Black Mage Sanctuary - A Guide to DPS

[Waliel]

Ravana DPS also relies heavily on raid DPS to skip phases.

[Xenitan]

The reason expressing potency changes as a % of some previous base would be to provide a common metric between classes.

Well if their base DPS is lower and Det is only used in first degree polynomials, then the common metric wouldn't be so common. If X det increases your potency coefficient by Y and this holds for all classes, then for two potency coefficients P and Q with P>Q we'd have that an increase of Y/P and Y/Q which wouldn't be equal suggesting that X det on P is less valuable than x det on Q.

Whereas if you just look at what each stat gives you as a product of DPS increase, we remove that problem entirely.

Unless by percentage you're referring to an actual value of 1/100 in the potency coefficient formula. It's weird to refer to it as that, but I guess it wouldn't be incorrect or anything.

[Dervy]

Well, Arisue_Neetsha over in JP land has DET scale similarly for all jobs, which to me, makes the most amount of sense. Why would other jobs scale CRT differently, or skillspeed? If DET truly is just a plain old multiplier, then all jobs will scale DET exactly the same.

Any inconsistencies are due to rounding errors, which, I still believe all decimals are truncated to the X decimal place.

I mean Monk, Machinist, Bard and Dragoon all use the same DET coefficient with my tests, so I highly doubt it's different for BLM/SMN.

[Xenitan]

I mean Monk, Machinist, Bard and Dragoon all use the same DET coefficient with my tests, so I highly doubt it's different for BLM/SMN.

I don't think it is either- we both have the same numbers either way. I think there'll be some confusion about what the stats actually *do* later down the road though, since we'd have a problem where we would later say Det scales downward the more det you have (which would be untrue) because you need more Det to get 1% higher in DPS. People said the same thing about crit too.

[Kenji1134]

So, its not broken weapon data, but here is some controlled Det change data for War using a high potency Flash.

STR WD Det Hate
710 74 218 3873
710 74 250 3887
710 74 268 3896
710 74 274 3898
710 74 281 3905
710 74 295 3912
710 74 304 3914
710 74 309 3921
710 74 327 3933
710 74 338 3935
710 74 345 3939
710 74 360 3949
710 74 370 3951
710 74 384 3956
710 74 390 3962

Using the 218 Det value of 3873 as a baseline, an extra 1% would be 3911.72, which occurs at 295 Det. 2% would be 3950.46 which is is pretty much 370 Det.
295-218 = 77, 370-295 = 75, okay so some cherry picked values work. The thing is if you start to look at more minor changes, of 20-30 Det rather than 70-80 Det at a time, the numbers are off.
For instance 390 to 360 is 30 Det, 3962/3949 is .3292%, which gives us 30/.3292 = 91.13 Det/%... This is turning into a bit of a headache =P

[LathusWHM]

Most BLM out there need to remember that our abilities are situational. Yes there is an optimal rotation... But that's not the only way we play 100%.

You need to learn how to time certain abilities with different phases erg. Pre planning movement fights and such.

Yes, this means we have to think. Yes it's harder then the 2.x rotation but it separates good from bad.


With that said. Yes I do think there need to be qol changes in the future.

[Dervy]

So, its not broken weapon data, but here is some controlled det change data for War using a high potency Flash.

STR WD Det Hate
710 74 218 3873
710 74 250 3887
710 74 268 3896
710 74 274 3898
710 74 281 3905
710 74 295 3912
710 74 304 3914
710 74 309 3921
710 74 327 3933
710 74 338 3935
710 74 345 3939
710 74 360 3949
710 74 370 3951
710 74 384 3956
710 74 390 3962

Using the 218 Det value of 3873 as a baseline, an extra 1% would be 3911.72, which occurs at 295 Det. 2% would be 3950.46 which is is pretty much 370 Det.

http://puu.sh/jMeDD/c883e5944e.png

Well, this makes a lot of sense if Hate is supposed to have 0 variation......
/s

This is why I hate this game sometimes, lmao.

[Kenji1134]

Well, Flash has 0 variation, which is why its nice to use for gathering this kind of data, if only on a tank.
Though when you start using smaller Det changes, it moves away from the 75-77 value, as you can see with how jagged the data set is when plotted.

There's also still minor rounding issues. For instance 360 Det produces 3949 hate using Flash... so does 362... So with that in mind one could argue that under the assumption that this is in fact a linear behavior, we should use the min and max values to compute a slope to minimize rounding errors.

So 3962/3873 = 2.298%, 390-218 = 172 Det, so our slope is 74.8477 or about 75 Det/% on WAR.
Now if we go back a few pages to my "nice denominators BLM formula", I had an average projected Det/% of 74.7, which matches rather nicely to the 74.8477 from WAR.

I would like to get more data of this form, or more specifically I will redo this general test with a different weapon to see if the ~75 Det/% value holds across different initial conditions, and for other classes as well.
If this kind of examination continues to produce ~75 for various classes w/ various initial conditions (WD and Main Stat), then we may have "definitively" cracked part of the potency formula.

Healer data would be great as well!

[Vateri]

To be honest, I don't believe in Crit. I only pick whatever Speed piece meet the accuracy check of A4S which is why I prefer the Eso 210 weapon over the Gordian one.

What are you going for with BiS? I'm also starting to lean towards more of a SS/Det build... Is the added SS really that worth it in 3.0?

[Flipside101]

Alright, so when is it OK/Not OK to use the Thundercloud proc? What's the minimum duration of the dot for proc to be a DPS gain against F4? Is it OK to use the proc in UI even though I'll clip the dot big time? Do I use it after F1 refresh only on the fresh enochian timer/ 2nd refresh timer?

I'm seeing a variety of answers when to use/not use the proc. So, really which is it?