 |
 The Voracious Resurgence |
 
| |
| Prime Weapons • Ultimate Weapons • Ultimate Augments • Abjurations iL119 • JSE Necks • Divergence Augments • Escutcheons | |
| Reforged Armor Artifact: +1 • iL109 • iL119/+2/+3 Relic: +1/+2 • iL109 • iL119/+2/+3 Empyrean: +1/+2 • iL109 • iL119/+2/+3 |
Guides • Crafting • Trusts • Apex Monsters |
User talk:Izanami
Mochizuki Kyahan +3
Curious about your recent edit to Mochizuki Kyahan +3. The page now contains both:
- Increases Ninjutsu Magic Attack Bonus by +5% Ninjutsu damage per merit level, for a maximum of +25%. This is a separate multiplier than Magic Attack Bonus.
- Increases Magic Attack Bonus of elemental Ninjutsu spells by +5 per merit level. This does not apply to weapon skills, or other forms of magic damage.
Is the implication of this based on your testing that the Augment effect applies both the previously-stated 5% Ninjutsu damage per merit level which is a separate multiplier than MAB, and a +5 MAB per merit level? Hya (talk) 14:29, 23 March 2025 (UTC)
--
Yes, I find that the augments on Mochizuki Kyahan +3 provide +5 Magic Attack per merit to Ninjutsu spells in addition to the +5% damage dealt per merit.
In 2022, I fully constrained the Ninjutsu damage formula for all possible dINT values for all three nuke tiers (see https://www.ffxiah.com/forum/topic/56749/updated-ninjutsu-damage-formulae/). I just now got around to writing a code to return enemy INT given player stats and observed damage dealt. However, I realized that my testing on Lv0 huge hornets (INT=6) with Mochizuki Kyahan +3 equipped did not match the +25% damage dealt as suggested on BG Wiki. Mochizuki Kyahan +3 do not have any INT on them so testing this is straightforward.
When compared to no gear equipped, equipping only Mochizuki Kyahan +3 increases damage by about 46% against Lv0 Huge Hornets in South Gustaberg (and by about +33% with a nuking set equipped), which is clearly in conflict with the +25% originally on BG Wiki. For specifics, damage dealt with Katon: [Ichi, Ni, San] is [313, 717, 1141] without any gear equipped. Equipping only Mochizuki Kyahan +3 increases damage to [458, 1048, 1665].
Based on the results of my complete Ninjutsu damage testing in 2022, I am able to calculate the observed damage dealt perfectly for an INT=6 Huge Hornet in multiple nuking sets only if I use a multiplier of (1 + 0.05*merits; 1.25 max) and increase my Magic Attack by +5*merits (+25 max). I tested this with 3/5 merits as well and found a 1.15 multiplier and +15 Magic Attack worked perfectly. I additionally tested Mochizuki Kyahan +3 with NIN/BLM using Fire1 and with Blade: Ei. Neither saw any benefit from the hidden "Magic Attack +5 to +25" augment.
Similarly, my code to calculate enemy INT successfully recovers "Apex Toad 293 INT and 25.0% MDT" and "Apex Vodoriga 318 INT with 12.5% MDT".
I use the following equation to calculate damage dealt, which works perfectly for Ichi, Ni, and San nukes on my test cases above.
dmg = floor(player_mdmg + v + dint*m)
dmg = floor(dmg * skill_multiplier)
dmg = floor(dmg * matk_multiplier)
dmg = floor(dmg * (1 - mdt/100))
dmg = floor(dmg * (1.0 + 0.05*relic)) # "relic" is an integer with possible values 0,1,2,3,4,5
--Izanami (talk) 17:04, 23 March 2025 (UTC)
--
I appreciate the thorough explanation. I was able to replicate your 46% increase in damage dealt against Huge Hornets wearing only Mochizuki Kyahan +1 (and no Ninjutsu-specific Group 1 Merits, only 5/5 Ninjutsu MAB), so it appears safe to apply the same edit to the lower tiers of Mochizuki Kyahan as well. Thank you for your work. Hya (talk) 18:16, 23 March 2025 (UTC)
Hey, I see you are running into the issue of exceeding template limits. To get around this name a page "Title/<job>". Then on the main title page, in this case "All Jobs Gear Sets" type:
{{:All Jobs Gear Sets/<job>}}
For example:
{{:All Jobs Gear Sets/Warrior}}
This will call that page to the main title page. So you can have as many templates for gear as you want by breaking it up this way within the limit. You will have to edit those pages individually (there is also a recache period) if you have changes, but this is how massive pages with tons of template calls have to be handled. --Spicyryan (talk) 21:13, 30 April 2023 (UTC)
Thanks. Unfortunately, it seems that method is not enough to get around the limit. The main title page appears to be stuck at a shared 2MB "Post-expand include size" limit. Each single gear set appears to take up ~19KB worth of that limit due to the 16 32x32 pixel equipment icons included per gear set. This leaves enough space only for 110 gear sets total per single BG page, which is only about one-third of what I was hoping to include.
