The Sealed Dagger: A Ninja Guide

I'm elbow deep in Excel file atm adding in GEO buffs and the relic business to my version here. Also going to be adding in some gear.

Hope I can upload it before lunch time for peer review~ that .ODS file you have Ores is just weird as ***to me :D

I cannot open it in any program I have that it doesn't lose some equations or anything.

Adds radiance, not light.

I mentioned this earlier, but I tested Metsu with Kikoku. It's MUCH weaker than Kikoku using blade Hi.

Shun is another story, as It's been updated since the spreadsheet was created, and we don't know what SE did to it. If I were to compare Shun using the dps sheet, it wouldn't be very realistic. It is quite good now days though.


No, I'm not sure I'm doing it correctly, because there's limited data on this. I went back and checked, and I was only calculating the attack bonus for the main hand. Good call. I've added the attack bonus to Hand #2. I was also not considering it for WS, but that's fixed now as well.

What about Dakken though? Can you check sometime and let me know if it's also affecting your ranged attack?

Hand 1 TP
Base Att 1097
Food Att 150
Berserk 0
Chaos 0
Geo-Fury 0
Total 1247
Total with AM 1356

Hand 2 TP
Base Att 989
Food Att 150
Berserk 0
Chaos 0
Geo-Fury 0
Total 1139
Total with AM 1237

Yes, they are. Keep in mind Empy's Occ attacks X times can proc on any swing from the main hand. Technically, you could proc OTD on all 4 hits from a Quad attack for example. With Relic's OTD, you can only proc on the FIRST swing from Hand #1. This makes the already low proc rate for Relic even lower by comparison sake.

This was a bit jacked up. I had mistakenly assigned Kikoku Quadruple damage instead of Triple damage, and Mote had set the proc rate for relic to 12%. I set it back to x3 dmg, and adjusted the proc rate to 16%. (are we sure about this?? It seems higher than other relics)

I wouldn't be so sure Kannagi gets anymore benefit from sam's roll than Kikoku does. Remember Kannagi's largest benefit is it's enormous boost to white damage from regular melee swings. Of coarse +50 AGI helps Hi's damage, but it's not a huge difference. I'm sure Kikoku's attack helps a bit for Shun in the same regard, and Nagi's ability to proc DA/TA on WS helps it.

Uncapped attack is uncapped attack. Try and remember this is a simulation. Our goal is uncapped attack vs. capped attack. I'm not putting Risenjima NM's into this because not a single one of these Katana's gives any benefit to accuracy, and that's not what we're comparing here. Tojil has high as hell Defense, but I'm sure there's some Risen NM with even higher defense. I'll bump up his stats to simulate something tougher like 135 gramk droog or Old Shuck, but not going to touch his AGI/Evasion.


Agreed here.

Anyways, here's some more results, with the above changes mentioned in effect. (am issues and otd issues fixed, as well as no more corsair rolls)
I'll provide more info to add clarity, but I'm sure you'll like the results now.

Kikoku AM - 1063 DPS. (4776 avg WS DMG, and 836.7 avg melee dmg per round)
Kikoku no AM - 1015 DPS. (4573 avg WS DMG, and 776.8 avg melee dmg per round)

Kannagi AM1 - 1164 DPS. (4948 avg WS DMG, and 975.3 avg melee dmg per round)
Kannagi no AM - 1008 DPS. (4948 avg WS DMG, and 706.3 avg melee dmg per round)

Nagi AM3 - 1028 DPS. (4681 avg WS DMG, and 776.2 avg melee dmg per round)
Nagi no AM - 909 DPS. (4395 avg WS DMG, and 681.8 avg melee dmg per round)

I also stacked STP in the Nagi AM3 gearset to try and mimic reality a bit more, but it's still disappointing.

One last comparison. I bumped Tojil's defense from 1900 to 2500, and his VIT from 200 to 250. (approx 25% increase)

Kikoku AM - 867 DPS
Kikoku no AM - 831 DPS

Kannagi AM1 - 955 DPS
Kannagi no AM - 833 DPS

Nagi AM3 - 832 DPS
Nagi no AM - 740 DPS

Here's Serac Rabbit, with sole sushi instead of RCB.

Kikoku AM - 950 DPS
Kikoku no AM - 889 DPS

Kannagi AM1 - 1035 DPS
Kannagi no AM - 895 DPS

Nagi AM3 - 876 DPS
Nagi no AM - 775 DPS

Download Apache's Open Office. It's Free :p

HERE is my latest DPS sheet with Kikoku's AM added, and a number of issues fixed.

Does your sheet implement the 40% bonus for Kikoku/Metsu?

Indeed it does.

