The Sealed Dagger: A Ninja Guide

relc head has Elemental Ninjutsu III (San spell) damage +5% per merit level. This is separate from Magic Attack Bonus. You sure your winning lol?

I have a herc head with 28 mab and 5 burst dmg, never tested, but if that can win, yippy!

Ah... That would be it, I just have 1/5 in each San spell. So I guess for your one or two 4~5/5 San spells, Relic hat is indeed best.

Still not convinced elemental ninjutsu is particularly worthwhile anyway though! ;)

ya at 1/5 it wont be big, i suggest doing 5/5 on 2 diff.

while i still dont got a good nuke setup, its worth it in apex at least if your nukers cant down a mob in 1 volley.... or solo.

119 relic head should beat anything else with 5/5 in one san spell(+25% dmg Seperate from MAB) if you don't have one maxed out then yea herc is likly better.

while I don't have any SSs for you, in CP pts when I nuke with a geo I get aprox 30k Hyuton:San 20k suiton: Ni and with capped casting times JP it's pretty much instanuke on NI, 50k total nukes from nin is on par with average blm in cp so considering I've done pts where I solo darkness SC then drop 50k nukes I can say I was the main part of that pt instead of just making SC and waiting to make SC again.

Do Mujin Mantle and Shigi bypass the 50% recast time reduction cap?

I'd be curious to read more tests about these myself.
Also when does the -delay thing get applied? Before FC/Haste or after?


On a side note, impressions on the Hizamaru+1 set? It's pretty meh, but if your courrent +acc augment on Herculean suck, it can be an interesting option for heavy acc sets? At least for some slots, not talking about going 5/5.
I guess this is a valid option only for average players who lack good options for all possible situations, not for the very dedicated NIN players.

Last but not least, anybody tested legs on the spreadsheet?
Wanted to know how they compare to other typical WS options for various weaponskills.
Before Hizamaru I used Samnuha (if I recall?) for most WS and Herculean (acc/att+25, agi+5, Crit rate+3%) for Blade Hi.
I haven't tested but I feel like Hizamaru likely wins for Blade: Hi and probably for Ten as well, probably for all NIN WS to be fair, with the exception of Shun where Samnuha might still hold because of the Dex+12 I suppose.

Did some very small/limited testing myself on the WS part. Totally non conclusive tests but so far I'm getting these results.

Blade: Hi
Herc posted above > Hizamaru (this is probably because of the very high Agi on Herc)

Shun
Samnuha > Hizamaru

Ten
Hizamaru > Samnuha and Herc, good margin (around 200 damage difference)

Kamu
Hizamaru > Samnuha (small difference)



Talking about perf Samnuha of course. Alas I still have a -1 (close to perfect, but not there)
These tests do not take into consideration the fact that Hizamaru has much higher acc than the other options I tested.
In a situation where you need every single point of acc you can get, Hizamaru will likely win.

Anybody mind doing more accurate tests than my very superficial ones?

I was pretty let down with the set. I had hoped for big things like the last set did for those jobs. I use the body in my idle set and tbh the regen is really cool IMO, but all that work for regen? ya....

the head I hear is very good for mnk, the legs are a good ten piece and thats its from my understanding. The hands are useless, and the feet do have a spot in my DW set now, which odly enough gets used to much in party's (lazy mages), but in themselves are nothing to brag about and outdone.

overall I was/am still upset since all these jobs needed a boost a lot bigger than what was given. Good for returning players though, but lets be honest.... in the 20-30 hrs of ambuscade bs even a new player could farm 30mil or more (90 if smart). You could merc all adhemar and buy all abjs with that, or farm all herc and have all pieces with 4TA.

It should have been strait -dt set like last months, or should have really catered to mnk. Tons of MA so they can at least cap delay reduction and some obscene amount of DA/str too. Im talking +65 str or more and like 20-30 MA a piece, then mnk could have maybe came out from way last to, meh. and sam needs higher ws numbers...

Overall I am just sad with the lack of real quality from SE on this. Lucky my direct3d error is preventing me from playing atm so i can write this wall!

"very good" is an hyperbole.
In uncapped attack delay situation it can be a decent option for MNK and PUP, supposing the 6% Haste will still put you at capped gear haste, but that's it.
Perf Dampening Tam, Adhemar Head and good aug (TA) Herculean are probably gonna be better.
Like said before it has lots of acc so if you lack good acc options it's good for that at least (same can be said for all of the hizamaru set)

Hizamaru set is nothing spectacular but I think people are exagerationg in saying it's complete ***.


Yeah from my short tests it proved to be really hot for Ten.
I guess it could be decent for Hi too if you lack good augments on herculean.


