The Sealed Dagger: A Ninja Guide

Double posting to address the capped/guide methodology discussion:


I get that the varying levels of buffs and targets can be a real struggle to capture without having 50 gear sets...

But FWIW, I do agree that NIN in uncapped attack situations is not a very unlikely scenario in current endgame activity for most players. It's also one of the reasons Kikoku remains a useful tool, and I sometimes find myself using it!

As far as targets, I'd kinda think reasonable tiers of current endgame mobs might be stuff like:
* Trivial - I don't really care about optimizing sets in guides for this kind of stuff
* Lower-end endgame Omen fodder, easier Ambuscade, maybe up to Divergence normal mobs pre-midboss
* Higher-end endgame Divergence (post midboss), D~VD Intense Ambuscade, HELM, etc.

TP Bonus then MAB/WSD depending on the amount you can get. I had my CP cape locked so wasn't using a JSE back and the set was largely my RUN's lunge set with slight modification. Those three WS can be really really REALLY strong. Highest today was an 81K Chi but it normally hit around 40~50K unless a hit whiffed and then it was in the mid 20's. To was the next highest hitting 30~40K unless a hit whiffed, then Tekke was a bit weaker but still would see the occasional 48~55K appear.

Chi -> Tekki -> To -> Tekki is a four step Fragmentation that should kill just about anything, Apex bats were normally dead after the To.

These are like Jinpu, so Frailty + Maliase + Chaos Roll / ect.

Well describing an Attack Capped situation is very easy.
I think describing a no-buff situation is somewhat easy as well.
Just assume no buff at all against a high level target and literally no buff (capped Job Points).

So we got both ends of the scenario here, but the real problem is picking the one in between.
Like you said what target? Which buffs? Which values?

I reckon saying "somewhere in between the other two extremes" is very generic, but then again the sets list should be something indicative, a guideline, something approximate but still helpful.
So, if you ask me, whichever criteria you were to pick, it would still be very useful as long as the criteria remains the same and you describe it in the guide.

Anywhere around the middle would be useful. Just assuming a target where you got some att buffs but not too many, where you start getting an interesting amount of +att, but you're still far from capping it.
So really, anywhere around this description, no matter exactely where, would still be useful for a potential reader of the guide and, ultimately, for us all.


Again I'm just sharing opinions, far from me putting pressure on you Langly. In the end I think the guide is fine as it is and you did an awesome job out of it.

Absolutely agreed. It's the same thing I suggested to the guy who asked about it before.

But still one can mix & match pieces.
For instance what I do on most of my jobs (kinda depends on the job, the gear, etc) is the following:

1) Check which gear is BiS at capped attack
2) Check which gear is BiS at mid-attack
3) Take the set from 1) and mix it with a couple of pieces from 2) (how many and which depends on a slot-by-slot basis, depending on which pieces I have access to, which pieces produce the biggest increase and keeping the goal of trying to swap the least amount of slots possible).
4) After that I take this set and use it as my default set for that WS.

I also apply some Accuracy logic to my sets, adjusting where possible, when necessary.
I try to keep the accuracy level of my WS the same as my TP set.
If I have, say, ~1200 acc in my low-acc TP set, I try to make so my low-acc WS sets have around 1200 acc as well.
If my high-acc TP set has ~1400 acc, I try to make so my high-acc WS sets are as close as possible to 1400.
So, again, I kinda mix and match pieces according to these values.
Sometimes I might decide to pick a piece, for a certain slot, that has 2 less STR and 10 less att, just because it has 25 more acc and brings me closer to the overall acc value I'm aiming for.


I mean I dunno how everybody applies logic to the choices they make to their sets.
Personally for me it's a race trying for the best compromise.
I don't have the patience and the inventory space to create 2000 different TP sets and 3000 different sets for each different WS, so I kinda have to draw a line somewhere and try for the best compromise, the one that's gonna be useful for me in the highest possible number of different situations.
I really have no interest, most of the times, in creating and mantaining a set that I'm gonna use once every 6 months.