Thanks for the clarification. Though I must admit, it does make me a bit less excited about the weapon. The BG page lists Light as the property while under AM so it was a bit misleading.

I may have to double check that then. My understanding is that it grants the merit WS the ability to close the radiance/umbra skillchains as if they had T3 properties. If it does in fact allow them to create double light/dark skillchains as well then that'd open up some flexibility for Heishi, as you'd have the option of spamming Ten (max damage per ws/darkness with a distortion partner), spamming Shun (self-light), or going for the four step radiance skillchain (max sc damage).

Does the spreadsheet replicate the 40% bonus to all hits/multi hits? Cause its hard to believe kikoku is still doing worse with 160 attack over kannagi in uncapped situations.

If you only use AM1/2, you should be able to spam Light with Shun.

Blade Metsu is a single-hit WS (all relic WS's are single, despite their animations).

The numbers I posted for comparison with Kikoku vs. Kannagi were both using Blade: Hi. I was making note that Blade Metsu's average damage was lower than Hi for Kikoku. Sorry if that was confusing.

The data above should clarify how Kannagi manages, if you look at it's average damage per attack round.

Kikoku did beat Kannagi when both had AM down against the original target I specified, so there is that..

offhand/multiattack?

This ^
I wasn't refering to the animation making it multihit, because its been known for a long time now that relic ws's are 1 hit.
But the damage bonus from relics/mythics applied to the ws gets added to all hits landed.

In short, yes, but I can't really speak to this as well as Mote could, but he's MIA so I will try...

His DPS sheets simulate approximately 500 weaponskills with the given set of buffs/gear, etc. The WS damage provided is the average. All scenarios are covered, like main hand proc's TA, offhand miss, or main hand miss and offhand DA's. etc.

Mote already had the relic bonus, so it's not something I did. He takes the final average damage output from the WS simulation mentioned above, and he then scales that final value by 140%, so yes off/hand and multihits are definitely factored in.

Just ran a few more tests, strictly using Kikoku (final form) to compare WS results.

- Uncapped Attack
Metsu avg = 3384
Hi avg = 4776

- Capped Attack
Metsu avg = 9114
Hi avg = 8588

I didn't compare capped/uncapped situations when I first compared Metsu, so I missed out on this scenario.

Excel Workbook (3/18/2016)
DPS - Calculator NIN.xlsx


What I've added for Gear:
Kannagi 119 III + Logic for OTD (Works the same as it did before, use the Setup page to declare your OTD/ODD rate. The data page will detect which version of Kannagi you chose in the gear page)
Kikoku 119 III + Logic for 16% Triple Damage (Works the same as it did before) + Toggle for AM on Setup Page for 10% Attack (not Rattk unless someone tells me differently - similar to Ores's setup page)
Nagi 119 III + Updated Stats (The OAx distribution hasn't changed has it? If it has, that's easy to adjust in the Gear Lists page)

Fixes to some Adhemar and Ryuo gear with incorrect values. Added some random new accessories like Enuma Mantle, Lissome Necklace, etc.

What I added for buffs:
Geo toggles for buffs/debuffs and their values. The debuffs will effect the targets values at the top of the setup page they are NOT set specific. (Much like Dia wasn't set specific before either) Frailty is implemented as an additive value with the value found in N4.

The values in the green box are adjustable as always in case you have an Idris geo present, adjust the values as necessary. The values shown should be those for a default 900 skill geo + dunna.

What I added for food:
Sublime Sushi

What I added for targets:
By calculating values from Falkirk I posted here: See my post

I created 3 new targets: iLvl135, iLvl140, and iLvl145.

The values falkirk gave aren't perfectly accepted for ALL target mobs of that iLvl, but it's something to go on. Using it's evasion as a measure, I created stats that would line up with the other creatures values. Until we get some serious data samples we won't have anything concrete.
Ex:
iLvl 135: Defense 1600, Evasion 1385, AGI305, VIT305, INT305
iLvl 140: Defense 1900, Evasion 1530, AGI355, VIT355, INT355
iLvl 145: Defense 2200, Evasion 1810, AGI420, VIT420, INT420

These could be entirely wrong (outside of evasion values, as that's based on empirical data) so if you feel they are wrong, they're in the "Other Lists" tab. You can go change it for your simulation.

And of course, double check the work if something seems fishy. Let me know where I've failed and I'll do my best to look it over and help correct it.

Edit: Still need to enable the Bolster/Blaze of Glory/Ecliptic effects.

What offhand should we be using for each of these RME? Is Kikoku/Kannagi even smart at all if you happen to be spamming Hi? To me, it doesn't seem like there are any decent RME options for offhand unlike other 1-handed jobs. Also, will we typically see more ninjutsu damage from RME/RME or RME/Ochu?

