Some interesting numbers on 4.0 BLM

[soslinky]

What are your openers and rotations exactly? I'm getting vastly different numbers than you.

Opener:
B3 Pre>Eno>Swift>B4>LL>T3>F3>Triple>F4>Sharp>F4>F4>F1>F4>F4>F4>F3P>Convert>F4>F4

4.0 AF to UI:
F4>F4>F4>F1>F4>F4>F4 >>> B3>Foul>T3>B4

3.0 AF to UI:
F4>F4>F1>F4>F4 >>> B3>Foul>T3

That's the opener and rotation I used for the spreadsheet.

[soslinky]

@Galvuu
The table uses the nerfed F4. It's 491.4 potency including AF3 and Enoch.

I don't have the energy to manually compute this so I used the DPS calc spreadsheet. Author claims it's updated and so far it looks like it is.

@Rakhabit
How are you computing for your numbers? What results did you get?

[Galvuu]

@Galvuu
The table uses the nerfed F4. It's 491.4 potency including AF3 and Enoch.

I don't have the energy to manually compute this so I used the DPS calc spreadsheet. Author claims it's updated and so far it looks like it is.

@Rakhabit
How are you computing for your numbers? What results did you get?

You can ignore Enochian since whichever comes ahead without it also comes ahead with it.
This is exactly the same thing I used for the main rotation. Huh, one of us is wrong. And I hope it's me xD
Is Excel lying to me or something.

Wait, losing a F4 has to be a pps loss, because the "fraction" of your overall rotation that's spent in AF is smaller, and the UI cycle has lower average pps than the AF one (the more overall time is spent on UI, the worse it is).
The reason why 3.x pulls ahead is because removing B4 from the UI phase makes it lose 1/3 of it's duration (and the third with the least potency, too, making the average UI potency for 3.X way higher).

[soslinky]

So am I correct if I understand it like this:
On a dummy 4.0 rotation is better as more F4=more dps, as one would think.
In an actual fight you should use an UH opener but then go back to 3.x rotation, cause, let's face it, an actual fight isn't a dummy and you will be forced to move. As soon as you lose 1 F4 in 4.0 rotation due to movement/mechanics it falls behind 3.x rotation.

On a side note, did anyone try not using B4 in normal rotation and only using it in combination with Triplecast?

My latest parse on Susano had me using only 6 B4s and I only used it when:
A) I got a slow mana tick and foul wasn't up, and when
B) I had the purple marker on my head so I knew I had to move and would miss out in completing the 6 F4s.

I cast a total of 82 Fire4s so on average that's 15-18 AF windows (this is pure approximation).

I did not purposely hold off on Triplecast and used it on CD and proceeded to 3x F4. From what I've surmised, if I think I can get away with 6 Fire4s I go ahead and B4. If I have T3P up during Umbral I weave a Swiftcast to B4 just so I can get back to AF that much quicker.

[soslinky]

snip.

I don't know what to tell you. I was curious as to how big an impact losing one out of the 6 F4s per AF cycle was to 4.0 rotation so I took a single one out and the PPS came out higher (by 1 point). O_O
I hope to our lord and saviour Ex-Death that the spreadsheet is correct though.

What are your thoughts on prolonged PPS calculations? If the percentages I've seen from +2 AF versus +4 AF are valid then maybe we've been doing the math on too small a window and we need to parse/math a whole fight duration to get a better understanding of this.

[Galvuu]

snip

No, it's good that you question the results and that we discuss it. I mess up a lot too.
The spreadsheet does have a lot of weird multipliers- maybe something went wrong when we went to SB?
I'm looking at my (painfully simple) tables and I cannot see where they're wrong... and I want them to be wrong, too.

You only need to compute the pps of a full "rotation" (AF+UI cycles). Steady dps just repeats these ad nauseam, so you just find an arbitrary least common multiple for (opener)+K*(rotations), for some integer K, and then compare those when normalised for that overall length of time.

EDIT: Note that opener+2AF cycles for 3.x is way less "casting time" than opener+3AF cycles for 4.0. For the same time window, you can probably fit almost 3 AF cycles from 3.X VS 2 for 4.0. That may be another factor causing discrepancy (though the pps per cycle difference is still very odd- since they're normalised in respect to time, they shouldn't be that different after the opener).

[Rakhabit]

@Rakhabit
How are you computing for your numbers? What results did you get?

Here's my opener(199.99 PPS): http://i.imgur.com/CpMrlr2.png

Here's my opener with 2 cycles of 4.x Rotation(167.99 PPS): http://i.imgur.com/06fZUy6.png

Here's my opener with 2 cycles of 3.x Rotation(172.64 PPS): http://i.imgur.com/16H1vOO.png

I don't think I made any mistakes entering these, but I could have. I still haven't slept.

[soslinky]

snip.

Tired of looking over this spreadsheet so I ran an actual test.

I ran two runs on a dummy. Same opener, took out any T3P/F3P procs to maintain consistency, cast foul during UI only (if it wasn't up I used it on the next UI cycle - 3.0).
I capped the duration of each run at 2minutes, if I was still in AF by then I kept it going right until B3 then stopped recording. Sharp, Triple, Swift, Leylines were used exclusively on opener.

Here are the results:
(2:08min) 4.0 = 4,206.20 / DirectHit(12%) / Crit(20%) / CritDirect(4%) / All UI contained a Foul / Total of 51 Spells
(2:04min) 3.0 = 4,024.01 / DirectHit(10%) / Crit(18%) / CritDirect(3%) / One UI missed a Foul due to it not being up / Had perfect mana ticks so no delay in spell queueing / Total of 50 Spells

That's 182.19 DPS of a difference. I don't know how the 1 or 2% from the DH/Crit/DHCrit variance would pan out in DPS (would depend on what spell actually proc'd, less of an impact if it was the B3). In the end, this result is still disappointing. We went through a rework and we only came out higher in DPS by only that much.

[soslinky]

Here's my opener(199.99 PPS): http://i.imgur.com/CpMrlr2.png

Here's my opener with 2 cycles of 4.x Rotation(167.99 PPS): http://i.imgur.com/06fZUy6.png

Here's my opener with 2 cycles of 3.x Rotation(172.64 PPS): http://i.imgur.com/16H1vOO.png

I don't think I made any mistakes entering these, but I could have. I still haven't slept.

I can't seem to open Imgur at all. Been like this all day. Would you mind uploading from a different photo hosting site?

[Rakhabit]

I can't seem to open Imgur at all. Been like this all day. Would you mind uploading from a different photo hosting site?

Sure thing.

Opener: https://ibb.co/iY65FF
3.x: https://ibb.co/cibSUa
4.x: https://ibb.co/cbfnUa