How to Dragoon per Second: A DPS Paradigm

[whiskeybravo]

Hell yea, grats Jack! Sounds like a solid # to me ^_^

[Mantrus]

Hey guys, sorry if this question was already addressed or brought up, but for the opener is it worth doing HT > IDC4 > TTT4 > TTT4 instead of HT > IDC4 > TTT4 > IDC4 (i.e. 2 triple thrust combos in a row instead of alternating) because the first Chaos Thrust will be buffed by B4B, Litany, IR, possibly Trick Attack, and a potion?

[JackFross]

Not sure. Gonna run a quick examination. I feel intuitively that it would be a loss, since Disembowel is going to fall off before you re-apply it, by a substantial margin.

When clipping (assuming you didn't also have Arrow), you lose ~1.8-2 (~2 when Fey Wind is a thing) ticks of the hyper-buffed dot and replace it with the not-so-buffed one.

1 dot tick is 35 potency
HT+B4B+IR+BL+TA+Potion
35 => ~75 potency.

Just HT (what you'd get re-upping it combo 3)
35 => ~40 potency.

So you're turning (adjusting for the 1.8-2 ticks) 134-149 potency into 72-80 potency, a gain of 62-69 potency on those two ticks.

Running a quick test myself on a dummy at 604 skill speed, opening with IDC4>TTT4>TTT4>IDC4 causes Chaos Thrust to be inactive for ~7s. That would be ~2.2 extra ticks of the ~40 potency dot, which is a gain of ~88 potency just from not letting Chaos Thrust fall off. We can also factor in the fact that, in my parse, Disembowel falls off at :31, and is re-applied at :37, 6 seconds of time during which I cast Phlebotomize, Impulse Drive, and Disembowel. These are 170+180+220 = 470 total piercing damage that loses this buff:

HT+Dis
470 => ~594 potency

HT
470 => ~540 potency

Which is a further loss of ~54. Altogether (if you carry the decimals that I truncated for easy reading), we lose 142.6 potency from Dis and CT falling off, but gain 62.2-69.1 potency due to the 1.8-2 ticks being buffed. That's a net loss of 73.5-80.4 potency up through the re-application of Chaos Thrust in this modified opener.
That is, we're comparing:
H>IDC4>P>TTT4>H>TTT4>P>IDC
H>IDC4>P>TTT4>H>IDC4>P>TTT

Images of these timers below. First is buff/debuff, second is showing the rotation I used.


You can probably extrapolate this and figure out exactly how much you would have to clip the initial CT before it'd be worth it to NOT clip, but I think this should do enough to explain it and why it's a loss.

Also, if you actually intended H>IDC4>TTT4>TTT4 and not H>IDC4>P>TTT4>H>TTT4, you're going to lose even more potency on the loss of Heavy Thrust on the Full Thrust and 4th (650 potency losing a 15% buff) hits of the second TTT4 rotation, as well as still losing Disembowel for the 220 potency Dis re-application AND still losing a full tick on Chaos Thrust's timer. And none of that factors the loss of a fully-buffed Phlebotomize.

If you meant H>IDC4>P>TTT4>TTT4, skipping HT, you lose the 15% buff on VT>FT>4th (850 potency) AND you lose Disembowel on ID>Dis.

Without running calculations for those alternatives, this one I ran is the one with the least loss, I would assume.

[whiskeybravo]

It's in the OP:

Quote Originally Posted by Thendiel View Post
In this post, I want to compare a rotation that involves Disembowel and CT clipping:

HT - ID-Dis-CT-4 -Phleb - TT-VT-FT-4 - (repeat)

...to a rotation that does not involve Disembowel and CT clipping:

HT - ID-Dis-CT-4 -Phleb - TT-VT-FT-4 - HT -TT-VT-FT-4 - Phleb - ID-Dis-CT-4 - HT - TT-VT-FT-4 - Phleb - TT-VT-FT-4 - (repeat)

Basically, I was wondering if and when it would be useful to clip Dis and CT vs. letting them fall off for a few seconds.

[JackFross]

I was gonna reference that, but the question was specifically regarding the opener, and the math there only compares the general case. They were looking for specifically how the numbers shake out regarding a fully buffed CT being clipped early in the first 30s of the fight.