Thanks Saevel. I tried to have a conversation about these WS a few pages back but I don't think I had enough buff support. Gonna try a CP party with NIN NIN COR GEO BRD XYZ. I never knew about the 4-step hybrid chain either

You can actually go in circles with some of these WS's. They have really bad properties for T2 SC's but really good properties for linking several T1's.

Ei -> Chi -> Tekki -> To -> back to Ei

Or

Yu -> Tekki -> Chi -> To -> back to Yu

The two possible upgrades to T2's (with hybrid WS's) are

To -> Tekki = Fragmentation, pretty decent can follow with double Shun for Radiance.

Chi -> Yu = Distortion, medium depending on magic buffs / debuffs.

Chi is earth element, To is ice element, Tekki/Yu is water element. Use appropriate /SCH weather and obi swap.

Good stuff Saevel. I like To > Teki > shun > shun. If it survives, MB Raiton San should cover the apex. I have never really experimented with these with optimal buffs (frailty/malaise/chaos/fury), so it's good to see a use somewhere.

Chi is the better option since it's 2-hit. Is there any other chains that involve Chi besides the one above you mentioned with all 3 hybrids?

I'm guessing the best damage output would be the chi > teki > to > teki, and the one with radiance above. Very fun stuff.

(what are you using to calculate these skillchains? I'm using an outdated chart and its a nightmare to figure out)

My issue is, I have the gear with exception of sup3 +1 (waiting for it to come up on my server) but my ws's are lackluster on high levels. Think my tens were doing like 10k on apex crawlers and my sets are pretty well good. Haven't messed with the sets that saev posted but I am going to look into them. I just want to deal some damage!

Now question is, I have aeonic... do I finish kannagi or make relic?

In my opinion, stick with aeonic. Get extra support. Are you soloing apex? Or do you have a geo there?

Like most everyone says, and I agree with, Ninja needs a lot of attack. As well as attack, your fSTR is probably going to be negative also.

When supported you'll notice Ten doing much better. Without support, shun gains an advatnage. Hell, even Hi gains an advantage thanks to crits effect on pDIF.

Yeah, and to tack onto what Langly mentioned, you didn't really tell us what buffs you were getting? Also, apex crawlers use cocoon, so you can see better results from Raptors or even Bats. What TP are you spamming Ten at? Your gear looks fine, and su3 will barely push your ten damage any higher (absolutely zero su3 kendatsuba gear is needed for that set, anyways).

Saevel just posted a really neat hybrid skillchain setup that should demolish anything, support depending. I don't think your Ten damage is low, just unbuffed. Try Qultada+geo and if you can handle it, eat meat.

All hybrids are physical WS first, meaning off hand attacks and multihits apply. Chi is two main and one off hand for example. These hybrids have ridiculous TP scaling, aeonic and moonshade are huge.

What offhand do you recommend for hybrid spamming, capped Ochu?

Yeah, Ochu is going to probably be the best choice, maybe like a WSD kanaria but ochu having mab and all stats+ is hard to beat I think.

If you can swing the not insignificant accuracy drop-off, Malevolence might also be a contender. Capped Malevolence has a MAB+30 advantage over capped Ochu and lower delay (201 to 227) for more frequent WS.

INT is about the same as Ochu (10 versus 12), though Malevolence does lack the STR+22 that Ochu provides. FWIW, most of the hybrid WS are STR/INT mods: Teki/Chi are STR30%/INT30%, To/Ei are STR40%/INT40%. Yu is the weird one with DEX40%/INT40%

Kanaria can get weird stuff on DM campaign in particular, so might be somthing to toss at Oseem if you aren't after something else. I have a WSD+7% DM katana, and have seen some really high MAB ones as well.

It's not like it's ever a gamechanging DPS difference if you already have Heishi, so you probably don't NEED anything else. Even when one of Kikoku or Kannagi is better for the situation, Heishi is never far behind.

That being said...

