Salvage II - Slavage Harder

[Motenten]

i have a mnk/war and thf/dnc on it, with the thf/dnc using haste samba and opening SC with shark bite and mnk closing light with shijin spiral.

I actually had a question on that. One of the previous runs we had -- where I was healer, and thus had damage filters on, so don't have parse data to examine -- the DDs were sam + thf, and constantly doing either Evis>Fudo Darkness, Konzen>Fudo Light, or whatever pair of weaponskills the sam used with Sekka or during Meditate. However things went bad around 20-30% where they complained that PDT was getting overly problematic (and then I decided to jump on the hate train with a couple sch nukes to 'help', and that was pretty much it).

From reading on here, skillchains -should- help prevent the PDT buildup. Does Cerb heavily resist Dark skillchains, perhaps? All the comments have been using Light skillchains. Or maybe they were just giving me bad information.

[Byrthnoth]

I do not think that skillchains are as effective as magic damage for some reason, or it could be that only light-half elements reduce his shield. When I spammed Evisceration -> Pyrrhic Kleos, he built a strong shield and I pretty much had to turtle up and wear him away. When I starting using Aeolian Edge -> Pyrrhic Kleos, the fights went quickly and easily.



Darknesses during fights without AE:
Seether: 584 (-73.5% expected), 1952 (+23% expected), 768 (-33.5% expected)
NQ: 660 (-49% expected), 1559 (+5% expected), 1936 (+21% expected), 1771 (+24% expected)

Darknesses with AE spam:
Seether: 159 (-87% expected), 1740 (+12% expected)
NQ: 2096 (+10% expected), 2363 (+24% expected), 1404 (+33% expected)
Seether: 2151 (+10% expected), 244 (-87% expected), 1743 (+28% expected), 1927 (+30% expected), 1913 (+30% expected)

Apart from the italicized values, they all matched this equation, where N is a whole number:
FLOOR( FLOOR( Closing WS damage * Skillchain Bonus ) * (1+N/100) )

The first italicized value works out perfectly for a 1/2 resist. The second one would require some kind of day/weather tomfoolery or my gear not swapping properly and ending up wearing AF3+2 hands - combined with a resist. You can see that even in the AE spamming fights where I'm comfortable with his level of shield, I still get a 20~30% magic damage bonus onto my skillchain (though it's unclear how much of that is caused by the closing WS.) You don't need to be particularly close to an even ratio of physical to magical damage, but I'd say that you need to stay about 75:25 to be optimally safe.

[mdkuser]

Assuming the last part of the formula used in Kaeko's enmity tables continues up to level 99 (it's 3 piecewise-linear equations), the amount cured (outside Cure 5/6, and not counting -enmity) needed to reach capped CE at 99 is 17250.


On our last Bhaflau run where I was whm, NQ Cerberus required 17564 HP cured, though that included 4 Cure V's (so 13380 cured outside of that). I did not pull hate.

HQ Cerberus required 22732 HP cured, and I did pull hate right at the end. Removing the Cure V and Cure VI used after that point (I was going for MP efficiency rather than low hate, and the DDs never took enough damage to really require Cure V), 20454 HP was cured. There were at least a couple other cures in the last moments as well, so I'd estimate I pulled hate around 18-19k HP cured.

My cure set only has -6 enmity (grip and Orison Earring). -6 enmity would push the CE cap from curing up to 18350 HP.


Obsidian's parse has him being cured for 27935 HP on NQ Cerberus. Removing the Cure V/Cure VI healing, 17353 HP was cured. The whm took 2 hits in that fight, so did manage to pull hate at some point. I expect the whm just barely crossed the hate threshhold at the end.

On the HQ Cerberus, the whm cured 25990 HP, or 20883 after removing Cure Vs. The whm took 1 melee hit in that fight, so pulled hate there as well, though it seems it was well controlled.



