(Don't) Fear The Reaper: A Dark Knight Guide

Lib gains so quickly that you should *try* to WS at 1k, because functionally overflow will put you closer to 1500-2k simply due to latency. Insurgency at 1k TP is functionally unrealistic.

That's exactly where the initial claim came from.
/edit
I should say that the anecdote came from comparing a full Dynamis Bastok run with a LSmate, he was on R15 Calad the whole time, I was on R15 Lib. Same party, same buffs, one stupid death each. Now I wish I saved the parse, as I generally don't. That was where the 65% more WS number came from, in general 50% seems more of a reasonable comparison, from my experiece, with large room for error considering how many things can and do throw real numbers off.

At this point the only thing that makes sense to me is not a parse, but comparing the two on video.

Are you mixing me up with the other person? I never claimed 50-65%, nor have I been claiming that Liberator is stronger than Caladbolg.

I have both parses (though long gone by now since I was just doing it for myself; I could try to drum up a COR and go beat on some Campaign fortifications for new ones, I suppose) and calculations that show that, while the difference isn't as big as the other person claimed, it's still quite significant (~30-40%).

Geriond, want to put this to rest? We are both on Asura. I dont have a Rostam, but I do have Regal neck. Campaign fortifications Seem reasonable, I have OBS set up but my upload is crap so it could take a while to post video.

ItemSet 369395
LibAM3 set I'm using.

Valorous greaves-Critical damage +5%, Acc 29, att +2
I'm sure it's not bis, but works well enough for me. And I definitely can tell that I weapon skill far more frequently than every one else around me in dynamis. I do exactly what Mims said I just go right at 1k.

Even so, you shouldn't be trying to WS anywhere before 2k because by 2k your fTP has increased almost sevenfold. And even at that point Insurgency's fTP continues to massively increase even up to 3k TP (12 times as much fTP as 1k), so the aim should be 2k + overflow instead of spamming. You're wasting the stacked WSD otherwise.

Insurgency is a non-ftp transfer 4-hit WS that has a high chance of multi attacks from AM3. Your total fTP doesn't increase nearly that much with higher TP.

Sorry to kill the good discussion going on. But figure id ask. Does anyone have a good drk lua? Ive tried using the one posted on the main page but for some reason i cant get it to stay on my weapon of choice lol.

It should matter because you can stack significant amounts of WSD for that first hit that gets tremendously stronger with just one extra attack round to go from 1500 to 2000+ TP. And you will have to deal with fewer overall forced WS delays. Multiattack procs are far less reliable damage spikes in comparison.

It does matter, but it's not a "seven-fold' fTP increase from 1000 to 2000, and even at 2000 TP WSDMG applies to less than half of the WS.

It's roughly equally optimal to use Insurgency at any TP between 1750 and 2750 (so 1500-2500 with earring, less with Warcry/Crystal Blessing).

Obviously I was strictly talking about the fTP increase for the first hit when I said sevenfold, not the entire multihit WS.

If you properly gear for a decent Crit build to improve white damage and take the 2s forced WS delays into account then holding your TP should absolutely be a DPS increase. Again, you won't proc extra multihits every WS, so a stacked first hit is more consistent high damage.

But yes, if you are getting WAR main Warcry or Crystal Blessing then you'd obviously WS much earlier.

Dark Knight doesn't benefit too much from a white damage build unless using Empyrean. Gearing for primarily STP and WSing at those TP values gives the highest total DPS.

It's not that difficult to gear for a 3-hit STP build without neglecting TA and Crit rate/damage.

ItemSet 369387

Petrov Ring may be better than Flamma, but the difference is extremely marginal.

In any case, I will try to parse the two different setups at some point in the future just to make absolutely sure.

No, but that reduces your TP speed you get (with some exceptions for TA/QA on some slots), and white damage is a very low proportion of the damage with a non-Empyrean 3-hit setup.

Its easy to get blinded by the raw fTP value of Insurgency, its non-replicating multihit nature adds a bunch of damage from extra hits.

One way to get a better picture of how the extra hits factor in is to treat them as added http://fTP. I know its not a perfect model, but for the most part it works. I threw things into a spreadsheet and will now proceed to get the code tags to format properly:

Assumptions:
68% Weapon Skill Damage applying to first hit only
AM3 provides 0.8 additionall attacks per proc
0 additional multiattack, fudging those numbers on top of AM3 is more than I want to do right now
Fotia belt only, no gorget
if AM3 provides .8 additional attacks per proc and can proc twice per Insurgency, there will be 4.6 extra hits per WS from multiattack

Now, latency and reflexes being what they are, functionally at capped haste you will get an attack round in between hitting the WS key and the signal actually getting to the servers in Japan, which makes the WS fire off closer to 2k TP due to PS2 limitations.

From the table above, 3k TP insurgency is only about 44% more damage than 2k TP, assuming 0 multiattack outside AM3.

And in practice, even trying to WS at 1k every time there are enough functional delays to WSing that if you're going to be even further delayed.

/edit Table formatting is being stupid, let me try to straighten it out

Its easy to get blinded by the raw fTP value of Insurgency, its non-replicating multihit nature adds a bunch of damage from extra hits.