Kannagi is more often the best choice for pure DPS spamming Ninja rage, if you can manage Innin and go nuts with Hi (see my sets/comments at the end of previous page). The worse your buffs, the bigger the gap in favor of Kannagi versus Heishi (which is really at its best with strong buffs, assuming Ten spam). I've been finding Kannagi as the ideal DPS weapon in most realistic situations I find myself in lately, but it's less flexible and really needs use of Innin/Hi to perform at its max.

Kikoku: its place is when you can take advantage of that huge Atk+10% AM (and Atk+60 on the weapon), so you will be maximizing its potential when you're far enough under atk cap. YMMV based on how you play NIN and how often you find yourself starved for attack - it could be useful often for you, or the huge attack advantages could be totally useless for you. Metsu is cool too, has really great SC properties and solid damage - so might be a reason to choose Kikoku when damage is otherwise similar. Really though, even when you find a particular situation where it beats Heishi, it's not by much, and more often Heishi will be a bit ahead - so I wouldn't stress it if you have Heishi already. Good option for people who don't have Aeonic though.

Also, if this hybrid WS discussion ends up helping push the meta toward those WS, that's clearly something where Heishi has an advantage due to the impact of TP Bonus on those WS.

If you ask me I wouldn't bother with Kannagi, that gil is better spent into buying something else.
I dunno whatever, plenty of HQ choices you could pick for NIN.
HQ Kendatsuba, HQ Ryuo, HQ Staunch, HQ Yetshila, HQ Moonlight, I dunno, lotsa stuff there.

Kikoku on the other hand, in addition of being cheaper than a Kannagi (I think?) is an excellent weapon if you find yourself soloing a lot.
When you're low on support or got no support at all, Kikoku is probably a better weapon than Heishi Shorinken (and it looks awesome!)

I'm kinda with you that a Heishi owner probably doesn't really need any other RMEA, since Heishi's performance is always up there at or at least not far behind the top.

However, you're definitely underrating Kannagi. Kannagi/Hi spam actually beats Heishi/Ten spam in a wide range of scenarios - more often than not, from what I can tell. Depending on what you do/how buffed you usually are on NIN, it might very well be the case that Kannagi is better for the vast majority of situations you find yourself in. Definitely YMMV though, I don't want to make assumptions that somebody doesn't always find their NIN with 2-4 COR rolls, Honor March BRD, and two Idris GEOs.

Caveats: Kannagi has less WS flexibility than Heishi (while Ten spam is good, Heishi can do a better job of mixing in Shun as needed for SC purposes). Kannagi also needs Innin to reach its maximum potential (though that tends not to be a big issue and even Trust tanks are generally fine - August holds hate off me no problem with Innin/Yain on ilevel enmity system content, even if I'm going nuts with Hi and a near perfect Kannagi TP/WS set)

Is Kannagi's not infrequent DPS edge big enough to warrant the expense if you already have Aeonic? Eh, I'd say probably not for most people. But it's better a good bit of the time.


This is usually not true. It will be a closer call, but even in situations most advantageous for Kikoku with minimal buffs, uncapped attack, etc. - Heishi still wins most of the time. When Kikoku does win, the advantage tends to be pretty small. Not really meaningful enough that I could suggest someone who already owns Heishi to make a Kikoku too.

Kikoku is still a very respectable weapon, with a nice relic WS and good SC flexibility, but it's sort of the cheaper RMEA option for people who can't get or prioritize Aeonic Katana. Kannagi is more expensive and a little more focused (Innin/Hi), but its ceiling is higher and tends to perform best in the kinds of situations I think NINs find themselves in a lot in modern party content.

Are you talking from in-game realistic parsedata results?
Or from theoretical scenarios calculated in the spreadsheet?
Because last time I tested Kannagi in the spreadsheet I was having a hard time creating fictitious situations to make it win over Heishi, and even then it was proving to be pretty hard even though in the end I succeeded.

In-game the only situation I can remember where Kannagi outperformed Heishi was against that Qutrub Ambuscade fight, and it was mostly due to the AM, not the WS spam/approach per sè.

