Hi, I have been busy for the last week doing some moving, I should clarify that I wrote the simulator to see how different openers affected the overall DPS and how much of an impact Wildfire made on our overall damage, as well as things like using Reload when it's up, or lining it up with buffs, lining up Hypercharge with buffs or using it whenever it's up, etc. So it is very rough, I can improve it if you need it to model.
Turret: I didn't get the chance to test it, the turret was set to WD because just from playing it seemed like the turret hit around the same time as my auto attacks.
I used the formula for a regular potency skill like Split Shot for the turret damage.
1. I set the loop timer to 0.1 seconds, it should really be 0.01 seconds but python takes too long to run 10000+ simulations for a long period of time and I didn't want to spend 10 minutes waiting for it to run 10000x times.
2. I was just using 1 second as the animation delay, as in game 1 second accounting for latency was roughly the time to get off a buff in between GCDs.
3. c_gcd is the gcd, o_gcd was there to keep track of popping buffs.
4. When I read up on previous forum posts about how different buffs stacked, they mentioned it was multiplicative, so I just assumed it so.
5. DoT snapshots are handled by setting the potency tick with whatever buffs were active, and letting it run for 30 seconds, ticking every 3 seconds. The potency is calculated with skillspeed in mind.
As to how accurate the Dragoon formulas are; the auto attack potency seems to be different, the Dragoon formula gave me really high auto attacks compared to my in game auto attacks. The potency formula however seems to be roughly accurate; I didn't include the +/- 5% variance in damage but over 1000 weaponskills and turret damage logging through ACT, the calculated damage was close to the damage in game.