If MB'ing I'd say RME/Ochu or even Ochu/Ochu would win due to the +13 MBD and MAB that isn't on RME despite significant loss in magic damage ~80.

Free nuking.. I'd say RME, but when does NIN ever free nuke?

Hey buddy, just curious where you got that info for Shun? I couldn't find anything, but haven't really been following BG much.

Only things I really saw, was Metsu lacking relic bonus. Also, BGwiki is saying it's 5.0 http://ftp. Otherwise, your results seem in-line with mine, after a quick test.

I think the bbcode for your URL is busted.

Ha ha ha ha ha ha. In regards to Blade: Shun, I have no idea as I don't ever remember changing those rows. I remember reading there 'seemed' a potential shift to the Attack penalty/bonus and the anchor point for 1000TP was 0% attack modification instead of -8%, but now I'm trying to find it.

Thanks for the heads up about Metsu and my stupid BBCode link.

I can confirm that Kikoku aftermath does not affect ranged attack.

This is what I remember reading. (And now I'm trying to find that..)

Edit: Found it!

Something seems off; most of the things I plug in show Kikoku beating Kannagi in most situations with my gearsets. However, in practice, my Kannagi is miles ahead of Kikoku when I parse them against each other on the same NMs.

Edit: Also, Shigi is currently under the old stats. It's probably safe to assume that people will upgrade this weapon to the new version.

Which ws do you have each katana using in your comparisons?

I had them both using Ochu at the moment.

It might be the buffs? Just grabbing Langley's linked sheet, I changed no gear/weapons/formulas and just started tweaking buffs. For instance, turning off Berserk/Aggressor, Ionis, and Chaos Roll changed it from a significant Kikoku lead to Kannagi slightly ahead. Then turned off Minuets (leaving March/March/Mad) and the gap widens a little more.

I still think the Kikoku AM is not quite modeled correctly though. You can see this easily by turning the AM flag off, and simply manually adding Atk+100 to TP & WS set. The manually added Atk+100 (which is actually undervaluing the true amount) gives higher DPS than the modeled AM, which shouldn't be the case. To explain what I mean about undervaluing, listed spreadsheet base attack (TP/WS) was 1130/1179 for Hand 1, 1046/1089 for Hand 2 using the default gear (see rows 59 and 114 in the "Data" tab). 10% AM = 100/1024 = 9.76% times base attack, which results in over 100 for all of those attack values. But I just lowballed it and plugged in 100 for the sake of being excessively conservative.

Unfortunately I'm not your guy for figuring out how to do it RIGHT, but I have enough understanding to think something may be off.

I also won't discount the possibility that I'm misunderstanding how the relic AM works. I've been under the impression it's a simple attack+10% based on base (pre-food/buffs) attack whenever the AM is active, and will apply to all hits on both hands. If anyone knows that to be false, please clarify.

I fixed Kikoku AM in my sheet. You can add 100 attack to a piece of gear, and it's actually less DPS than turning AM on. It wasn't like that before I had added it to both hands, and to WS as well.

As far as Langly's sheet showing Kikoku beating Kannagi. Make sure both Katana's are using Shun lol. He's got Kannagi using Hi with Kikoku using Shun, which has REALLY good ws mods in his sheet.

Kikoku's AM and Kannagi's AM both have identical timers. If we're to assume Kikoku has time to pop AM with Metsu and start using Shun then we can assume the same for Kannagi. Empy AM1 does overwrite itself just like relic AM as well. Has to be apples to apples.

I'll take a look at the Aftermath implementation a bit later today and see if I can balance it out. Ores, did you ever change your pDIF implementation in yours?

What's a good target to test this on? I haven't killed much with Kikoku (actually couldn't kill Belphegor before I died using this) but each time Kannagi seemed to be ahead by at least 25-33%. Using my own WS set, the spreadsheet is spitting out way higher Metsu number than I've seen.

It seems inserting it where I did, lumped the 10% attak in with things like minuet, which I'm assuming it doesn't do...

I've replaced it as it's own line, only adding 100/1024 attack from base att and att from strength.

As for my Shun modifiers, I took gdiShun's input from his tests in my previous post and I undervalued his expectations and have it at -8@1k, 25%@2k, and 50%@3k. We certainly need more testing on it.

Edit: Capuchin: For kikoku's aftermath, does it add a flat 10% of base attk derived from (STR + Skill + gear(?)) to both hands equally? Or does it add 100/1024 to each individual hand.

For now, I'm putting it as 100/1024 to each hand separately including attack from gear (outside of skill), but would like clarification.

Changes made 3/19/2016 Kikoku Aftermath, Shigi attributes updated, geo Buffs moved to separate block.
Did not do: still waiting on implementing bolster etc).