Both.

Spreadsheet-wise, are you actually using good Blade Hi sets (like mine at the bottom of the previous page) and setting Empy AM3 at 50%? I'd also suggest setting Innin on to allow Kannagi to take advantage of that.

I have a spreadsheet that's a bit customized, but try something like starting from:

Set 1: Heishi/Ten
Set 2: Kannagi/Hi (be sure OTD rate is 50%)
- Sublime Sushi, Innin, Kakka on
- No other JA on to start (Sange, Berserk/Aggressor, etc)
- Haste 1, Dia 2, Boost DEX
- I put BRD Marches on as default mainly just to simulate capped magical haste from whatever source
- Target: I usually start from Apex Bats (136) or Bhoots (139). Pick one of those?

At this starting point, Kannagi/Hi should be well ahead. Now start adding some buffs - it should take pretty substantial buffs to get Heishi to close in and overtake Kannagi, but eventually with enough buffs it probably will.

Also, be sure to double check GEO buff values on your sheet are accurate - for instance, I've seen some past version NIN (and other job) sheets with way overinflated Frailty values that result in inaccurate modeling. Depending on whether you have the luxury of always assuming Idris GEO availability, might even be a worthwhile practical check to input values for a 900 skill Dunna GEO instead of perma-Idris.

Gearsets:
I used the same TP set for both sets. Maybe some people have minor quibbles, but this is a pretty near max TP set for either weapon.
ItemSet 358992

Blade: Hi set from previous page (make sure your spreadsheet correctly accounts for Mummu+2 set bonus).

Blade: Ten set mostly like the one on page 1 of this guide, except I used Adhemar +1 body/hands (which may spreadsheet better than Herc with STR+10/Acc&Atk+20/WSD+4~5, they do for me in most cases), and I tend to prefer not to assume Lugra latent (so I use Ishvara for WSD-favoring WS).



IDK, I do find Kannagi pretty effective in practice. I've outperformed other Heishi users on same content (but that's not necessarily indicative of weapon potential), and my Kannagi usually crushes my Kikoku in the vast majority of content (I usually pull out the Kikoku more for SC purposes than anything else). But I don't have all 3 myself: Aeonic is a pain for our non-SMN LS, and katana isn't in my top 3 priority Aeonic weapons anyway.

Also, the AM for Kannagi is ALWAYS what makes it perform well. Kannagi/Hi will have lower WS damage but significantly higher white damage - often enough to make up for the WS damage deficiency. The more you buff the WS numbers, the better Heishi does. To me, NIN is still one of the jobs that aren't quite as heavily skewed toward WS as some others though... which is precisely what makes a weapon that really maximizes the white damage perform pretty well.

Remember too that in reality, you might not WS at the ideal time all the time. Amnesia, waiting to WS in order to SC (or not interrupt another SC), just slow WS due to human error... All of that plays more in Kannagi's favor than Heishi's, since Heishi's main advantage is in stronger WS (so the more WS over time, the better Heishi's overall numbers). For Kannagi, some delay in WS isn't quite as big of a problem because you're still hitting all of those large damage mainhand crits (900~1100 a pop normal attacks sure add up on a job that attacks as fast as NIN).

Kannagi also has the issue of absolutely relying on AM3, but it's not that hard to get it up pretty quickly with NIN's attack speed and use of tools like Sange for a quick TP boost. Once it's up, it's kind of just an art of timing it to maintain reapplication of the AM effect - same as any mythic/empy. But, admittedly, some stuff like encumbrance or TP wiping moves can really play havoc on an AM-reliant weapon.

I actually got pretty interested in messing with Kannagi and Empys after doing a lot of ranged stuff with Armageddon and Gandiva RNG taking advantage of strong crit AM3 sets and performing REALLY well in practice. I've been nothing but pleased with the Kannagi so far.

Oh hey look, mho was right about this months ago while you guys were all crazy for Heishi. Maybe mho knows some things, Capu-chan~ Finish, your Nagi and learn to tank with it :]

How can they be both if you say you have still to complete your Heishi D:

But regardless, let's keep talkin about spreadsheet and hypothethical situations.


