So for a while now I’ve been using a rotation as BLM that is different from the standard fire spam. It is a set of six rotations organized by a priority system (e.g. rotation 1 is performed if possible, rotation 2 is performed if rotation 1 cannot be performed, rotation 3 is performed if rotation 2 cannot be performed and so on).
If you just want to see the rotations click the link below or read the TL;DR right below all the links.
If you’re skeptical and need math go down below the TL;DR.
-----
BLM Rotation Video Guide - https://www.youtube.com/watch?v=Ds8tPTAXAfY
BLM Training Dummy Parse w/ my rotation – https://www.youtube.com/watch?v=wZ8rk6E9yzg
BLM Training Dummy Parse w/ Standard rotation – https://www.youtube.com/watch?v=RZL1ycT6ko4
BLM T6 Clear W/ my rotation - https://www.youtube.com/watch?v=H8g3XBUPrNk
BLM T7 Clear W/ my rotation - https://www.youtube.com/watch?v=4J-zGvtcKMU
BLM T8 Clear W/ my rotation - https://www.youtube.com/watch?v=BVyJ4vv8quU
(On T8 I save cooldowns for triple/double flaring the dreadnoughts)
BLM T9 Clear W/ my rotation – http://i.imgur.com/gmhJVRM.jpg
(It is a screenshot as I don’t have a good POV video at the moment)
Numbers are available w/ overlays on all the vids. Please square have mercy
-----
In case the Video Guide goes down here’s the six different rotations in order from first priority to sixth priority. This is basically the TL;DR.
Opening Rotation – Priority 1
Fire 3 -> Fire 1 -> Thunder ½ (depends on mana) -> (Be sure to use procs) -> Fire 1 -> Raging Strikes -> (Be sure to use procs) -> Fire 1 -> Fire 1 -> Hardcast flare -> (Be sure to use procs) -> Convert -> Fire 1 -> Swiftcast flare -> (Be sure to use procs) -> Transpose -> Blizzard 1 -> Thunder 2 ->
Rotation w/ Swiftcast + Raging Strikes + Convert up – Priority 2
Fire 3 -> Fire 1 -> Raging Strikes -> (Be sure to use procs) -> Fire 1 -> Fire 1 -> Fire 1 -> Fire 1 -> Hardcast flare -> (Be sure to use procs) -> Convert -> Fire 1 -> Swiftcast flare -> (Be sure to use procs) -> Transpose -> Blizzard 1 -> Thunder 2 ->
Rotation w/ only Swiftcast up – Priority 3
Fire 3 -> Fire 1 x5 -> Swiftcast Flare -> Transpose -> Blizzard 1 -> Thunder 2 ->
Rotation w/ a Firestarter proc on the second to last fire (i.e. the fourth one) – Priority 4
Fire 3 -> Fire 1 x3 -> Fire 1 number 4 is cast -> Start casting Fire 1 number 5 -> (Oh boy I have a proc!) -> Finish casting Fire 1 number 5 -> Hardcast flare -> Use proc -> Transpose -> Blizzard 1 -> Thunder 2 ->
Rotation when you have nothing up – Priority 5
Fire 3 -> Fire 1 x5 -> Blizzard 3 -> Thunder 2 -> (you can do blizzard 1 / scathe here if that’s your thing) ->
Rotation when you have to reapply damage in fights like Leviathan or T9 – Priority 6
Fire 3 -> Fire 1 -> Thunder 2 -> Fire 1 x3 -> (Use the highest priority rotation available)
-----
When I say BLM standard rotation I mean
Fire 3 -> (Raging Strikes ->) Fire 1 x5 -> (Swiftcast Flare -> Convert -> Fire 1 ->) Blizzard 3 -> Thunder 2 ->
Everyone has their twist on this “standard” rotation so don’t shoot me if you do something different. This is just a good benchmark for comparison between rotations without many flares and my rotation which uses a ton of single target flares.
----
Math + Reasoning for skeptics
Swiftcasting flare without convert and Hardcasting flare both seem like they should be DPS losses at first because you have to use transpose to get back to full mana and there’s a RNG period of time where you sit at 0 mana waiting for mana ticks so rightfully you should be skeptical.
However, flare is a boatload of potency with three stacks of AF. It’s 468 potency with the 80% AF buff. This is loads more than Fire 1 which sits at a measly 306 potency in AF by comparison. Even Fire 3 falls short at 432 potency.
