Prime WS Information, Testing, Discussion

You can still use prime gkt for zanhasso builds but the tp gain is noticeably less vs having zero DA in tp set. But the gains with AM3 and slightly stronger base ws of mumei edges just by a small margin.

Small margin. Sure.

I don't save or store gameplay data.

Outside of simulations and only looking at gameplay parse at the end of the run.

Say in sheol C farm for example. Identical buffs, same party members.

You take a sample size of 10 runs with a r15 masa. And you take a sample size of 10 runs with a stage 5 gkt prime. The end parse will be a difference of a weaponskill or two in my experience.

But you're probably a better sam than me and I'm probably doing something wrong. /shrugs

I like it for everything except tp sets. Mobs have a tendency of running off during segments a lot more than they used to. It's great for everything else you're saying.

You mean the TP gain feels less.

7% chance of not having a 25% chance of a good TP round. It will affect a little less than 2% of your attack rounds, and the other 98% of the time, the TP gain will be the exact same.

Exactly its so little it doesn't matter. WS delay makes it even less of a factor and Mumei wrecks Fudo.

DA of stage 5 is at 10%, DA will proc ahead of anything else. I'm sure you know that. DA percent seems small on paper but when it procs that's already 90% that you won't get the good tp round.

I'm not analyzing this from what it's saying on paper but from experience.

Yes mumei hits harder than fudo. But if you take what we're talking about above, the fudos will give you straight light, producing larger overall SCs. And the speed of the run "feels" the same as well.

As I mentioned in a 30min sheol c run in exact conditions and a decent sample size, prime gkt stage 5 will edge the masa r15 by a ws or 2 in the end.

Considering the time it takes to take a prime to stage 5 vs making a masa r15, it's a small margin of gains to me. If you're considering making a stage 5 gkt, it's BIS. I'm not advising against it at any means.

But overall ignore what I'm sayin, equip it and go do a few runs.

Sheol C is irrelevant as content to compare weapons with. It literally does not matter. Parses lie because of overkilling.

I was answering a question and tried to apply it to content that everyone was familiar with. You are parsing yourself and the speed of the runs.

90% of not getting Zanshin assumes 100% Zanshin rate which is impossible (on attacks that hit). If you are wearing 100% Zanshin and St5 prime then your breakdown looks like this:

10% DA
22.5% Zanshin
65.5% Single hit

Without prime it's:

25% Zanshin
75% Single hit

You're losing 2.5% chance of getting a Zanshin hit, but gaining 10% chance of a double (that doesn't include Zanshin ATK and TP bonus). That you (and others) present this as being a disastrous, noticeable difference is laughable to me, but I'm not going to try to deny your lived experiences, you do you boo.

I said < 2% because I was thinking of St4 prime which more people are likely to have.
7% DA
23.25% Zanshin
69.75% Single hit

While I generally agree with your logic, I would like to point out, that Zanhasso rate is not 25%, but 35% (gifts add 10%) and then you can also use Ryuo feet +1 for double zanshin attack when it proc, so overall breakdown would be slightly different.

I coded this into my local version now, so it will be more accurate. I will recheck Gae Buide and post in DRG thread.

Maybe someone should update the BG page that says cap is 25% because it's 1/4 of the Zanshin rate you have in gear/traits. How does this work then, it just adds a flat 10% after all other calculations, after the quartering...?

I ignored the Ryuo feet because that's an even more minimal difference...you're talking about 10% DA reducing your Zanshin+Ryuo rate by what, .3%?

Without prime it's:

31.15% Zanshin Single
3.85% Zanshin Double
65% Single hit

Prime St5
10% DA
28.05% Zanshin Single (-3.1%)
3.45% Zanshin Double (-0.4%)
58.5% Single hit (-6.5%)

Prime St4
7% DA
29% Zanshin Single (-2.15%)
3.6% Zanshin Double (-0.25%)
60.4% Single hit (-4.6%)

Instead of 25% of Zanshin rate, with gits its 35% of Zanshin rate.



I know, but I'm a nerd

BG page is correct, Sam main just has a 10% on top of that from JP gifts which is not mentioned.

I wouldn't call something "correct" which lists a hard limit AND fails to mention another stat which goes over that limit, especially considering it's under the header "Hasso/Seigan and Zanshin-SAM Main Job Only". I mean...it's clearly a section talking about SAM Main, with Hasso up, swinging on hits that connected...and then says 25% is the cap, with no mention of the gifts. This seems like a pretty major oversight on the page (and by me, because I was sourcing my info from the page).

If you want a real head scratch here's one, does the 5% Zanshin from merits count towards Zanhasso or just missed hits? Up to cap or over?