Overall, if you have a mnk tank (and thus one who has enough HP to absorb a full Cure V's worth of healing), a whm using Cure V should easily be able to stay under the hate cap. If you have sch curing, Regen V should compensate for a huge amount of the hate you'd otherwise receive, as a Perp+Regen5 can restore over 6k HP. If you have a brd, rdm or sch healing, or DDs that don't have enough HP to warrant Cure V, and thus generally using Cure III/IV, a decent amount of -enmity should suffice; -20 to -25 enmity is fairly easy to get while maintaining 50% cure potency, and sch can boost that a bit with Animus Minuo. And finally, if you have a thf along, a couple uses of Collaborator should keep any healer below the radar for the duration of the fight.

those numbers are a bit excessive . My last two parses had total HP cured 11,000 (verethragna MNK/WAR, so less on spharai in general). This include 2 bad TP moves for 1400 dmg, and several uncountered 600 dmg crits.

my mnk doesn't even have formless strikes. i don't see how people are having trouble with magma hoplon and the pdt. i have never had the stoneskin last for more than 10 seconds, and that's being conservative. i have a mnk/war and thf/dnc on it, with the thf/dnc using haste samba and opening SC with shark bite and mnk closing light with shijin spiral. thf is of course using feint/sneak attack and trick attack. never used collaborator or accomplice. i don't save tp for hoplon, i just do SCs through the fight and both fights take hardly any time. the whm starts out the fight with boost-str, auspice, barfira, shellra5, and dia2. redo the buffs right after a tp move after a couple minutes then back out to 20' range. dispel spikes from hoplon.

so THF and MNK DD not MNK solo. Obviously you'll not pull hate due to killing much faster and Skillchainging, even better if teh THF only uses aeolian edge. I have yet to pull hate since I merited formless strike on that mule.

[Obsidian]

I actually had a question on that. One of the previous runs we had -- where I was healer, and thus had damage filters on, so don't have parse data to examine -- the DDs were sam + thf, and constantly doing either Evis>Fudo Darkness, Konzen>Fudo Light, or whatever pair of weaponskills the sam used with Sekka or during Meditate. However things went bad around 20-30% where they complained that PDT was getting overly problematic (and then I decided to jump on the hate train with a couple sch nukes to 'help', and that was pretty much it).

From reading on here, skillchains -should- help prevent the PDT buildup. Does Cerb heavily resist Dark skillchains, perhaps? All the comments have been using Light skillchains. Or maybe they were just giving me bad information.

Dunno about Darkness, but Blade: Shun -> Victory Smite last night was doing maybe <300 damage per Light. On the Seether I was using Hundred Fists as ended up doing Smite > Shun for a Light that did ~200, and then Smite again for double Light which did 3000. That was the only Light I had break 400.

Edit: Those numbers were excessive because I was trying to evasion tank it without Counterstance like people have been suggesting, lol.

[Motenten]

those numbers are a bit excessive . My last two parses had total HP cured 11,000 (verethragna MNK/WAR, so less on spharai in general). This include 2 bad TP moves for 1400 dmg, and several uncountered 600 dmg crits.

Note that for my parse, that was with 3 DDs, so AOE damage is going to scale faster than against 1 or 2 DDs.

Dunno about Darkness, but Blade: Shun -> Victory Smite last night was doing maybe <300 damage per Light. On the Seether I was using Hundred Fists as ended up doing Smite > Shun for a Light that did ~200, and then Smite again for double Light which did 3000. That was the only Light I had break 400.

Does a 3-part skillchain have better magic accuracy than a 2-part? IE: less likely to get resisted? On the other hand, Byrth's numbers seem to indicate not nearly as much resistance on the Dark skillchains he did.

[Yugl]

SCs reduce magic evasion (Or SDT? Not much tested), so the first SC could have enabled the 2nd SC to do full DMG.

[Byrthnoth]

SCs have a large MAcc bonus over what they used to have. I do not know whether "continuing" skillchains have higher MAcc than two-step, but most of the things I fight don't even resist two-step Skillchains.

I do suspect Skillchain magic accuracy depends on the closing player's MAcc, which could be tested using Storm spells, Obis and Klimaform on SCH. We know that the skillchain chooses the damage type that the monster has the lowest resistance to at any given point (and has a priority order that we figured out), so if you can change that damage element to a low priority element using Klimaform then it also depends on the closing player's MAcc, and overall depends on the Magic Hit Rate.


As for why his resisted and mine did not, here are four options:
1) I showed several examples above of large MDT penalties against the Cerberus monsters. It's possible that the skillchains were just after the start of the fight or using Formless Strikes, in which case we'd expect him to have a large MDT penalty.
2) The sample size is small. Maybe it was just bad luck.
3) Skillchain Bonus could also give Skillchain MAcc (although MNK also gets level II, so I wouldn't expect it to make such a huge difference).
4) I'm a taru. High INT, yo!

