Fire Mages are alive again (Updated for 6.2.3)

Guides and discussions of all things specific to the Pyromaniacs of Azeroth.
yungg
Posts: 62
Joined: Mon Dec 01, 2014 12:54 pm

Re: Fire Mages are alive again (Updated for 6.2.3)

Unread postby yungg Sat Feb 13, 2016 11:27 pm

so am i getting this right, as long as you have gcd cap and start from 18y away, you can always get IB before ignite ticks?
Alzer
Posts: 50
Joined: Tue Jan 27, 2015 3:29 pm

Re: Fire Mages are alive again (Updated for 6.2.3)

Unread postby Alzer Sat Feb 13, 2016 11:58 pm

so am i getting this right, as long as you have gcd cap and start from 18y away, you can always get IB before ignite ticks?
You will get it before the 5th ignite tick yes.
yungg
Posts: 62
Joined: Mon Dec 01, 2014 12:54 pm

Re: Fire Mages are alive again (Updated for 6.2.3)

Unread postby yungg Sun Feb 14, 2016 12:17 am

aight cool.
Alzer
Posts: 50
Joined: Tue Jan 27, 2015 3:29 pm

Re: Fire Mages are alive again (Updated for 6.2.3)

Unread postby Alzer Sun Feb 14, 2016 12:19 am

That doesn't necessarily mean it's the best thing to do however.
yungg
Posts: 62
Joined: Mon Dec 01, 2014 12:54 pm

Re: Fire Mages are alive again (Updated for 6.2.3)

Unread postby yungg Sun Feb 14, 2016 12:29 am

Is that because you lose a blast wave on crystal? or other reasons?

Edit: ah, never mind, I reread the post. Interesting.
skiz
Posts: 182
Joined: Fri Aug 22, 2014 12:27 pm

Re: Fire Mages are alive again (Updated for 6.2.3)

Unread postby skiz Sun Feb 14, 2016 6:37 am

I have 194 haste and getting ib before 5th ignite tick is no problem. 810 cap with my (troll) profile might be because robomage doesn't combustion correctly and the extra ~600 haste helps with ib multistriking.
This. Robomage:
IB
wait gcd
Combustion/IB (on the same global)
We want this:
IB/Combustion (on the same global)
wait gcd
IB
The gain of a hastecap is simply to fit in your rotation before ignite ticks and to combustion right after it. GCD cap helps for that. Any more haste is probably wasted. As a non troll this 810 maybe even 900 haste is really a huge gain. 810 is probably enough haste because of : 9% + 5% + 6% haste randomly proccing on pyros. However, with archimonde's staff in mind you can easily get 810+ haste as a troll, too. But without it I would probably dodge every hasterating as a troll.
That doesn't necessarily mean it's the best thing to do however.
A)
(Fireball -> Pyroblast) -> Pyroblast -> Pyroblast -> Infernoblast-> Combustion -> Infernoblast

B)
(Fireball -> Pyroblast) -> Pyroblast -> Pyroblast -> wait if you're to far away -> Combustion -> Infernoblast

Pretty sure that version B) is superior to A) if you do have enough haste to build up ignite before it ticks away and you're running into your prismatic crystall. If you do not run B sucks. Executing that you'll have the problem of hitting combustion fast enough before your last pyro multistrikes. You do have the same problem with version A) and Inferno Blast multistrikes.
Infernoblast has no traveltime so this works:
/castsequence Inferno Blast, Combustion

You could do the same with Pyroblast and Combustion for B). But it will come down to : Is your latency higher than the travel time of your last pyro? Which is literally: Are you able to get the traveltime to mostly 0.
Last edited by skiz on Sun Feb 14, 2016 8:06 am, edited 1 time in total.
yungg
Posts: 62
Joined: Mon Dec 01, 2014 12:54 pm

Re: Fire Mages are alive again (Updated for 6.2.3)

Unread postby yungg Sun Feb 14, 2016 6:50 am

and B is supposedly better because it builds a bigger combustion? or is it because you get combustion off earlier? Because I can't see why we wouldnt want to spread the ignite twice, if the first inferno blast wont decrease our combustion.
Alzer
Posts: 50
Joined: Tue Jan 27, 2015 3:29 pm

Re: Fire Mages are alive again (Updated for 6.2.3)

Unread postby Alzer Sun Feb 14, 2016 7:08 am

Since I have been very curious about IB vs. no IB myself and several people have asked me about it I'm gonna make an attempt to simulate it properly.

