A DPS Simulator 2 - SimFF

[pandabearcat]

I mean theres no reason I can't add war/pld dps.

Adding classes is probably the easiest part of the entire simulator (I designed it that way).

But no class aside from baseline drg (no cross class abilities yet...sry) is going to be done until the core functionality of the simulator is complete.

Also if you want a head start to scripting your own rotations...start looking at simcraft documentation. The grammar that will be used is going to be very similar.

[Dervy]

Straight Shot is 10% flat added on top of your Critical Hit %. IR is the same.

[pandabearcat]

I added Dervy's stats in. Something is wrong with the autoattack formula.

Unless DRG are really doing 20k autoattacks in which case bravo!

Dervy if you could check those numbers (AA aside) on a dummy to see if they are off significantly that would be good.

Furthermore, I am unsure of the actual damage contribution of Disembowel. Is 10% piercing resist reduction 10% damage increase? It seems like it would not be that simple.

EDIT: I fixed the autoattack bug. But now autoattacks seem very weak.

[Dervy]

I'm sure if Sunny can elaborate a bit more, but from his testing, yeah, it's literally 10%.

You sure it's weak? It's based all off my recorded Data, and my AA Formula, surprisingly, has a much lower variation vs recorded damage, than my WS Formula.

((WD*0.0496151+1)*(AP*0.0595451+3.3394725)*(DET*0.0014344+1) is my latest one as I'm filling it with more low-AP, low-DET parses. I just did another parse with the stats I sent you a few pages back:

AA_DMG: 57.22
AP: 646
DET: 408

Recorded Min: 242
Recorded Max: 267
Recorded Average: 254.6

Formula Min: 241.7
Formula Max: 267.1
Formula Average: 254.4

[pandabearcat]

This is the formula you gave me:

(AA_DMG*0.00388814+1)*(STR*0.0717908+3.5421066)*(DRT*0.0014237+1)

And it is giving 96 damage baseline for autoattacks.

[SunnyHirose]

Furthermore, I am unsure of the actual damage contribution of Disembowel. Is 10% piercing resist reduction 10% damage increase? It seems like it would not be that simple.

It is that simple. On the backend they could just make it a 10% buff or they could do it more circuitously as:

damage buff = (200 - (damage type resistance))/100

Either way, I've tested it on NIN and DRG as being 10% and definitely not 11.1% (using Aeolian Edge and Full Thrust to really exaggerate the difference).

I've heard anecdotally that Shiva's debuffs (which really do drop your resistance stat) suggest it works the more complicated way.

[Dervy]

Oh. The issue is 0.003. Remove one of the 0s, lmfao. But yeah, use the newer model I gave you I gave you, also use AA_DMG, not WD, so it'll be:

((AA_DMG*0.0496151+1)*(AP*0.0595451+3.3394725)*(DET*0.0014344+1).

Sorry for being a massive dumbo here :P Been working with numbers all day.

[pandabearcat]

I've put in the new formula. I would appreciate it Dervy if you would check the numbers of the skills and AA to see if they are reasonable.

If any other dragoons would like to help test the accuracy of the sim in relation to their stats, please post here and I will PM you a pastebin of a sample parse.

In the meantime work on the engine has been "completed", so the next project is the parser.

Once a reasonable portion of the grammar is complete, I will release the sim for people to try out and test.

A few more notable things that need to be done however, but they are somewhat lower priority.

1. Actual spell animation delays. Ignoring Leg Sweep (since its weird), I've currently set GCD skill delay to 1.2 sec, and jump delay to 1.5 sec. This is obviously not really correct, so it would be appreciated if someone would time these. I asked before on the DRG thread but people kind of didn't pay attention to it.

2. I need to know the general damage "spread" % (instead of an attack always doing the same damage) so we can do more realistic sims. Statistically it won't make much difference in the average but it would obviously affect the deviations of data and can show general lower and upper bounds of damage more easily.

[Viridiana]

2. I need to know the general damage "spread" % (instead of an attack always doing the same damage) so we can do more realistic sims. Statistically it won't make much difference in the average but it would obviously affect the deviations of data and can show general lower and upper bounds of damage more easily.

Can't help much with the animation delays, but shouldn't this still be 21/19=1.1052? That is, plus or minus 5% with rounding preventing perfect ratios? At least, that's the max/min I've always observed, and the few times I couldn't get the base numbers to actually display a 21/19 split, if I accounted for fractions I could almost always get it.

[pandabearcat]

Wow its not a uniform distribution? Well thats super lame.

Are you saying the STD is 5% or are you saying the full curve is +-5% (so STD would be closer to ~1.7%)?

Is it hard bounded at +-5% or is there a super slim chance of doing a billion damage?