Consequently, all of the six rotations I listed out attempt to fit in the maximum number of flares when Flare’s massive potency can be abused by minimizing casting downtime.
Using transpose after a swiftcast flare or after a firestarter proc when at 0 mana gives the caster approximately 2.5 seconds where they can receive mana ticks. Mana ticks occur on approximately 3.5s intervals which leaves us with at most only a second of downtime before resuming casting. This means that you spend no longer waiting for a transpose mana tick than you do a mana tick after casting Blizzard 3 -> Thunder 2, which is why I exclude mana tick wait time entirely in my calculations below.
From experience it seems as if on average you spend less time waiting for mana ticks after transposing than we do after casting Blizzard 3 and Thunder 2, but I honestly have no idea how to model this.
-----
This is the part with math + too much theorycrafting
For the sake of brevity I assume a standard global cooldown of 2.5 seconds. All spell speed does is make DPS increases and DPS losses larger, so a DPS gain in my calculations is still a dps gain with whatever spell speed you have, and the same goes for losses.
Starting with the opening rotation
Casting Fire 3 with no stacks of umbral ice on leaves you with only enough mana to cast Fire 1 four times. Casting Fire 1 an additional fifth time if you have the PIE buff lets you cast Blizzard 3, but there is a large chance you will have to wait for a mana tick to cast your thunder DoT.
Instead, we can use this extra mana in our AF (Astral Fire) portion of the rotation in order to check for a firestarter proc while also applying the thunder DoT early.
Going Fire 3 – Fire 1 – Thunder 2 – Fire 1 eliminates the 16% chance that we will have back to back firestarter procs. This means this cast order allows us to use 2 firestarter procs when we would only get 1 in 16% of our opening rotations.
The same logic applies for raging strikes. Going Fire 1 – Raging Strikes – Fire 1 immediately after that eliminates another 16% chance we get firestarter procs back to back.
Additionally popping raging strikes this late in the opening rotation nearly ensures that raging strikes is up for both flares, adding a whopping (468 x .2 =) 93.6 potency to each flare cast. Raging strikes only falls off just before the swiftcasted flare if we get nonstop firestarter + thundercloud procs. At that point we have gotten the value out of our raging strikes regardless.
The hardcasted flare comes out to 140 average potency per second with raging strikes on, which is only slightly less than a Fire 1 w/ AF and RS at 146.88 potency per second. This is a dps gain, however, due to the fact that in the standard BLM rotation – some of the raging strikes duration can be wasted on Blizzard 3 which has a measly 115.2 potency per second under raging strikes even with the spell speed buff from AF 3.
Finally, the fire cast after convert is optimal as it has a 40% chance to trigger a firestarter proc while under raging strikes, and a swiftcasted flare under raging strikes becomes a whopping 224.64 potency per second.
The rotation is optimal with an average number of procs. No procs = you have a bad day and the weaving is a noticeable DPS loss. There are scenarios where this rotation is outperformed by a standard opener of Fire 3 -> RS -> Fire 1 x4 -> Swiftcast Flare -> Fire 1 -> Blizzard 3 -> Thunder 2 -> Repeat and vice versa, but this rotation has the higher DPS ceiling.
---
The second rotation in my priority system is just the opening rotation without the thunder cast so all of my reasoning + math above proves it is optimal as well.
This rotation also includes 1 instance of weaving with raging strikes so it loses to the standard rotation with no procs, but beats the standard rotation on average.
The Standard BLM rotation with all cooldowns does 3429.4 potency over 25.5 seconds for 134.49 potency per second
This BLM rotation does 3898.2 potency over 29.5 seconds for 132.14 potency per second
From experience, the 2 potency per second loss with no procs is more than made up for by the times when the weaving does result in casting 2 firestarter procs.
As a sidenote, a nice aspect of the opener is that use of all the cooldowns at the start of the fight usually results in them coming up at the same time 3 minutes later, and 3 minutes after that, etc. so that you can consistently perform a very high and sustained burst rotation every 3 minutes into a fight.
---
The third priority rotation does
2631 potency over 23 seconds for 114.39 potency per second
The Standard BLM rotation when it has no cooldowns up does
2161 potency over 20.5 seconds for 105.41 potency per second
---
The fourth priority rotation does
3063 potency over 27 seconds for 113.4 potency per second
The standard rotation does
2593 potency over 23 seconds for 112.73 potency per second
Both of these calculations include 1 firestarter proc cast under AF3
---
The fifth priority rotation and sixth priority rotations are no different from the normal BLM rotation.