I've been doing IB before combustion most of the time but lately I've been exploring ignite tick timers more and come to the conclusion the IB might not be the best option even though it's very clearly a dps increase in SimulationCraft.

As skiz mentioned, these two openers (FB->pyro->FB->pyro->pyro->pyro and FB->pyro->FB->pyro->pyro->pyro->IB) play out differently. Without IB you wanna move towards the crystal as you are buiding your ignite, lowering the time between your first FB+pyro hit and your last pyro hit, effectively squeezing your ignite into a shorter window and thus making it bigger. With IB this is irrelevant since your last spell is instant and it's impossible to squeeze your stuff into a shorter window, here you simply wanna stand at 18+ yards all the time.

What it really comes down to is 2 fireballs + 3 pyroblasts before the 4th ignite tick compared to 2 fireballs + 3 pyroblasts + inferno blast before the 5th tick.

The problem with this is that robomage does not run towards crystal and thus when comparing the two rotations in SimC you are essentially comparing 2 fireballs + 3 pyroblasts before the 5th tick vs 2 fireballs + 3 pyroblats + inferno blast before the 5th tick. Naturally the IB rotation comes out ahead.

The real issue here is there appears to be no way to either tell robomage to run towards crystal or to alter the flight time of the last pyro (to emulate being at point blank).

So I'm gonna try to squeeze in the opener before the 4th ignite tick in SimC by setting the GCD and flight time of pyroblast to 0 and 600 respectively (don't go crazy pls). This means it takes 0.033 seconds for pyroblast to hit the crystal.

Code: Select all

override.spell_data=spell.11366.prj_speed=600 override.spell_data=spell.11366.gcd=0
Obviously this poses alot of problems that need to be taken care of. We need to introduce a GCD and a flight time for the earlier pyroblasts another way.

Here is how the burst looks:

Code: Select all

actions.combust_sequence+=/sequence,if=buff.bloodlust.up,name=precombustion:fireball:wait,sec=0.700:pyroblast:wait,sec=0.300:fireball:wait,sec=0.700:pyroblast:wait,sec=0.800:pyroblast:wait,sec=0.5:pyroblast:wait,sec=0.2:combustion:wait,sec=0.8
Instead of writing down what this means I decided to show it in a picture instead: http://i.imgur.com/1G58Rwq.png Apologies for poor paint skills and no sense of symmetry.
The purpose of all these weird wait segments is to emulate a GCD and a flight time.

Due to how this rotation greatly affects the normal rotation, we need to do the same thing with the IB rotation to be able to compare them, that one looks like this:

Code: Select all

actions.combust_sequence+=/sequence,if=buff.bloodlust.up,name=precombustion:fireball:wait,sec=0.700:pyroblast:wait,sec=0.300:fireball:wait,sec=0.700:pyroblast:wait,sec=0.300:pyroblast:wait,sec=1:pyroblast:wait,sec=1:inferno_blast
No picture for this one, the last one took much longer than I would like to admit.

First off, both of these profiles need to be at crit cap, otherwise robomage will start hardcasting pyroblast once something doesn't crit. The default profile only casts prismatic crystal when sandman's pouch has procced so this is already taken care off. However, due to the interaction of a zero flight time pyroblast with fireball, robomage will sometimes cast crystal without both Pyroblast! and HU. To prevent this I replaced fireball with scorch (scorch also has no flight time). I added

Code: Select all

actions.single_target+=/scorch,if=time<20
to the normal single target rotation.

To confirm nothing crazy was going on I did a log=1 in SimC.
This is a log from SimC with all the irrelevant stuff edited out: http://pastebin.com/pVqeF3g0

Alright, assuming the crazy pyroblasts affects the normal rotation for both profiles equally we can start comparing these two rotations.
Image
40s, 1 target, 0 multistrike, 1100 haste

Image
40s 2 targets, 0 multistrike, 1100 haste

They are both run with 0 multistrike to avoid any weird multistrikes with instant flight time that interfere with ignite timers.

So, are these results reasonable? In the no IB profile we expect see a higher combustion (due to running towards the crystal), a higher crystal blast wave damage (due to not having to spend a GCD on IB) and a lower ignite (from not double spreading it (T17-style) to the boss). This is indeed what we see:
Image

Keep in mind these numbers are not tick damage or anything like that, it's simply the contribution to the overall dps.