@Zixxer. A few of my LS mates share your opinion. The DA is detremental in practise and very noticeable for them. I / they also agree that the dmg between mumei and fudo is pretty small if non existant thanks to aria.

I have the Gs and would also agree that the empyrean (Calad) does more dmg then the prime. This is probably only a drk and sam problem though.

All comments are based on stage 4 but I don't know that triiple dmg on the first hit is going surpass empyrean white damage and capped damage ws are capped damage.

I have Prime GS4, Prime GKT4, Calad15 and Masa15. Your LS mates are not correct. Mumei hits much harder than Fudo. (15% range)

Fimb and Torc are closer but the STR/VIT mod typically pulls Fimb ahead. I hit 99999 often with Fimb (effective 3000tp) and rarely with Torc.

Plus the flexibility since both weapons allow you to make light or darkness with their best WSs.

The DA being noticeable is placebo at best. Mumei + a DA is almost 1000tp depending on roll/saveTP. At worst you are WS in 2 attack arounds which with WS delay is not a loss.

The DA is helping you not whiff a WS which in the middle of a 4/5 step is a much larger DPS loss then anything mentioned.

ITT: people with selective memory and DRK who hit nothing but 99,999 Torcleavers every time they push their WS button.

It's even funnier, that DRK has (beside Hover Shot) potentially the most powerful damage enhancing JA in the game. If you use Scarlet Delirium optimally, you can indeed create a lot of selective memory 99999 WS moments XD

Did a 20 sample test just now on Locust Crawlers



Edit:
Control
Set a ws rule to ws once tp hits over 1k. Stayed at AM1 throughout testing for both gkt.
Buffs = cornelia haste, hasso, haste2, mad, march.
Tested on GKT5 not 4.

Lol, thanks for the support, I'm getting crucified here. XD

So I'm just going to state what should be obvious right now before this gets out of hand. We're in a thread about testing and data for primes. Doing a test with minimal buffs for a weapon line whose highlighting aftermath is PDL (except mages) is disengenous.

Please present data that would highlight information that people would actually want. If you continue down your current line of discussion at least have the decency to put disclaimers on your posts that you're sandbagging your tests to highlight situations where primes are not good so that less knowledgeable people who are looking for recommendations or data won't miss the point.

I know Zixxer personally and don't believe he would intentionally post misleading information in bad faith, as that's not the kind of person he is. I think he was simply eyeballing his experience and voiced/attempted to support it by backing it with a sample. With that being said, the test was flimsy and I agree with Iocus that it should be tested in the proper context so that a fair comparison can be made, at the very least attack capped with pdl+ gear and aria vs same targets using both weapons.

I happened to compete vs a St4 prime SAM as my R15 Masamune SAM in last month's ambu, and I honestly almost told myself that the weapon wasn't impressive due to what numbers I was seeing, but I had to remind myself that parses and eyeball tests are terrible ways to gauge individual weapon strength unless it's controlled test. It's very hard to make good assumptions just based off of what you are seeing, since there's too many variables to solve for that aren't necessarily noticed at the time.

I was answering this question.

"Bahamut.Drumskull said: »
But with the prime weapon u cant play with a zanshin build no? Because of the double attack on the gkt? Isint it just slightly better than r15 masa?"

As the small test sample dictated, DA affected the kill time and WS frequency.

I'm sorry, but calculating and simulating is much better solution here.

With 95% hit rate a 100% Zanhasso build with 3%QA and 0%DA AND Ryuo feet +1 will have 0.21% more avg TP per round over the same build, but with added 10%DA.
Without Ryuo feet +1 it's actually the set with 10%DA winning by 0.45%

So like you see Maletaru, those feets are actually important ;D

Ofc not really. Both sets with and without those feets are still almost equal, but I find it kinda funny, that this super small change they make is actually what pushes the advantage one way or another, even if the advantage is microscopic :P

Now lets make a full simulation of both builds with both weapons and both WSs.

Targeting 1250+TP and assuming Overwhelm is active.
KusanagiV+Mumei DPS: ~16240
Mumei median: ~66031
MasamuneR15+Fudo DPS: ~13800
Fudo median: ~55034

So while TP gain is almost equal in the Zanhasso build, DPS is not really even close. Kusanagi-no-Tsurugi V clearly smokes MasamuneR15. Ofc this DPS doesn't include skillchain damage.

Torcleaver allows you to do both light and dark. Also, I too hit capped more often with Fimbulvetr but I hit 90k plus with cald just as often (and capped dmg quite often) and calad has better white damage. I also hit harder with Torcleaver under 2k tp. I fail to see Fimbulvetr winning in overall damage at stage 4.

Also, you can claim mumei is king and the 7 to 10 da isn't noticeable. I don't have the weapon and can't confirm or deny but I'll take my static members word for it.