In addition to the "Post-expand include size" limit, I'm half-way to reaching the 2MB "Template argument size" limit with only 5/22 jobs included.
Unless there is a way to get around these limits, then I think the only option is to include links to the job sections instead of embedding them into the main page, which is less interesting.
-Iza
I've not messed around with a limit after transclusion to verify this. I was under the belief that transcluding several pages into one would bypass the template limit. I guess I can check with your pages later and find out. Still appreciate you adding these here though.--Spicyryan (talk) 13:52, 10 May 2023 (UTC)
Souleater
Hey, I am interested in the update you made on souleater and that it is affected by weapon skill damage bonus stat. Is there a post or data verifying that? I maintain a personal excel calculator that I like to update to reflect the equations on bgwiki but I also like to validate the sources.
--
I don't think I ever made a public post for the data, but I tested this myself as NIN/DRK with Blade: Kamu (FTP=1.0 at all TP values) against Lv0 Huge Hornets in South Gustaberg in early 2025. This could use more testing to really finalize everything, but taking the values from my Souleater testing spreadsheet: my testing is as follows:
- Used Blade: Kamu 27 times against Lv0 Huge Hornets (average damage = 2434.8, max/min = 1.046, the max/min should have a max value of 1.05 from the 5% random factor to physical damage.)
- Used Blade: Kamu 20 times with Souleater up against Lv0 Huge Hornets (average damage = 2594.9, max/min=1.042)
To calculate damage dealt and compare to observed damage I had the following stats:
Weapon Damage = 157 (Gokotai for Regain set)
WSC = 348
fSTR = 25.444
FTP = 1.0
WSD = 26%
PDL_Trait = 0.3
PDL_Gear = 0.0
HP = 2611
- With these numbers, I can insert the Souleater bonus at all possible locations in the damage equation and see how the final predicted average damage compares to the observed average damages.
When adding Souleater in different parts of the equation, I calculate the following expected average damage dealt:
No Souleater: 2432.0 (2.8 damage off observed)
With base damage: 2938.6 (343.8 difference) (base_damage + SE) * FTP * (1+wsd) * PDIF * (5%_Randomizer; used 1.025 multiplier)
After FTP: 2938.6 (343.8 difference) ((base_damage)*FTP + SE) * (1+wsd) * PDIF * (5%_Randomizer)
After WSD: 2834.1 (239.2 difference) ((base_damage)*FTP*(1+wsd) + SE) * PDIF * (5%_Randomizer)
Before 5% Randomizer: 2545.6 (-49.6 difference) ((base_damage)*FTP*(1+wsd)*PDIF + SE) * (5%_Randomizer)
After everything: 2542.5 (-52.4 difference) ((base_damage)*FTP*(1+wsd)*PDIF*(5%_Randomizer)) + SE
With WSD, no FTP 2571.2 (-23.6 difference) ((base_damage)*FTP*(1+wsd)*PDIF*(5%_Randomizer)) + SE*(1+WSD)
With WSD and FTP 2571.2 (-23.6 difference) ((base_damage)*FTP*(1+wsd)*PDIF*(5%_Randomizer)) + SE*(1+WSD)*FTP (FTP=1.0 in this case, so no change from previous)
With WSD and PDIF 2926.3 (331.4 difference) ((base_damage)*FTP*(1+wsd)*PDIF*(5%_Randomizer)) + SE*(1+WSD)*PDIF
This test best matches the observed average damage if the Souleater bonus gets WSD, but no PDIF.
Edit: My spreadsheet tested "With WSD, no FTP", but had the label as "Souleater with WSD, not PDIF". I just edited my response here to correct the labels in the above table.