I'm currently using path D Rawhide in my DW sets. Need to recalculate values, maybe I can swap a few things out with Hizamaru, but it has to be seen wether or not it will be worth it.
Except the already mentioned acc-need situations of course.

Was under the false assumption that there is still a maximum recast reduction. However, more details as well as exact formulars are given here:
https://www.bg-wiki.com/bg/Spell_Recast

If its true that there is no cap the minimum Utsu San recast would be (caped Haste + FC + Mujin Mantle and Shigi):

((100 sec * (1-80%)) * (1-40%)) - 1 - 3 = 3.2 sec

But I guess testing is needed to verify whether that really works.

There is a recast cap, it's 80%. Just like casting time.
This is why stun doesn't drop below 9s.

Though I think there are somethings that can overcap recast.
But that's mostly 2hr/JA stuff.
However -Xs could likely break the recast cap (which is at least haste+FC+recast%).
It's a weird stat that is on only a handful of things.

It wouldn't break the cap if it's calculated as a reduction before the % coming from haste/FC and if the haste/FC reduction still use the original (before the -delay) duration to calculate the 80% cap.

In such a scenario -delay would be useless if you're already at 80%.
Which is why I was curious to see if anybody ever tested if the -delay gets calculated before the fC reduction, after, if it works outside of the cap or not etc.

There needs to be some testing, and I think it has to be item-specific because this is such a rare stat I wouldn't be surprised to see different behaviours from the same stat on different items.

More or less agree with this analysis.

Head is indeed pretty solid for H2H users who aren't at capped delay reduction (i.e. not getting capped magical haste). It's actually comparatively more useful for PUP than MNK, since MNK has access to Adhemar/Dampening and PUP doesn't (Herc and Taeon only). See my post in PUP forum if you're interested in more detail of various MA/delay capped scenarios.

Feet... yeah, it's SAD that I need to use a DW set, but I do indeed need it sometimes these days. It is a pretty healthy chunk of DW on one slot though!

FWIW, NIN needs:
* DW+39 to hit capped delay reduction with ZERO magical/JA haste.
* DW+32 to hit capped delay reduction with Haste 1 only.
* DW+21~22 with Haste 2 only, or Haste 1 + Mighty Guard
* DW+1 (or, don't bother with DW gear) with capped magical haste

Hiza+1 feet fit nicely into those kinds of builds. My DW+39 set (which I DID have to use in 6/6 ambuscade party the other night...) uses:
ItemSet 344009

-I've added in the new Hizamaru Gear +1 into the spreadsheet as well as some WS Dmg options for Herculean on both the Dex and Agi side of the camp.

-Also added the Heishi Shorinken, something I didn't think I'd need to bother with but since obtaining it, decided to look at it's performance math-wise.

(My god Blade: Ten with Heishi and Moonshade is stupid good.)

-Corrected an error referencing a gear column in the data tab.

Found Here: DPS Calculator - Ninja.xlsx

The front page of this is horribly out of date and I now see that I am an allowed collaborator... so it's about time I look into what I can do for the front page. ***looks freaky... O_o

(I have not yet gotten to adjusting the geo bubble toggles for both sets yet)

Awesome, Langly - thanks for your work!

Questions:

1) Are we sure Shun is reflected right in the spreadsheet? It seems crazy low.

I did a lot of VD Ambuscade on NIN in the past week (+1 Regen body get, woo!) - Kikoku, relic AM, generally with GEO x2 or GEO x1/COR x1. Did enough to tweak sets and use it as sort of a calibration tool, played with some variations across different runs and comparing myself to my same DD partner in crime as a pretty reliable baseline (a good Tizona BLU, and we were generally very close in DPS throughout our ~15k+ Ambuscade points together). I was finding Shun to perform REALLY well there versus Metsu/Hi/Ten. Admittedly, my Ten set might be a little less than flawless (but still respectable), and my Metsu/Shun sets are pretty well on point... but it seems in real life to perform way better than it shows on spreadsheets.

Part of this might be YMMV based on weapon, as Shun should perform better both with Heishi (associated WS) and Kikoku (that Atk+ AM is very useful for Shun), but even for other mainhands I'm a little suspicious here that something is off.

2) On the current spreadsheet, it's looking like Kikoku beats Kannagi in... most situations I've looked at (in a few minutes of quick playing around - put a solid Hi set together, made sure Kannagi was reflecting OTD, etc.) Is that your experience too? What changed from older sheets where Kikoku won sometimes, but Kannagi more often?