I decided to give it a go and use your sets (which, while not BiS, are definitely superior to mine!).

Here are my results. Kannagi/Hi on the left, Heishi/Ten on the right, 2 different scenarios, overall dps.

1) 1605 / 1265
2) 3634 / 3865

First is your scenario, which is basically completely devoid of any attack buff.
Second is what I'd call a pretty common/average scenario. Just Honor March, Minuet 5 and Frailty. No Marcato, No Berserk, No food, no BoG, no EA, no Idris, no Bolster.
Heishi/Ten is already above :x
Granted the difference is small, whereas the difference in scenario 1 is bigger.

There are some things to take into consideration though.
First: Scenario 1 greatly favors Blade: Hi, it's basically the ideal situation for Hi and how often are you so devoid of buffs unless you're soloing with trusts? In this scenario even Heishi/Hi is better than Heishi/Ten (still inferior to Kannagi/Hi though)
Second: This is not very realistic. The Spreadsheet doesn't take into consideration the fact that Innin won't be active 100% of the time, this only affects melee damage for Heishi, but both Melee AND WS damage for Kannagi.
Furthermore you told me to consider 50% OTD, which I did, I'm not sure that offers a realistic comparison with Heishi though, because the spreadsheet doesn't take into account that to activate 50% OTD you need to get 3k TP, that means "wasting" 2x WSs and that reduces your overall damage over time, every 180 seconds. This is not calculated by the spreadsheet but it would produce lower numbers for Kannagi/Hi in reality.
A more realistic assumption would be 30% OTD imo. The numbers I posted would be balanced pretty much the same, with Kannagi/Hi winning in Scenario 1 and Heishi/Ten winning in Scenario 2, but the difference would be smaller for first scenario and bigger for the second.
Last: in a real situation you're going to unavoidably end up with overTP while WSing. The best way you can simulate this without affecting TProunds/waits I think it's to add a custom TP bonus in the WS sets. This won't affect much Hi damage (doesn't really scale that well with TP =/) but it's gonna be a big bonus for Ten.
Once more, I think the numbers above would be spread pretty much the same way, but smaller difference in scenario 1, bigger in scenario 2.


In the end I had no claim to say that Kannagi is a bad weapon, it's awesome! But that it needs pretty specific situations to outdamage Heishi Shorinken. I don't think this statement of mine got proved wrong?
Yes, situationally Kannagi will outperform Heishi, but on average I think Heishi will produce better results.
As such I wouldn't really suggest someone who already owns Heishi Shorinken to pursue Kannagi, unless he's well aware of what we just said.


If I may add one last consideration, over the last year+ we saw the addition of a lot of interesting and powerful Crit gear, like Mummu for instance, and more. This provided a great benefit to Blade: Hi, whereas the situation for Blade: Ten is not really that much different.
Tl;dr: using Kannagi and Hi with nowadays' gear, will surely produce better results than 1+ years ago, whereas the damage on Ten didn't increase as much.
i.e. : the damage difference has certainly evened out a fair bit, in favor of Kannagi/Hi.

Sorry I did not explain better. Besides the new relic pants and sup3 I have basically the exact same sets as the guide. tp sets are fine I have ryo head +1 and adhemar +1, just do need to get a nice set of high acc legs without dw in my high acc high haste set.

I was looking at guide for the sup3 that its in and alot of other sets posted here have it but like I said I am kinda low on gil with only 188mil (sounds like alot but that set is 170mil on odin).

Some pieces could use better augments on herculean side but that entire system hates my guts.

I mainly solo with nin or blu but I do dyna D with ninja. I do not have a geo or cor. I usually lead the melee party as tank/dd and we usually get a bard. (we break into 2 groups for farming one mage heavy for blue eyes and one melee heavy with a couple blms for statues that kill green eyes).

Attacked starved..very much so.

