 [240-324] (acc damage is affected by aura/skills/talents). The first two are uncomfortably close, so if we can rule out case 3 we'll switch to a different test.

Case 3 doesn't seem to work.

Her accuracy buff at the very least isn't affected by your fairy FP aura. This isn't ideal, as it means it's a smaller % buff the stronger your fairy is (think SVCh). Now the only question is if skills/talents/tiles can apply to the acc buff.

With the pictured tile setup (no fairy), we're looking at either [140+(0.2*83*2.3)]*1.88=334.98 [287-388] or 140*1.88+(0.2*83*2.3)=301.38 [259-349].

And the results:

We're seeing too many small numbers for FP tiles/skills to be applied to her acc increase. VSK's damage in-skill can therefore be roughly thought of as [(base FP*aura*FP skills/tiles/talent)+(0.2*acc*acc aura/tiles/skills)]*crit damage. These findings are consistent with a KR investigation. (Note: I'm lumping tiles/skills together as they are, based on my previous tests, calculated/recalculated in the same step all other skills/talents/effects are.)

Night

VSK is said to be less effective at night because her acc-to-damage buff uses her post-night penalty acc. For now, let's just verify this is the case. We'll talk about the impact of this later.

With the same tile setup as above, we have either 140*1.88+(0.2*83*2.3)=301.38 [259-349] or 140*1.88+(0.2*83*2.3*0.1)=267.02 [229-310] damage.

As claimed, the night penalty does reduce her acc-to-damage conversion. It's adding an incredible 3-4 damage here.

Skill Analysis^#

Before you go and scrap VSK due to her acc-based selfbuff not scaling with FP buffs, let's go over some example daytime numbers.

In a completely base state, VSK goes from 140 to 156.6 damage - a 12% increase.
With Calico/P22/Stechkin and no fairy, VSK goes from 204.4 to 239.26 damage - a 17% increase. With skills active, it drops to a 13% net increase.
Adding a 5-star Artillery fairy to that changes these conditions to a 11% and 8.5% increase, respectively

So - I would estimate it as a 5-20% damage increase. No matter what, it's nothing incredible. Acc tiles push it towards the upper end of that, FP buffs lower its effective increase. This doesn't mean it's bad to buff FP - it just means her acc-based buff is worth less consideration.

There's also the other component of her active - the first 3 attacks (at SL9+) each skill activation deal two instances of damage. (Technically, it's delayed by 5 frames, but that's splitting hairs.)

During the first few shots after activation, VSK has an effective 100% skill DPS multiplier, plus any accuracy conversion. This makes her quite strong in battles that last for 6-9 seconds. You could interpret this as VSK preferring FP buffs (to maximize each of the doubled attacks), but I don't necessarily think this outweighs the built-in preference rifles have for ROF. Rather than endlessly speculating and crunching numbers, let's run some tests and see how VSK actually does compared to other rifles - at night and at day.

Test Data^#

This is about as high as you can get for accuracy buffs without making any really dumb teambuilding decisions.

WA2000 was selected as a generic selfbuffer, who is likely to have an advantage here due to the lower amount of HG ROF buffers. VSK's performance looked solid at day, but seemed to drop a bit when switching to a night battle. Liu and WA2000 performed similarly to each other. Keep in mind that damage dealt statistics are difficult to compare between rifles.

Px4 Storm works nicely with VSK on paper - lots of acc from her tiles, and crit damage from her skill. Round 2, with SL8 as the common pair:

Note that SL8's tiles mean we expect most rifle's damage dealt to be higher than hers.

These zombies got a bit trickier at night - they have 22 EVA, and this echelon only had Px4's tiles for accuracy. VSK performed similarly to R93 at day, but fell behind at night. VSK's echelon even wiped a few times (those are the upper end of the damage taken chart, with 1920 total team HP). SVCh's surehit did not seem important, as she was somewhat worse than R93 either way.

While you'd expect SRS to get better and VSK to get worse at night, it doesn't seem like a big deal here. These Strelets have 36 EVA (Doppels 10), but VSK didn't lose much ground relative to Rico when switching to night. Echelon damage taken increased dramatically, but that's only because one doppel would get an autoattack grenade or two off against SRS/VSK - its guaranteed linksplash damage inflates damage taken, after only a slight slowdown in kill speed.

Conclusion^#

VSK-94 seems to trail a little behind the top selfbuffing rifles (e.g., Liu/R93/Rico/Type88), but her performance is solid. Despite her accuracy component being less effective at night, its impact is relatively minor and she can still do well - especially if you can find a use for her frontloaded/bursty double shots.

Is VSK the new meta rifle? Not at all. But she's a reasonable option, with some interesting quirks. If specifically targeting 6-9 second fights or otherwise benefitting from her doubled attacks, she can be quite powerful.

Minor notes^#

VSK has a unique firing sound. No impact on gameplay, of course, but it's neat.
As mentioned earlier, Px4 is nice to pair with VSK due to the accuracy bonus scaling with crit damage. It's sort of the same situation as with Liu, though - other options (P22) may still be preferable if you need what they offer (frontline shielding).
VSK's unique tile shape (covering 4 and 5 from 7) works quite nicely from position 7. When using P22, the other rifle only needs a single forward tile to have at least some CDR on all your HGs.

Example Echelons^#

Standard disclaimer: teams should be built as needed according to the circumstances of specific maps and enemies. Do not blindly copy-and-paste teams. With that in mind, here are some not-entirely-awful ideas to illustrate where VSK can be used.