3) I'm TOTALLY not complaining about your fantastic work for the community! However... can we please at least turn Ionis off as a default buff on the spreadsheet ;)

I've commented before that people playing with the sheets need to remember the defaults are often buffed to the moon to the point where it might not be realistic for your actual situation. Maybe you actually are getting Haste II + BRD + COR + GEO when you play NIN, so hey, you do you. But Ionis is very clearly not a buff most people will have on any content of consequence in 2016. If you're trying to optimize for Incursion or Vagary, turn it on yourself ;)

I would be entirely unsurprised if Shun's attack mods are currently incorrect. Not out of any anecdotal evidence, but simply because nailing down attack mods has always been a bit of a pain. Might have also changed at some point. It's a little bit easier to get conclusive values nowadays, so further testing is certainly worth consideration.

Shun had broken TP mod a couple of months ago, but I reported it and Langly fixed it supposedly.
Before that, Shun wasn't calculating any TP above 1000.

Make sure that old error didn't surface again on the spreadsheet. I'll check it myself later.

Looks like Blade: Ten is messed up to me (2.5/11.5/15.5 in the spreadsheet) although I have not used these too much.

I'm always happy to have multiple people checking the work.

@Capuchin, I have found Kikoku 119III to win in most instances against Kannagi 119III. And by most instances, I'm referring to iLvl 135 and up. Due to some unknowns revolving around mob stats, it's probably fair to say that all three katanas are very well balanced, each with their own pros and cons. Though I will admit to putting my kikoku aside and having hellacious fun with this Heishi Shorinken. Do you use scoreboard for your parsing? If you do, what kind of wsavg do you see with your Shun on the Tauri? We don't really know their stats so it's going to be a bit of a crapshoot, if you're using my generic mobs on the spreadsheet, know that they may have inflated stats outside of Evasion.

Shun had gone through a lot of changes recently, Sechs had pointed out that it was not carrying fTP which was corrected as well as the Attack Bonus not finding its way into the WS set columns. Based on gdiShun's data I linked previously, I have given Shun the following attributes (again, which need to be conclusively nailed down):

Blade: Shun anchor points
fTP 1000 : -8%Attack
fTP 2000 : +25%Attack
fTP 3000 : +100%Attack

This could be a fairy tale or it could be close to reality, I'm banking on it being wrong. I really do not know for certain. In the spreadsheet it is reflected as a compounding value, so you would see -8 > 33 > 75.

@Snaps, even without a great Ten set, that WS has been blowing ***up for me with Heishi. The fTP mods for Blade: Ten were modified as of December 2014? Tested and linked on Blue-Garter wiki.

Makes me wish Ninja had a force crit JA... could feel like Rudra for ninja!

I do /thf on WAR from time to time. Not because it's effective, but because it's fun. But I have Berserk and Warcry, so there is that. Maybe with a solid Geo doing Frailty+Fury and a Chaos Roll you can cap pdif.

I would really like any input on setting up a test case for trying to get Shun's attack modifier figured out.

Based on gdiShun's testing here. I would like to go back to those same Tarichuk's and wail away for another batch of samples.

If we knew the mob stats exactly, then this should be elementary. But we don't, so I kind of need to start there. Tarichuk's are warriors and as far as I remember, mob stats (non-NM) follow fairly similar to player character stats, at least up until level 99.

If anyone knows of any conclusive way to get mob stats pinned down (do I really gotta get on PUP and go shoot at them with Sharpshot frame?). I'd love to hear it. Once I know those, this should be relatively straightforward.

It's FOUR katanas!!!one!1! /slap
(lol)

Do you guys use Isvara Earring for Ten? Can you guys post two Ten sets? One for low acc and one for high acc? I'm curious about what you're using.


Ten is cool at 2000+ TP but it's nowhere as good as Rudra. Katana WS are all in all pretty meh. They don't "suck", but they sure don't compare to other weapon categories. With better WSs NIN overall damage would be so much more competitive.
Ten with Heishi is probably the best we can get out of NIN atm.

MNK is kinda in the same situation, but it's even worse. MNK's "Ten" is Raging Fists.
Kinda ironic when you think about it lol, to think how broken Vsmite was before the WS reform.
Thing is Vsmite stayed kinda the same, while everything else skyrocketed over9000 lol



For the Shun test make sure to have Battlemod loaded up so you can check the monster's level before engaging, and separate the results you get in different columns according to the monster's level.
The range is usually 3, max 4 levels afaik. Maybe even less on Apex, I dunno.

Thanks for the note about battlemod Sechs.

Blade: Ten is actually very similar to Rudra's in regards to fTP scaling. At 2000 TP Ten sits at ~11.5 and Rudra's has ~10.2. We just don't have a JA that ups the base dmg and adds a crit to it (SA or TA).