---
Thank you for taking the time to read this great wall of text and I hope you took something valuable away from it ^.^
Also if you still hate me, think I’m a bad BLM, etc. please at least give my rotation a spin or 2. I’ve been using it on second coil for months now. I'm sure there are plenty of people out there with better Personal Records on coil DPS, but I feel like the difference on a training dummy can speak for itself.
Black Mage Rotation Guide
Printable View
-
08-09-2014, 10:23 AMEric_WindthiefBlack Mage Rotation Guide
-
08-12-2014, 07:31 AMArgyle_Darkheart
Your calculations are wrong.
You're assuming you start each rotation from a UI3 F3, but that isn't the case. Every time you Transpose, you're re-entering Astral Fire with an Umbral Ice 2 Fire 3 (which is a 3.5s cast), but you're never calculating for this.
That is, each rotation should be starting from Astral Fire 3 Fire 1, and ending with the appropriate Umbral Ice 2/3 Fire 3.
For example, you have the Priority 4 rotation calculated as a slight DPS increase, but it's actually a DPS loss. The correct rotation would be:
[Astral Fire 3] F1 (2.5s) > F1 (2.5s) > F1 (2.5s) > F1 (2.5s) [Proc] > F1 (2.5s) > Hardcast Flare (4.0s) > Firestarter (2.5s) > Transpose > B1 (2.5s) > T2 (3.0s) > Umbral Ice 2 Fire 3 (3.5s) - 3087 potency / 28s = 110.25 potency/s
Versus:
[Astral Fire3] F1 (2.5s) > F1 (2.5s) > F1 (2.5s) > F1 (2.5s) [Proc] > F1 (2.5s) > Firestarter (2.5s) > Astral Fire 3 Blizzard 3 (2.5s) > T2 (3.0s) > Umbral Ice 3 Fire 3 (2.5s) - 2593 potency / 23s = 112.74 potency/s
All that said, I don't know why you parsed a DPS increase; I certainly haven't experienced any DPS increase (usually the opposite) while testing your rotations. -
08-12-2014, 08:57 AMXisin
i recognize you from ffxi arg, and if what my my buddy says is true then you are from darkhearts on ramuh :) I never thought of you to be the black mage type. Not that I really spoke to you but I was in competing linkshells of yours for years.
edit to make this relevant. I can't imagine the hardcasting of flare in a rotation is a good thing mainly due to raging strikes and well the speed of the phase transition. Not going to really debate on it though, and honestly you need parse numbers (nevermind this i see your transparent parses in the corner now) to prove anything... Parses on dummies are hogwash as a blm will actually parse higher in most fights than what they do on a dummy. (exceptions t7 sometimes, depends on waiting..., t9 cause of waiting, ifrit ex due to freaking waiting, and sometimes titan ex depends on waiting lol)
I'm no mathematician though, I've always done what feels best and really haven't followed a posted rotation. (cause theres no such thing.) -
08-12-2014, 10:28 AMArgyle_DarkheartHey; long time no see. I still main a tank job (WAR, in this case), but I play all classes/jobs.Quote:
Originally Posted by Xisin
On topic, I would like to add that hardcasting a Flare can be useful in a burst scenario (i.e. to fit as much damage into a single Astral phase as possible), but I don't think there's any single target scenario where it equates to more sustained DPS (on average). -
08-13-2014, 01:18 PMEric_WindthiefREI did make a math mistake on the fire 3 casting time. I'll adjust math later. To be totally honest, I'm not sure why my experience is so radically different from anyone else testing the rotation. I'm currently checking to see if I actually am just under performing using the standard rotation. If that is not the case I will then review all of my math and revise it. I plan to try and simulate the effect of waiting for mana ticks in future dps comparisons.Quote:
Originally Posted by Argyle_Darkheart
The only reasons I can think of for why my personal experience using this rotation differs so widely from the rest of the community is that I wrote my guide wrong and it is misleading people in some way or I am doing something wrong myself when using other more conventional rotations. I certainly don't have a small sample size for comparison though, I've been using this rotation since 2.2 and found it preferable again and again.
Thanks for the feedback. -
08-14-2014, 06:47 AMSkeith-Adeline
Sorry, but, I'm completely at a loss. @. @
My numbers are a 405 constant but... would this rotation make a noticeable difference? Seems like a lot to learn, err, releearn/
I'm a 106 blm with the choker and other ring at 90.