[Brehn]

my whm has never pulled hate off the mnk tank. if your whm is taking hate, get a better tank.

Never said MNK. We were talking about a NIN/THF on it at the time. Like I said, its easy in a prolonged fight where you dont have some form of magic damage (Which again implies no MNK, Mura SAM or Twilight DRK.)

On top of that, if Cerb decides to just be an ass and use hoplon over and over again when he starts spamming TP the fight drags on for a considerable amount of time. EDIT: In that situation (super high PDT.)

[Obsidian]

They were happening through the fight so they were all crap except the one double Light. I was closing with Smite for 1000-1500~ usually so they were getting nerfed pretty hard. Unfortunately I didn't save the parse from last night, but I can save the one from tonight if you want it.

[Protey]

Never said MNK. We were talking about a NIN/THF on it at the time. Like I said, its easy in a prolonged fight where you dont have some form of magic damage (Which again implies no MNK, Mura SAM or Twilight DRK.)

On top of that, if Cerb decides to just be an ass and use hoplon over and over again when he starts spamming TP the fight drags on for a considerable amount of time. EDIT: In that situation (super high PDT.)

the only so called magic damage i have is in the form of SCs and Dia2. and like i said, bring a better tank, nin is lulz. if you don't have a thf with you to SC with urdoinitwrong.

[Cairthenn]

Implying anything more than a THF mule is worthwhile.

[Byrthnoth]

Or use Blade: Yu.

[Brehn]

the only so called magic damage i have is in the form of SCs and Dia2. and like i said, bring a better tank, nin is lulz. if you don't have a thf with you to SC with urdoinitwrong.

You did not read any of the posts did you Just that one I made, then took it out of context............

[Obsidian]

Here's the parse from today. Lights did a bit better, got one to break 1000, but overall still heavily resisted. Are you spamming Stutter step Byrth?

[Byrthnoth]

Nope, not using it at all. Just Elegy + Dia II + Box/Feather

I close skillchains with this set: http://www.ffxiah.com/item-sets/287555
I should have 91 or 93 INT in that set depending what food I'm eating.

[Obsidian]

Hm. I use this set for Victory Smite, and I eat Red Curry (Hume male, I have no idea what my INT would be at). I wanted to do one cerb with Lights and the other with Darknesses, but trying to make Darkness on MNK with a NIN is next to impossible, since MNK doesn't have access to a single skill with the Distortion property, and most of NIN's skills are Gravitation.

[Motenten]

A few points of int hardly seems enough to expect to have any significant effect, though. Obsidian's total should be... 79 int if mnk/war eating RCBs, so 12-14 point difference.

Do we actually have any understanding of what influences skillchain accuracy? Magic Accuracy seems silly, considering skillchains are created almost exclusively by melee, though can't be fully refuted due to the period where skillchain damage was horribly resisted on almost everything of note (IE: poor SE coding).

Perhaps something more like blu spells, which are primarily influenced by weapon accuracy?

Byrth's acc:
420 skill w/merits
35 acc from trait
40 acc in gear
89+12 dex if dnc/war
61 dex in gear
Total: 554

Obsidian's acc:
440 skill w/merits
32 acc in gear
91+12 dex if mnk/war
22 dex in gear
Total: 510
Also: 6 m.acc

A difference of 44 acc (minus maybe 6 m.acc). Enough difference that one might start thinking there could be a notable difference in resist rates, if it were based on that, but still a hard sell.

If Cerb was notably weaker to one of the dark elements, though, that could be significant. Hard to test without changing the current setups of those who've provided info so far, though.

[Darkmagi]

Kaustra resisted on seething cerberus quite heavily, so I don't think it's weak to dark.

[Motenten]

Well, wouldn't be surprised if it was weak to water. Doubt it's weak to ice. Maybe earth, as Hydra is pretty weak to that, but there's no reason to expect those to match up. Be easy enough to test with by cycling enspells. May try that next time, bring a rdm along to try out all the enspells (or sch with portable thf grenade).

[Darkmagi]

I'm stupid and didn't read the information about skillchains picking the weakest element, that's good to know. Being weak to water would make sense resist rate wise.