To further confirm the accuracy of this simulation we compare it to the default profiles:
Image
40s, 1 target, 0 multistrike, 1100 haste, default profiles

The most important thing here is noticing that the IB default profile has the same dps as our modified IB profile, even though the latter has a pyroblast with no GCD and a higher velocity than a jet plane. This indicates that the wait segments fairly accurately emulated a GCD and a flight time.

I was gonna do some sort of plot/reforge plot/discussion about haste but this took quite some time and if I ever want to get some sleep I'm gonna have to stop here. Just don't do this with low haste unless you are a troll.
Last edited by Alzer on Sun Feb 14, 2016 7:26 am, edited 1 time in total.
yungg
Posts: 62
Joined: Mon Dec 01, 2014 12:54 pm

Re: Fire Mages are alive again (Updated for 6.2.3)

Unread postby yungg Sun Feb 14, 2016 7:18 am

Nice post, definitely answered my questions. Thanks for putting in this work!
EgonS
Posts: 4
Joined: Wed Aug 27, 2014 4:56 pm

Re: Fire Mages are alive again (Updated for 6.2.3)

Unread postby EgonS Sun Feb 14, 2016 3:17 pm

Does anyone have an issue where combustion is not spreading from PC on old world bosses? I've has it happen on Sha and Nalak now

Edit: Thank you to all that worked on updated the new fire APL
skiz
Posts: 182
Joined: Fri Aug 22, 2014 12:27 pm

Re: Fire Mages are alive again (Updated for 6.2.3)

Unread postby skiz Mon Feb 15, 2016 1:38 am

There is an issue ingame with ignite that is worth to be mentioned:

This log:

Code: Select all

https://www.warcraftlogs.com/reports/VCwnAW6tD8gpNH4z/#view=events&type=damage-done&pins=0%24Separate%24%23244F4B%24casts%240%240.0.0.Any%2483617373.0.0.Mage%24true%240.0.0.Any%24true%2411129|11366^0%24Separate%24%23909049%24auras-gained%24-1%240.0.0.Any%2483617373.0.0.Mage%24true%240.0.0.Any%24true%24186170&source=1&hostility=1&fight=26
00:00:07.811 Skizzaa Ignite Prismatic Crystal (Skizzaa) Tick 20164
00:00:07.811 Skizzaa Pyroblast Prismatic Crystal (Skizzaa) *159647*
00:00:08.636 Skizzaa casts Pyroblast on Prismatic Crystal (Skizzaa)

00:00:08.778 Skizzaa Pyroblast Prismatic Crystal (Skizzaa) *135206* <---- this thing
00:00:08.807 Skizzaa Ignite Prismatic Crystal (Skizzaa) Tick 27416

00:00:08.949 Skizzaa casts Combustion on Prismatic Crystal (Skizzaa)
00:00:09.120 Skizzaa Pyroblast Prismatic Crystal (Skizzaa) *44188* (Multistrike)
00:00:09.408 Skizzaa Pyroblast Prismatic Crystal (Skizzaa) Tick *22756*
00:00:09.408 Skizzaa Pyroblast Prismatic Crystal (Skizzaa) Tick *6827* (Multistrike)
00:00:09.835 Skizzaa Ignite Prismatic Crystal (Skizzaa) Tick 34772

I do have 2 pyros hitting between 1 ignite intervall. And 1 pyro multistrike between the other. On both intervalls ignite grows by roughly 7k. Pretty sure that the pyroblast before ignite does his contribution to ignite after the tick. Simply a pyro multistrike is unable to grow an ignite by the same amount as two pyroblast hits are able to do. Combustion goes off with the pyro inside of it. This seems a little bit buggy, but there seems to be no difference between getting a spell of slightly before ignite or slightly behind it.

Code: Select all

https://www.warcraftlogs.com/reports/VCwnAW6tD8gpNH4z/#view=events&type=damage-done&pins=0%24Separate%24%23244F4B%24casts%240%240.0.0.Any%2483617373.0.0.Mage%24true%240.0.0.Any%24true%2411129|11366^0%24Separate%24%23909049%24auras-gained%24-1%240.0.0.Any%2483617373.0.0.Mage%24true%240.0.0.Any%24true%24186170&source=1&hostility=1&fight=56
The pyroblast after ignite lets the ignite grow to 39k - again this happened after the ignite tick even though the spell has hitted before it. The spells hitting after the ignite tick let it grow to 42k. Combustion gets off with a 42k ignite.