[Mantrus]

Thanks for the calculations, Jack. I've been wondering that for a while and I always felt a bit bad overriding an empowered DoT, but now I won't have to feel like that. And yeah I meant to have HT and Phl in-between the other combos, but I forgot lol. I think you missed the extra 60 potency from the TTT combo over the CT combo, but it doesn't really matter since it still comes up short in potency even ignoring auto-attacks and ranged DPS piercing damage. And yeah I meant specifically just for the opener.

[CookieMonsta]

^ That's actually exactly my feelings regarding dungeon runs. If I'm running on DPS, I play BLM instead. It's super easy single-target and AoE in every level 60 dungeon.

ALSO.
http://www.fflogs.com/reports/naP9Kt...pe=damage-done

\o/ A4S down, yayyy. 1345 while doing adds seems pretty good? Not totally sure, though. I'm just happy to have it down.

Gratz!!!
Now time to slack!

[whiskeybravo]

I was gonna reference that, but the question was specifically regarding the opener, and the math there only compares the general case. They were looking for specifically how the numbers shake out regarding a fully buffed CT being clipped early in the first 30s of the fight.

I guess I just took:

"we are never going to have enough Skill Speed for a non-clipping approach to be worth it."

too literal :P

That's a net loss of 73.5-80.4 potency up through the re-application of Chaos Thrust in this modified opener.

I did a little chart out of boredom, basically came to the same result (81.65 pot loss):



For the dots I just added up the potency at the end/start of the next dot. 8 ticks on the clipping approach, all 10 on the non. Added leftovers at the end

[JackFross]

I think you missed the extra 60 potency from the TTT combo over the CT combo

I was comparing these two strings:

H>IDC4>P>TTT4>H>TTT4>P>IDC
H>IDC4>P>TTT4>H>IDC4>P>TTT

Stopping calculations at the point where CT would be re-applied in the first one. Since it's the same exact set of GCDs just in a different order, I got to ignore the differences in potency between the two different openers and just focus on the more minor differences (dots and disembowel).


Also, Whiskey. Yesterday, out of boredom, I spent ~7-8 hours straight updating my excel spreadsheet in preparation for 3.2. It's still not even close to finished yet. I'll put this junk under a cut since most people will be like "Ariyala exists u nerd" and my argument is "Yes, it does, but it's so ugly."

This is what the main page of it looks like:


All of the buttons are fully-functional, the radio buttons adjust dropdown menus in each of the slots, Race is fully integrated... It's set up where it CAN work for other Jobs, I just didn't care enough to fill in base stats or gear listings for anyone else, since I don't play any of the other jobs that Dervy has figured stat weights for.

O - saves the set on a new sheet
X - deletes the set's sheet, if it exists. doesn't let you delete important sheets.
PURGE - deletes all sets you've created.
Clear Gear - clears every Gear slot to blank.
Clear Materia - clears every Materia slot to blank.
Sort Gear Listing - sorts the gear listing for the Job that your set is for.

There's also sort functions for the Job page and Food page so you can add new stuff yourself and shuffle it in without much issue, using the custom sorts I've made to keep things organized properly.

Making a new Job gearset is as easy as copying the DRG Gear listing and naming it XXX Gear - where XXX is the 3-letter abbreviation for the Job. I *might* add in some other Job full functionality before calling it finished, but I think it's turning out well.

If anyone has ideas for stuff I should add, let me know! I'm slowly learning Visual Basic to program this sheet, too. C:

Currently, I'm in the process of seeing if it would be possible to have a dynamic picture on the set page that adjusts based on which Job you are...

//Edit: That was a lot easier than anticipated.

[whiskeybravo]

Nice. I spent a couple hours the other night trying to come up with some way to visualize and arrange the various buffs, like the alignment thing that was being discussed before. Still struggling a bit with it, what I came up with was just ms paint images to scale, like HT was 240 pixels, IR 600 pixels, b4b 800 pixels, etc. But having a hard time coming up with an accurate sort of grid to plot them on.

Edit.. That's a really nice spreadsheet, totally jealous lol. I know Excel is pretty powerful, but unfortunately outside of basic math formulas I'm a total noob with it lol.

I used to code in C++ and VB many years ago, recently downloaded android studio to play around with but haven't attempted anything serious. I'd like to have an unofficial Dragoon companion app :3