Next, I repeated with Fotia Belt and Fotia Gorget:
- Used Blade: Kamu 13 times against Lv0 Huge Hornets (average damage = 2848.4, max/min = 1.029)
- Used Blade: Kamu 9 times with Souleater up against Lv0 Huge Hornets (average damage = 3041.2, max/min=1.044)
With updated stats:
Weapon Damage = 157 (Gokotai for Regain set)
WSC = 348
fSTR = 25.444444444
FTP = 1.1953125
WSD = 26%
PDL_Trait = 0.3
PDL_Gear = 0.0
HP = 2390
When adding Souleater in different parts of the equation, I calculate the following expected average damage dealt:
No Souleater: 2907.0 (58.6 damage off observed)
With base damage: 3512.6 (471.3 difference)
After FTP: 3413.6 (372.4 difference)
After WSD: 3309.1 (267.9 difference)
Before 5% Randomizer: 3020.3 (-21.0 difference)
After everything: 2907.0 (134.2 difference)
With WSD, no FTP 3046.2 ( 5.0 difference)
With WSD and FTP 3073.4 ( 32.2 difference)
With WSD and PDIF 3497.8 (456.6 difference)
This test best matches the observed average damage if the Souleater bonus gets WSD, but no PDIF. This is true no matter which single recorded Blade: Kamu damage I delete from the small sample (ie this is not affected by outliers so I larger sample would help, but likely not change anything.)
Edit: My spreadsheet tested "With WSD, no FTP", but had the label as "Souleater with WSD, not PDIF". I just edited my response here to correct the labels in the above table.
- While both data sets use a small number of trials, together they rule out most locations fairly well.
- It's clear that Souleater bonus does not get applied with the base damage, after FTP, after WSD bonuses, or after everything. It is also clear that Souleater is not enhanced by PDIF.
- We could make a case that it gets the 5% randomizer, but that seems weird to me since we don't see that for auto attacks. We'd need a perfect max/min=1.05 to rule this out for weapon skills completely, though.
- I interpret the results to suggest that the Souleater bonus gets WSD, but not FTP. Thus it would be it's own additional term in the equation as "total_damage = physical_damage + (0.1*HP)*(1 + WSD)".
I believe this could still use more data (preferably a perfect max/min=1.05 in all 4 test cases), but the main result is that Souleater is clearly a tiny increase in modern FFXI and generally not worth the lost HP per swing. When I typed up a summary of these results to the person who requested the testing, I added:
"""
It's +10% of your HP (+5% subjob) added as damage at the very end of every other calculation step.
For weaponskills, it gets WSD+ since it counts as Weapon Skill Damage, but it does not get FTP or PDIF multipliers.
For example: A RDM/DRK doing 90,000 damage Savage Blade with 80% WSD stat with 3000 HP will instead do 90,270 damage.
A DRK using Origin for 90,000 damage with 9999 HP and 80% WSD stat would do 91,800 damage instead. But it is a large increase to white damage at +1000 damage per swing and +25 Accuracy. I believe DRK typically deals 1900~2100 damage per melee hit, so Souleater adds maybe +25~50% damage per swing depending on crits and AM3.
Going a bit further. I checked the TP/WS damage distribution for a DRK/SAM spamming Origin in a high buff situation and it's 17%/83%. Meaning that a 9999 HP DRK spamming Origin at 2k TP would gain 8.5% DPS by using Souleater.
This scales linearly with HP, so only 5000 HP would be a +4.25% DPS increase and 2800 HP would be +2.6% DPS.
This is only true if you're at full HP for every auto attack and every WS. So it only applies when you have Blood Weapon or a healer spam curing you.
In summary: Souleater is only worthwhile if you NEED the Accuracy+25, or if you're a DRK main job AND landed a decent Drain3 AND used Blood Weapon...
"""
I did not test, and do not know, if this suggests that Souleater gets WSD on multi-attacks for weaponskills. WSD in general only applies to the first hit of a weapon skill (outside of special WSD sources), so I assume that Souleater only gets WSD on the first hit as well, if my conclusion above is correct.
--Izanami (talk) 21:15, 23 March 2025 (UTC)
Thanks, I did some testing as well assuming Gyudon functioned on 1st hit only. DRK/SAM fighting ogrefly in litelor. 29dmg base weapon, Gyudon+1 (6%WSD),196 strength AF head(12%souleater), 0.5 dmg limit+, 2122HP
Max Smash Axe WS dmg without WSD applied to SE should have been something like: (29base+11fstr+196wsc)*1ftp*3.75pdif*1.05r*1.06gyudon + 2122*0.12 = 1239. And that is without flooring terms anywhere...
I did 12 WS at max HP with SE and Gyudon+1 and broke 1239 two times. One weaponskill for 1242 and another for 1245. My average was also much closer to the value with Gyudon applied. I just wasn't sure I had a high enough data either. Thanks for sharing.
And if you are interested, I did a similar test with overwhelm, subbing DRK, and it suggests overwhelm doesn't improve SE damage. Which i don't think contradicts this, i just think overwhelm and how it is applied is different than Gyudon and WSD 1st hit applications.
--Bolmster