Well thanks for explaining. Looking at your gear and your posted damage, it might appear a little low for a NIN, but not really that bad. My Ten damage unbuffed can range from 9k~13k or so if used around 1k tp. So i'm right in the same range as you, again, unbuffed. That's normal, and I have the 1200 gift and the +3 relic pants (but overall worse herc augments than you). As others have said, getting su3 and the pants wont magically skyrocket your damage. If you can't change your ws damage, change your buffs. Or change your ws. I dont think you're doing anything wrong.

If there's a GEO using frailty, you should expect that damage to climb a bit higher.

I've had the same experience as Sechs.

Apex Bat
-75% Def Down
-capped magical haste
-cRatio for main and offhand is capped, as is throwing
-Accuracy is capped, RAcc is capped
-Over-TP rounds set to .25

Using both sets for Heishi/Ten with Aggressor/Berserk/Sange on in set 1 and off in set 2. I kept the set ratios to 180 to 120 for berserk and aggressor though with Sange I would probably want to adjust that a bit. It won't mess this up much.

Combined DPS comes in at 5508.455 (5664.9/5273.7)

Using your sets for Kannagi/Hi with the exact same settings for the target and ratios + your gearsets on the last page.

Combined DPS was 5279.307 (5443.648/5032.795)

Just about everywhere I tried to get Kannagi to win, I couldn't.

Go make it, use it, and parse the two. Maybe then you'll understand it's practicality.

Comparing Kannagi/Hi to Heishi/Ten for low ratio situations is pretty irrelevant 90% of the time

Oh cool, more non-Kannagi's chiming in their thoughts about weapons they don't have or use.

Innin will not be applicable 50% of the time if you are even a mediocre DD, stop assuming it's benefits all the time. When you deal big damage you will pull hate, when you pull hate Innin no longer applies. The more badass a DD is the less benefit Innin gives and Innin is basically required to make Blade: Hi preform well, a +30% crit rate is huge on a WS that deals all it's damage in one hit that must critical.

When we discuss dealing damage we must first determine which context we're talking about and how it relates to the discussion. In FFXI there are two large categories things fall into which, first is when damage doesn't really matter because it's on a lower target, with few buffs and your screwing around with friends or soloing something. Second is where damage does matter because your fighting a big bad *** and loaded with buffs. In the first scenario dealing less damage doesn't materially effect your fight, in the second it can result in a loss and several pissed of LS mates.

It's pretty clear that Emp + Blade: Hi wins in the first because lower HP mobs, trust tanks and well under attack cap favor Hi. The second is where everything else starts winning because your now rocking plenty of buffs hitting hard as hell and having swings taken at you.

General Purpose DD Party Load Out

DD x2 or Tank + DD
WHM Healer
GEO Frailty + Fury, 900 Skill Dunna minimum, Idris on high level stuff
COR doing Samurai's Roll + Chaos Roll
BRD doing Macrato Honor March + Victory March + Minute + Madrigal

That is the context in which we discuss important damage. Nobody cares about the awesome WS done on a CL124 monster while solo, everyone cares about the damage done on a CL135~150 boss battle.

For low ratio, it should be compared to Heishi/Shun outside of instances where you want to form Darkness chains or want to avoid Light. Kikoku is also a more relevant comparison than using Ten when it's obviously at a disadvantage.

What is this, a "no u" argument in my beloved ninja forum?

It's not 2005, we understand the mechanics of the game. I don't have to own the weapon to understand it's "wtfbbq" awesome.

Please talk numbers, playstyle, examples, etc. Otherwise you're not helping your case. I'm happy to go back and forth with Capuchin because he doesn't just stroke his kannagi and say 'ur wrong.'

When I look at damage potential, I look at the peak. That means capped cRatio and so forth. Because Ninja's damage capacity exists underneath that curve. What wins where, is muddled a bit and instead of changing gear we should do what Kyte says and start using different WS's, skillchains permitting.

Having an opinion is great, but a "parse it" argument didn't fly in 2007 and it's not going to fly now.