Black Mage Sanctuary - A Guide to DPS

[HaroldSaxon]

It is. There is no difference in impact between spells.
So yes. Technically you can say that with longer spells, spell speed gives more benefit. However this is only in terms of S so unless you're giving me accurate potency per second per rotation values depending on spell speed in a way that includes procs and the fact that the time you spend in umbral ice is NOT dependent on spell speed, then for all intents and purposes the DPS calculator doesn't care about it and for any practical application SS doesn't care about the cast time.

Regarding Ice phase, it can allow you get into Fire phase quicker. In 2.x, we have T1+B1, using the B1 as filler if we needed a GCD due to a slow mana tick. Quite often, we wouldn't need the filler spell, with Thunder 1 sufficing.

If we get a slow mana tick now, we use B4+T1/B1 (or T1/B1+B4) as a filler if needed. B4 is a longer cast than T1. So immediately there are scenarios where you gain in ice phase.

On top of that, depending on Enochian timers, remaining thunder dot and thundercloud procs, B4+T1 (or T1+B4) would be improved with spellspeed and maybe even preferred to compared to just B4 when getting a quick mana tick. (depending on what part of the rotation you are in).

[Xenitan]

Regarding Ice phase, (spell speed) can allow you get into Fire phase quicker.

Not really. I think we can both agree that we're looking to get two ticks, each of which happen 3 seconds apart from each other. So we're guaranteed to be locked in for at least 3 to 6 seconds no matter how much spell speed you have.

(Even more math, Black Mage DPS is such a fascinating topic)

There are 3 "rotations" for Black Mages that I want to bring up.

1) You cast, in quick succession, Blizzard 3, Blizzard 4, Fire 3
2) You cast, in quick succession, Blizzard 3, Blizzard 4 but then have to wait a period of time before casting Fire 3
3) You cast, in quick succession, Blizzard 3, Blizzard 4, a filler spell and then Fire 3.

At some point during your Blizzard 4 cast you will know how long you will need to spend at a minimum in umbral ice on that 3-6 line.

To optimize DPS we want to understand when you should be doing either of the the three rotations. We understand rotation 1 is the highest DPS of them all and so from 3 to a point A on the 3-6 interval we cast rotation 1 where A is the point early enough to where you can get two ticks of mana even while chain casting your ice spells.

We understand that, given an infinitely small but still positive value e, At the point A+e Rotations 1 and 2 are indistinguishable. Given that Rotation 1 is greater than rotation 3, this yields the fact that Rotation 2 is greater than rotation 3 on some interval on this line. We need to find B the break even point where rotation 2 and 3 are equal.

This is actually easier to find than you'd think. When you cast Blizzard you are "assigned" the length of time you'll need to spend in ice. When you finish blizzard you are given the option of waiting W seconds or casting the filler spell. If you cast the filler spell, you will have spent a length of time longer than necessary in umbral ice equal to that spell's cast time minus W. The question you ask is if that value is worth the extra potency you'd get from the filler spell. In other words, we're considering that spell in two parts: the part you'd be obligated to wait anyway (where you produce 0 potency) and the extra time. We want to find the potency per second value of that spells potency over that spell's extra time. If it is equal to the expected potency per second of rotation 1, that's the break even point. It can lie anywhere from A to a value beyond 6 (that value meaning you never cast a filler).

Spell speed messes with all these values, obviously. But it doesn't do so at the same rate so we don't factor these values into the direct multiplier for DPS you get from spell speed. It can increase DPS in this way, but this way is much harder to quantify (and, in fact, we haven't). But it is only in this hard to quantify way that spell speed is "better" on longer partitions of time for rotations. What it doesn't do is affect how often you spend as little time as possible in Umbral ice. What it does do, if anything, is make you more likely to spend more time in Umbral Ice.

[HaroldSaxon]

Snip

You missed the point I was making. I was saying the minimum rotation we have to do in ice phase is B3, F3. If we get a slow mana tick, we are limited by time - but getting more spellspeed means that we can weave another spell in there as well as F3 (because you can "pre-cast" F3 just before the tick. Which is a dps increase because you are getting back into the phase quicker.

If you get lucky with your timings and get a quick mana tick, you are still going to be in ice phase longer than required in order to complete the single B4. If Thunder ends up being a DPS increase to use in that phase, then spellspeed is going to shorten your effective Ice Phase time.

Also, there is a 4th rotation if i'm being pedantic - B3, Filler, B4, F3.

In all of the rotation cases, if you get a quick mana tick, you are in the phase longer than you need to be for mana reasons. If you get a slow tick, then 3/4 will still gain because 3+2.5+1 is larger than 6. Puro quantified the DPS increase in his last spreadsheet, you can go take a look at how he did it.

And I would love to see the math regarding a B3, B4, Filler, F3 rotation with procs in the next fire phase vs B3, B4, F3 rotation

[StouterTaru]

And I would love to see the math regarding a B3, B4, Filler, F3 rotation with procs in the next fire phase vs B3, B4, F3 rotation

It's very tight to fight a proc into the next phase after using B4, the only time it's feasible is on Faust.

Ignoring procs, Thunder is about 2/3 of the dps of the entire rotation, so it beats waiting 1/3 of a GCD. That means a GCD of ~2.05 seconds before there's a chance that it can be the same dps occasionally.

Blizzard is around 55% of the full rotation dps, so it would beat out waiting for more than 45% of a GCD. That takes a GCD of ~1.95 seconds to rarely not lower dps.