One way to get a better picture of how the extra hits factor in is to treat them as added fTP . I know its not a perfect model, but for the most part it works. I threw things into a spreadsheet and will now proceed to get the code tags to format properly:
This is a visualization model for Insurgency, and should be tweaked should you apply it to other things.
Assumptions:
68% Weapon Skill Damage applying to first hit only
AM3 provides 0.8 additionall attacks per proc
0 additional multiattack, fudging those numbers on top of AM3 is more than I want to do right now
Fotia belt only, no gorget
if AM3 provides .8 additional attacks per proc and can proc once per Insurgency, there will be 3.8 extra hits per WS from multiattack

Now, latency and reflexes being what they are, functionally at capped haste you will get an attack round in between hitting the WS key and the signal actually getting to the servers in Japan, which makes the WS fire off closer to 2k TP due to PS2 limitations.

From the table above, 3k TP insurgency is only about 48% more damage than 2k TP, assuming 0 multiattack outside AM3.

And in practice, even trying to WS at 1k every time there are enough functional delays to WSing that if you're going to be even further delayed.

/edit Table formatting is being stupid, let me try to straighten it out
/edit2 The table seems to be aligning decently now, good god formatting is horrible
/edit3 The 1250 TP values are totally off, redoing them
/edit4 Values incremented by 250 TP, 4.6 model removed

The difference is in the balance. You lose very little as far as TP speed is concerned but you gain a decent amount of white damage.

It doesn't require Empyrean AM3 just to be worthwhile, Empy AM3 simply affects the ratios.

You lose enough of the TP speed for it to outweigh the white damage increase. I've checked it using my spreadsheet using a number of possible crit setups after Hjarrandi came out.

AM3 can only happen once, not twice

Beyond a 3-hit, how much TP speed do you realistically gain by swapping out Crit pieces for multiattack/STP, though? Take my set for example, which pieces would you replace to gain any kind of significant TP advantage? Mind you, any extra DA/TA devalues AM3 slightly and more STP in gear is devalued by Samurai Roll.

The spreadsheets are good, but they aren't 100% accurate.

Good catch, I wasnt sure about that which is why I included the 3.8 model in the middle, just to be safe. 2k to 3k is about a 48% increase by that model.

In that set, it would just be swapping any crit rate/damage augments on the feet to STP. More TP overflow will get you to that magic 1500 TP value in less rounds more often, or let you use Hetairoi Ring instead of Flamma/Chirich +1. The Yetshila +1 is the only slot where crit stuff might be better (I'd have to check again).

The augments on the legs and feet are STP by necessity to get a 3-hit build.

I was talking mainly about potentially improving on Hands/Ammo/Cape with Crit+10%/Ring1.

Aside from the misconception about AM3 procs on WS, I'm having trouble making sense of the point you're trying to make with that fTP table.

Assuming 100 WSs at 1250 TP with double procs vs. 62 WSs at 2k, you not only have more total "fTP" (nevermind that damage multipliers don't actually work like that) in the latter case, but also 76 seconds less forced WS delay.

Am I missing something here or somehow reading it wrong?

The point is to show how Insurgency damage actually scales with TP, multiattacks included. It does not factor in things like base damage, stat bonus, or Liberators insurgency bonus, as they are effectively just multipliers on top of these values. It does not take into account anything about TP rate, and is not meant to.

As for WSing at 1250, as I said earlier its unrealistic to spam at that value in any case and 1500-2000 is more realistic as a minimum value for spamming.

Even with your updated table, 100x1500 TP vs. 75x2000 TP WSs are very close in total "fTP", but you have nearly a minute less forced WS delay. I suppose it's like Geriond said that the TP range doesn't make a huge difference.

In any case, holding TP until 2k should have a decent advantage in longer, non-zerg fights.

I'm not getting into TP rate vs WS timing yet, at all. If you have a good way of modeling how AM3 interacts with multiattack from gear I'm all ears and would be happy to get all mathematical with it.

So just to be clear, when you spam Insurgency at minimum TP you're stacking multiattack gear instead of WSD with a set sort of like this, correct?

ItemSet 369401

I never said that, do not put words in my mouth.
You keep accusing me of pushing the "Insurgency at minimum TP" idea, which I do not and have never pushed.

I have specifically stated that focusing on multiattack is counterproductive and redundant, and while Insurgency looks similar to Upheaval at a glance, its pointless to mirror an upheaval build because WSing at 1000 TP + Moonshade is functionally unrealistic.

What I have stated, many times, is that even if you are aiming to WS at 1000 TP, latency, 2002 PS2 Netcode, and raw TP overflow will cause you to functionally WS around 1500-2000 TP, if not even more.

You want to see my Insurgency set?
ItemSet 360043
Here it is. I use WSD gear fulltime instead of multiattack at low TP because low TP, shockingly, isn't a thing, and will never be a thing before SE cleans up their netcode and moves their servers, which is not happening.

/edit
Since you seem to like making wildly inaccurate claims about what I mean when I say things, what I mean by "focusing on multiattack is counterproductive and redundant" is two things.

First, "counterproductive." I am not saying that adding multiattack on top of AM3 is somehow detrimental. Multiattack does increase damage. What is counterproductive, is when your client recognizes that you are at low TP, and your lua sends out a low-TP Upheaval style gearset, chances are good that through whatever Dynamis latency fog is going on, you are going to have more TP when the servers actually get around to processing your gear packet, pushing you into into a TP tier wherein WSD gear is comparable, if not favorable, compared to multiattack gear.

Second, "redundant." This is the obvious issue of multiattack taking priority over AM3. Gains exist, but are minor, and if you are using Insurgency without AM3, you are using the wrong weapon in the first place.