Lets say I can reproduce that issuse. Then I do get the following things:
- ignite is a little bit forgiving regarding to correct haste values.
- Double IB after running into PC may increase my ignite - even if it may sound decent to not run into the boss with it.
- Not running into the Prismatic crystal - no matter if double Ib or not - will result in a lower ignite. This kind of violates the idea of ignite as a bucket. But if running towards the crystal is necessary to force this situation....

Analyzing double IB vs no double IB:
- I do lose 1 GCD on the PC. This just matters as non troll with coldsnap beeing ready. Otherwise the gcd is spent on a spell that does roughly 60k damage.
- I do lose 1 tick of combustion on the PC because of an earlier one.
- I gain 4-5 ticks of ignite
- On which case do I get a bigger combustion?

For my gear setup it comes down to: When does my 4 piece set proccs:
- Pre building: Run into pc, followed by double IB: ~36-45k ignite (roughly jumps up by 10% after IB.)
- on the first cast: I'm not sure about that one.
- after it: does not work, double ignite just produces an equally / worse ignite. ~30k ignite
- I've used a specific gear set while testing it that has 15% haste and is critcapped.
- Ignite during BL is just slightly higher than that 45k ignite

Some more things:
- cant get traveltime of pyro below 0.2-0.3ms even if cuddling with PC.
- Different haste setup, different problems.
- seeing double ib beeing superior is something that seems to be covered by logs.
- fire 4piece feels like a 25% chance to have a 30% higher combustion.
User avatar
Komma
Administrator
Posts: 1486
Joined: Wed May 28, 2014 7:37 pm

Re: Fire Mages are alive again (Updated for 6.2.3)

Unread postby Komma Wed Feb 17, 2016 9:05 am

In case you folks aren't aware, analyzing IB vs not IB pre-ignite in sims isn't really worth it, due to multistrike's modelling. SimulationCraft core engine does not model multistrike's 333/666 ms delay.
Admin of Altered Time.

Have an issue with the website or moderation? Send me a PM!
User avatar
Curnivore
Posts: 827
Joined: Mon Sep 28, 2015 9:26 am

Re: Fire Mages are alive again (Updated for 6.2.3)

Unread postby Curnivore Wed Feb 17, 2016 10:52 am

SimulationCraft core engine does not model multistrike's 333/666 ms delay.
Got a URL explaining that problem?
User avatar
Komma
Administrator
Posts: 1486
Joined: Wed May 28, 2014 7:37 pm

Re: Fire Mages are alive again (Updated for 6.2.3)

Unread postby Komma Wed Feb 17, 2016 7:19 pm

There isn't a problem. It's just a design decision/limitation in the core engine. We're basically the only class that is affected by it. It does however mean that we should take it into account.
Admin of Altered Time.

Have an issue with the website or moderation? Send me a PM!
Searix
Posts: 233
Joined: Mon Feb 09, 2015 10:07 pm

Re: Fire Mages are alive again (Updated for 6.2.3)

Unread postby Searix Fri Feb 19, 2016 2:23 am

As skiz mentioned, these two openers (FB->pyro->FB->pyro->pyro->pyro and FB->pyro->FB->pyro->pyro->pyro->IB) play out differently. Without IB you wanna move towards the crystal as you are buiding your ignite, lowering the time between your first FB+pyro hit and your last pyro hit, effectively squeezing your ignite into a shorter window and thus making it bigger. With IB this is irrelevant since your last spell is instant and it's impossible to squeeze your stuff into a shorter window, here you simply wanna stand at 18+ yards all the time.
I'm assuming you mean run close enough to crystal to (almost) instantly hit it?

Also can we get some numbers for how much of a difference running in the last pyro is for overall dps? It's often a trade-off for fights where you can screw up the raid running in that close vs. staying out. If running is only like 1% overall dps it's probably not worth it most of the time
veezy
Posts: 2
Joined: Sat Feb 20, 2016 2:37 pm

Re: Fire Mages are alive again (Updated for 6.2.3)

Unread postby veezy Sat Feb 20, 2016 2:43 pm

Quick question for you guys. Does it matter much if in the burst phase I do 2 x blast waves before 2x dragon's breath?
Guidancea
Posts: 11
Joined: Thu Nov 26, 2015 5:56 pm

Re: Fire Mages are alive again (Updated for 6.2.3)

Unread postby Guidancea Sat Feb 20, 2016 3:08 pm

Quick question for you guys. Does it matter much if in the burst phase I do 2 x blast waves before 2x dragon's breath?
Generally DB deals more damage than BW.
Mage in <W U T> on Shattered-Hand.
twitter.com/smokym8
veezy
Posts: 2
Joined: Sat Feb 20, 2016 2:37 pm

