Dragoons: A Rotation Reborn

[Rystet]

I have been keeping up with this thread and I have not seen any mention of the server wide dot ticking mechanic. If this has already been mentioned I apologize, but I think it is relevant to the DOT Usage/Clipping Discussion.

All Dots in this Game Tick at the same exact time. Its like a server side DOT GCD of 3 seconds. So every Dot on the server "Pulses" at the same time.

So....

If you apply a DOT debuff at .5 sec into the Servers DOT GCD, the first tick will not happen for another 2.5 seconds.

The same situation will apply to the last tick of that same DOT. The Final Tick will happen and there will be window of 2.5 seconds where your debuff has fallen off, but you have not actually missed any potential DOT ticks. Applying the DOT debuff before the 2.5 would still ensure 100% uptime.

This can also work the other way, if a DOT is applied just before the Servers DOT GCD pulse, your debuff could be visible on the target for 2+ seconds, but the final Tick has already landed. So you Could reapply the DOT Debuff early without actually clipping.


Currently it is nearly impossible to track this pulse (set up a metronome?)... but in the future we may be able to create an addon that would impact our rotation.

[Sygmaelle]

There's something more about the dots and I think I kinda figured why FFXIV App is actually not able to track them down "properly", and instead doin a physical damage interpretation of.

Hint : a clue is in that sentence

Did anyone try to see if dots are hittin harder when you use a higher value of determination ? if yes its just a separated physical damage added to your attacks when you manage to land them when the dot(s) tick.

[Rystet]

There's something more about the dots and I think I kinda figured why FFXIV App is actually not able to track them down "properly", and instead doin a physical damage interpretation.

The parsers can not properly determine DOT damage because the combat logs do not include DOT damage, they only show that the Debuff has been applied. The damage is not actually recorded thus parsers have to estimate damage done.

I don't believe this is a coincidence, it is tied to the way DOTs tick (explained in my post above). The actual damage is done every 3 seconds on the server side when/if it registers the applied Debuff.

[Ayvar]

The same situation will apply to the last tick of that same DOT. The Final Tick will happen and there will be window of 2.5 seconds where your debuff has fallen off, but you have not actually missed any potential DOT ticks. Applying the DOT debuff before the 2.5 would still ensure 100% uptime.

This can also work the other way, if a DOT is applied just before the Servers DOT GCD pulse, your debuff could be visible on the target for 2+ seconds, but the final Tick has already landed. So you Could reapply the DOT Debuff early without actually clipping.

Currently it is nearly impossible to track this pulse (set up a metronome?)... but in the future we may be able to create an addon that would impact our rotation.

I think the reason there hasn't been much discussion on the subject is that it is completely out of our control. As you say, we may be able to do something to account for it later, but at the moment there's nothing we can do.

As for being able to reapply the DoT, even if it is done ticking, the DoT debuff is still registered as being on the mob, so an unbuffed DoT will still not overwrite a buffed DoT even if the buffed DoT has ticked already in the final 3s of its duration. You must wait for the DoT debuff on the mob to completely disappear before even casting the next DoT.

[Rystet]

As for being able to reapply the DoT, even if it is done ticking, the DoT debuff is still registered as being on the mob, so an unbuffed DoT will still not overwrite a buffed DoT even if the buffed DoT has ticked already in the final 3s of its duration. You must wait for the DoT debuff on the mob to completely disappear before even casting the next DoT.

Absolutely, that particular situation will only apply to non buffed DOT's.

But the point is there will almost always be some time Left on the Debuff after the final tick has been applied. The actual amount of time remaining could be anywhere from .1 to 2.9 seconds. So when factoring Rotations, and attempting to not Clip any DOTS (not just buffed ones) it is worthy of note that we have a bit of wiggle room.

On avg there will be a 1.5 sec gap between the final tick and the debuff falling off. Even if we play it safe and dont take the chance of factoring anything over 1 or .5... This could slightly change how we look at Skill Speed in relation to some of the suggested rotations.

[Ryuhou]

this is awesome rotation explanation. for almost DRG outside there already now how to rotate these skill, but this write is make my eyes open wider. good job!!

[notalltogether]

assuming 470 is the required accuracy rating (though this isn't necessarily decided yet). if we assume each point of accuracy gives the exact same value of hit. then a single point of accuracy would be worth .212766% chance to hit. this actually makes it slightly worse then determination if we assume that each ability hits on its own. however, missing an ability breaks the rotation, and really messes up a lot, so it should be prioritized. but for straight piece by piece comparison this is probably fair a good enough number to use until the impact of broken rotations can actually be calculated.

[Skies]

Hrm... If you need a stray GCD you might want to look at either Fracture or actually, if the boss isn't resistant, Feint.
Feint has a nice 20% slow debuff that is very helpful for your tank and also slows down move chargeups. Given your DPS might be more important but I like having chances to use Feint and help alleviate healer and tank lives.

[Zenmaku]

Hrm... If you need a stray GCD you might want to look at either Fracture or actually, if the boss isn't resistant, Feint.
Feint has a nice 20% slow debuff that is very helpful for your tank and also slows down move chargeups. Given your DPS might be more important but I like having chances to use Feint and help alleviate healer and tank lives.

Its rare that Feint actually works on endgame bosses these days, but it does work on one specific one that makes the fight feel easier (Garuda HM + her adds). Titan and Ifrit are immune to it I believe, since they have a specific pattern rotation that they follow.

[EasymodeX]

Generally speaking, clipping the CT itself is not a problem. The problem is when the first one is buffed and the second won't apply at all -- this will only occur every BFB cycle. Adding any GCD to the "IDC" rotation lowers the base DPS -- the only purpose here is to make sure the CT sticks.

For the H IDC P TTT H TTT // P IDC H TTT P TTT rotation, you only need to inject 1 Fracture (or any GCD attack) like so:

H IDC F P TTT H TTT // P IDC H TTT P TTT

That will delay the unbuffed CT by 1 GCD so that it sticks properly, and it only needs to be done once every 80s (or however often you are using BFB in the fight).

IDC-HP-TTT-TTT-HP-IDC-TTT-HP-TTT-TTT

I re-wrote this to get a better handle on it:

HP IDC TTT
HP TTT TTT IDC
HP TTT TTT

It's similar in approach to the original HFP rotation:

HFP IDC TTT
HFP TTT IDC
HFP TTT TTT

Except it uses the second TTT on the second row to space out the overall sequence, instead of Fracture on each row. Note: this "HP-based no-Fracture staggered IDC" rotation models for higher PPS (125) than the prior HFP rotation (124.2), but less than the "IDC" rotation (126.7). The IDC rotation with 1 injected Fracture rates at 126.3 (to solve the BFB CT clipping). The IDC rotation with a Fracture every IDC is 125.9.

Edit: Checked

HF-IDC-P-TTT-HF-TTT-P-IDC-HF-TTT-P-TTT

And that is 124.8. The staggered H/P give slightly better uptime for the various buffs.