Looks like a lot of STR gear and WSD + is the way to go for ten. This is what I've drummed up:

ItemSet 344074

As noted on the set~
Herculean gear will best perform as STR/Acc/Attk/WSD.
Andartia as STR/Acc/Attk/WSD
And of course Kanaria as DEX/Acc/Attk/TA

Some swaps to make if your targets are quite evasive without losing too much punch from Ten would be:

ItemSet 344075

Or if your herculean gear is DEX acc/attk based you'll gain a little bit from that at the trade off of some potential damage.

To steal from Falkirk, it's really starting to feel more like Blade: Ten out of Ten. :D

The only stats you'd need to identify for a Shun test are defense (easy) and maybe VIT (slightly more annoying, but not out of the question). No reason to bother with NMs here, so you can use /check to figure out defense. You could use throwing (bring a chakram) with capped ranged pDIF to establish mob VIT as well if needed, or just stack a ton of STR if you're confident you can cap fSTR. With that in mind, most any lowish level Adoulin fodder is probably fine. Maybe even something older depending on your attack.

From there you can generate data sets at fixed ratio and TP values and work backwards to find the penalty. Ideally you want to stay above ratio=1.5, as there's a region just below there where the lower limit is fixed at 1.0. That way we can stick to average values, though you can work around that if needed given a sufficient sample size (~50 WS is probably enough if you're ensuring only one hit lands).

If you want to be even more precise then you can abuse the new pDIF caps for a one-sided test, similar to crit rate testing in Abyssea. If you can figure out precisely when your lower limit drops below pDIF cap, that'll net the exact modifier value.

Ten's mods suck compared to Rudra's (30 DEX 30 STR vs 80 DEX) but the base damage on Heishi can make up for it somewhat. I think it should be revised to 40/40 to be competitive with other melee (as well as several other ninja adjustments.)

I'm seeing an agreement with my wsavg from Scoreboard with what I see on the sheets for Shun falling in line (At content level 135).

I will still see if I can devote some time today/this weekend to figuring out Shun's behavior. Be sure to pull down a fresh copy of the spreadsheet as I'm adding new gear to it continually (as I see things that improve Blade: Ten dmg lol).


Edit: By the way, what is the content level of VD Ambuscade? I never paid attention if the game told me while I was farming my Regen body. :(

Similar but inferior, and Rudra has much better mods =/ Which further scale well with SA but that's another story.


Opinions on Ishvara vs Tati vs Telos earring in place of Lugra+1?
Also I guess Suzaku ring (and Apate?) can be cheap-ish alternatives to Ifrit+1 if you don't have one.

Would be cool to have other options in place of Herculean with TEN specific augments. Personally I already have so many Reisenjima pieces for so many jobs for so many purpose I really don't have the will, gil and inventory space to make yet another set just for Blade Ten.
Don't follow my example I know I suck! Just explaining my reasons.

Similar situation for cape. I have to get so many capes for so many jobs and with the 3 per month I need to draw a line somewhere.
Originally I wanted to get a cape for TP (DEX/acc/att/DA) and one for Blade Hi (AGi/acc/att/WSdmg) which is a HUGE boost for Blade Hi. In this scenario, the first cape wouldn't be too bad for Ten too. Not optimal, but probably better than any other option.
But in light of having Heishi Shorinken I don't see me spamming Hi anymore, unless for SC purposes, so I might change my plans and make a STR/acc/att/wsdmg one I guess, dunno, so many options :(

When I checked the difference between my DEX/Acc/Attk/DA cape that I use for TP on Blade: Ten to a STR and WSDM one, the damage was heavy. 200+ on Lv135 or above. For Ten, Ishvara would be your better option. Both are inferior to Lugra+1 when it's active. Though for the sake of keeping inventory sorted, Telos isn't terrible for Ten. Works for both your tp phase and your ws! :)

All these gear options sure are a pita. So far with Herculean I have:

DT-4 set (head, legs, feet) @ 3
MAB/Macc set for COR and NIN (head, hands, legs, feet) @ 4
TA+4 Acc set for every job on it (head, body, hands, legs, feet) @ 5
Refresh Pieces from Dark Matter campaign (hands) @ 1
TH pieces from Dark Matter Campaign (body, legs) @ 2

And I won't get into valorous or chironic/merlinic.

One really does begin to wonder if it's 'worth it' considering melee, and NIN at the very least, is hardly used for content where this kind of min/maxing is even relevant. :(

If you also play THF, augmented DEX/WSD gear is useful for both Rudra's and Ten. So there is some overlap there. I think best melee set for any job that can use it involves some Adhemar HQ, so there is overlap there.

Also whether min/maxing is relevant is fairly objective.