Re: Fire Mages are alive again (Updated for 6.2.3)

Unread postby veezy Sat Feb 20, 2016 3:31 pm

Quick question for you guys. Does it matter much if in the burst phase I do 2 x blast waves before 2x dragon's breath?
Generally DB deals more damage than BW.
Yeah I understand that but would it affect anything if I can get both of them in on the crystal? (same procs/buffs considered)

Just curious as doing 2x Blast wave first allows a little more time to set up for a DB x2 with a proper angle. I am just wondering in case I am in a bad position and need a couple extra seconds to set up the perfect DB angle and I need to BWx2 first so my DBx2 can hit as much things as possible.
Skodmage
Posts: 1
Joined: Sat Feb 20, 2016 8:41 pm

Re: Fire Mages are alive again (Updated for 6.2.3)

Unread postby Skodmage Sat Feb 20, 2016 8:58 pm

quick question guys.- That 47% crit is that with or without raid buff? i mean fully critbuffed is with that too?:)
Petfood101
Posts: 41
Joined: Sat May 31, 2014 9:08 pm

Re: Fire Mages are alive again (Updated for 6.2.3)

Unread postby Petfood101 Sat Feb 20, 2016 10:21 pm

I have 194 haste and getting ib before 5th ignite tick is no problem. 810 cap with my (troll) profile might be because robomage doesn't combustion correctly and the extra ~600 haste helps with ib multistriking.
This. Robomage:
IB
wait gcd
Combustion/IB (on the same global)
We want this:
IB/Combustion (on the same global)
wait gcd
IB
The gain of a hastecap is simply to fit in your rotation before ignite ticks and to combustion right after it. GCD cap helps for that. Any more haste is probably wasted. As a non troll this 810 maybe even 900 haste is really a huge gain. 810 is probably enough haste because of : 9% + 5% + 6% haste randomly proccing on pyros. However, with archimonde's staff in mind you can easily get 810+ haste as a troll, too. But without it I would probably dodge every hasterating as a troll.
That doesn't necessarily mean it's the best thing to do however.
A)
(Fireball -> Pyroblast) -> Pyroblast -> Pyroblast -> Infernoblast-> Combustion -> Infernoblast

B)
(Fireball -> Pyroblast) -> Pyroblast -> Pyroblast -> wait if you're to far away -> Combustion -> Infernoblast

Pretty sure that version B) is superior to A) if you do have enough haste to build up ignite before it ticks away and you're running into your prismatic crystall. If you do not run B sucks. Executing that you'll have the problem of hitting combustion fast enough before your last pyro multistrikes. You do have the same problem with version A) and Inferno Blast multistrikes.
Infernoblast has no traveltime so this works:
/castsequence Inferno Blast, Combustion

You could do the same with Pyroblast and Combustion for B). But it will come down to : Is your latency higher than the travel time of your last pyro? Which is literally: Are you able to get the traveltime to mostly 0.
Okay. First of, hats off to Skiz and Alzer, makes a very very interesting read and the effort is just incredible. Huge thanks from me.
Now. I play from a different continent, therefore my best case scenario is 180ms or so. (Inb4 comments that say dont play with that ping, i've actually managed some decent parses with that so i can live with it.) I have noticed that even my pyro crits the ignite value seems to drop, however i do a (fireball+pyro) -> (fireball+pyro) - > Pyro -> Pyro -> Combust -> IB on the PC. Irrespective of whether i have BL or not. The difference being that during a bl i can get both 2x BW and 2x DB into the crystal. Outside of BL i usually manage to only get 1x DB into the PC.

So this brings up two questions for me.

1. What do you think is the best approach for me on the PC taking into account the ping i play with?

2. Is it better if i (Fb+pyro) ->Pyro -> Pyro -> Combust ->2x BW -> 2x DB on the pc? This means i get a smaller combustion value but however the ring burst might compensate or pull ahead. I cant figure out which is better, especially with the ping issues and all. Suggestions?

PS: Also, the way i see it, mythic Intuition's Gift, fully upgraded is better than Sandmans, but people dont seem to be using it. So i feel like i'm wrong for some reason. Why? Doesn't it reduce the RNG of sandmans and give more static intellect?

Return to “Fire”

Who is online

Users browsing this forum: No registered users and 9 guests