How to Dragoon Per Second: A DPS Paradigm

[Dervy]

Don't ask me how I know this, but there's two branches under Critical Hit Rating. Critical Power and I think the other is called Critical Strike. That's their official internal Development names.

Critical Hit Rating is just our flat stat that we see on our Character panel and it's this stat that affects both.

Critical Power is... Critical Power. Our Crit Damage. Critical Strike then is Critical Hit Chance.

Now the issue is that the English Localisation Team are scrubs and forgot to change the tooltips for the new Changes, which is what is causing the issue - or it's probably because they don't want players to know what's going on behind the scenes... (bloody Asian MMOs).

In the JP version however, the tooltip of internal release was changed from Critical Hit Rating to just "Critical", which I assume should of been localised as Critical Strike. Battle Litany is the same.

So no, those two skills do not affect Critical Hit Damage.

[pandabearcat]

Hi everyone,

Just an update on the sim - the reason numbers were slightly lower (~973 avg dps) was because I forgot to code in BotD actually affecting jump/ssd damage lol.

So with my dev build DRG are now simming 993.26 dps average, which seems about right (this was ilvl 180).

A new version isn't going to be released anytime soon, as I am working on documentation, which prompted a thorough code cleaning. It looks like I'm going to completely revamp the way skills are coded in, so whomp whomp. But in the future changes to skill builds should be much easier to implement, as well as testing theoretical dps as a result of hotfixes, etc.

PS: new stat weights

WD	1.0000
STR	0.1328
SKS	0.0195
CHR	0.0152
DET	0.0152

SKS is out of control! Either the sks formula provided is wrong or something is very different.

[HeavensSword]

A new version isn't going to be released anytime soon, as I am working on documentation, which prompted a thorough code cleaning. It looks like I'm going to completely revamp the way skills are coded in, so whomp whomp. But in the future changes to skill builds should be much easier to implement, as well as testing theoretical dps as a result of hotfixes, etc.

Cool. I have a UI coded and almost ready to go. Just need to make the appropriate calls to your DLL when you finish.

[SunnyHirose]

SKS is out of control! Either the sks formula provided is wrong or something is very different.

Probably something is very different. If you're doing what I think you're doing... unless you change something about the cutoff to scale with skillspeed, it's either going to have no effect when it doesn't squeeze in extra skills, or a huge effect when it does.

[enil]

SKS is out of control! Either the sks formula provided is wrong or something is very different.

SkS is not a very linear stat - is it more likely that at whatever gear level your sim is currently being run at that there is some sort of SkS breakpoint after adding x SkS (however your sim determines weightings)? i.e. the 2.4~ HT/PHEB breakpoint.
As you said running some more tests at seperate gear levels may identify the issue.

edit: the sim with the default rotation does not clip CT, i.e. skill usage of FT combo was about 2x as much as CT combo. changing the logic doing CT combo <12 seconds of CT buff instead of <4 seems to fix this resulting in a dps gain of 50~. The new rotation the SkS weighting becomes 0.0042 (approx 4x less than crit) the main issue was CT was not being clipped at all, leaving a large amount of downtime for CT and Disembowel (hence why SkS was so good, increasing uptime for both skills).

edit2: there's still something wrong with sim logic, total 13 F&C+WT vs 17 CT/FT. Optimal rotation should always have a 4th per combo. Will need to take a look.

[Ukitsu]

And where can we find this sim (the outdated one)?

[pandabearcat]

http://forum.square-enix.com/ffxiv/t...ulator-2-SimFF

[Dervy]

When I have the time I'll check out why there's a massive difference between our skillspeeds. There's got to be a difference in the way our models are calculating Skillspeed or DoTs. It's too much especially as skillspeed will clip DoTs more and more skillspeed we have.

Could someone try model:

H IDC4 P TTT4
H TTT4 P IDC4
H TTT4 P TTT4

On SimFF and tell me the DPS it gives?

[pandabearcat]

Hm, its actually really difficult to have a priority system do a straight rotation.

I am thinking of adding macros to allow users to track their own "states" in case they want to do more rigid rotations rather than priority. This wasn't really an issue in WoW because everything is proc based over there - here it is kinda lame that almost every class (aside from ranged) is completely static. At least it meant spreadsheets gave fairly accurate models.

Anyway the syntax would look something like this:

Conditional:

(p1.macro.test == 1)

Action:

{macro.test, macro.test.inc, macro.test.10, macro.test.dec}

macro.<id> : initialize
macro.<id>.inc : increment by 1
macro.<id>.dec : decrement by 1
macro.<id>.<#> : set macro to #

All macros would be integers. They are initialized to 0. You can create and set as many macros as you like (I suggest you do this in precast).

so for example (in the future) this rotation would be something like:

[p1.PRECAST]
(){macro.state_var}

[p1.GCD]
# still want to use 4th combo when we can
(p1.aura.enhanced_wheeling_thrust){wheeling_thrust}
(p1.aura.sharper_fang_and_claw){fang_and_claw}

# if state is 0, do first half first line
(p1.macro.state_var == 0){heavy_thrust, impulse_drive, disembowel, chaos_thrust, macro.state_var.inc}
# if state is 1, do second half first line
(p1.macro.state_var == 1){phlebotomize, true_thrust, vorpal_thrust, full_thrust, macro.state_var.inc}
# etc. etc, to reset after last line, it would look like
(p1.macro.state_var == 5){phlebotomize, true_thrust, vorpal_thrust, full_thrust, macro.state_var.0}

[pandabearcat]

Shrug! I don't know what the sim does.

I just punch in the numbers.

The sim ends when the last event occurs after the duration cutoff - the "true duration" is then recorded and the dps calculated. There is less of a hard cutoff of all or nothing.