Yeah I didn't mention mana gems usage because after the first one, it's basically "use it on CD". But as you can see, mana gems are indeed included in the simulation.

So with mana gems included, the timeline you get for a 5 minutes fight is:
1 - burn your mana until you're 12500 mana short
2 - use first mana gem to go back to 100% mana
3 - burn phase until 38% mana (approx 1 minute)
4 - evocation to go back to 100% mana
5 - burn phase until around 40% (during this burn phase, you should be able to use your second mana gem when it comes off cooldown)
6 - hover around 40% mana with scorch and FB weaving. Use your 3rd mana gem when off cooldown.
7 - if the time to kill is lower than 1 minute, burn all the remaining mana with FB until the boss dies.

As you said, in a non patchwerk fight, the burn phase will last longer because you'll cast some scorches when you move. On a movement heavy fight, it's even possible you won't be able to burn all your mana (we're talking a LOT of movement here though, don't know if it's realistic).

What would be the best potion to use as we have 3 choices available?


-Mythical mana potion: 9250 - 10750 mana
-Volcanic potion: 1200 SP for 25 seconds
-Mysterious potion: 1-15000 mana and the minimum amount gained of mana scales with your alchemy skill.



A small note for Blood Elf mages: Don't forget about arcane torrent as it's another 6% mana on a 2 minute cooldown. You can macro it with your gem as they're both off the GCD and share the same cooldown.

I'm putting this out there so other folks can try to make sense of it as well, this is a table of Pyroblast values as compared to combustion values at variant gear levels (Basically, putting one piece of my level 80 gear on at a time).

Spec for the test was Talent Calculator - World of Warcraft level 80 caster against a level 80 target dummy, process was to cast a Pyroblast, let it tick once, ensure neither the DD or DoT was a crit, cast combustion, record tick values, debuff durations (via a simple macro), number of ticks observed, and a summary of stats on the character sheet.

SP	HA	HI	CR	MA	Pyro Tick DMG	Pyro Dur	Pyro Ticks	Combustion Tick DMG	Combustion Dur	Combustion Ticks
180	0 (0)	0 (0)	0 (2.05)	0 (20.0)	319	12	4	127	10	10
332	72 (2.2)	64 (2.44)	20(3.22)	0 (20.0)	339	11.744	4	134	9.79	10
403	117 (3.57)	64 (2.44)	65 (4.63)	0 (20.0)	348	11.588	4	137	9.66	10
552	180(5.49)	64 (2.44)	151 (7.25)	0 (20.0)	367	11.376	4	144	10.428	11
622	248 (7.56)	64 (2.44)	196 (8.66)	0 (20.0)	376	11.156	4	148	10.23	11
785	248 (7.56)	128 (4.88)	276 (11.38)	0 (20.0)	418	11.156	4	155	10.23	11
902	298 (9.09)	128 (4.88)	318 (12.79)	0 (20.0)	434	11	4	161	10.087	11
1654	350 (10.67)	167 (6.37)	318 (12.79)	0 (20.0)	537	10.844	4	195	9.944	11
1724	400 (12.20)	167 (6.37)	369 (14.32)	0 (20.0)	546	10.696	4	199	9.801	11
1764	400 (12.20)	194 (7.40)	391 (15.04)	0 (20.0)	552	10.696	4	201	9.801	11
1864	400 (12.20)	194 (7.40)	391 (15.04)	0 (20.0)	566	10.696	4	205	9.801	11
1966	450 (13.72)	194 (7.40)	391 (15.04)	0 (20.0)	579	13.19	5	210	9.669	11
2069	450 (13.72)	194 (7.40)	444 (17.75)	0 (20.0)	593	13.19	5	215	9.669	11
2148	450 (13.72)	237 (9.03)	499 (19.43)	0 (20.0)	605	13.19	5	219	9.669	11
2400	450 (13.72)	237 (9.03)	597 (22.74)	0 (20.0)	639	13.19	5	230	9.669	11
2511	450 (13.72)	237 (9.03)	637 (24.27)	0 (20.0)	655	13.19	5	236	9.669	11
2678	450 (13.72)	309 (11.78)	717 (26.85)	0 (20.0)	676	13.19	5	244	9.669	11

Neither the linear difference between the individual ticks damage, nor their ratio stays constant. Likewise when their hasted DPS is taken into account. Or if it is considered to be a damage per DoT calculation assuming their base durations, etc etc. I'm still poking at it to see if I can figure it out, but extra eyes are definitely welcome.


Edit: Crit percentage value for 637 was off.

Here is another table of values for LB for a comparison. The two behave similarly, but not exactly the same. Spec for the LB test is Talent Calculator - World of Warcraft

LB Tick	LB Dur	LB Ticks	Combustion Tick	Combustion Dur	Combustion Ticks
604	12	4	239	10	10
658	11.744	4	258	9.79	10
682	11.588	4	267	9.66	10
735	11.376	4	286	10.428	11
759	11.156	4	294	10.23	11
857	11.156	4	314	10.23	11
900	11	4	329	10.087	11
1176	10.844	4	423	9.944	11
1201	10.696	4	432	9.801	11
1216	10.696	4	437	9.801	11
1253	10.696	4	449	9.801	11
1290	13.19	5	462	9.669	11
1328	13.19	5	475	9.669	11
1357	13.19	5	484	9.669	11
1449	13.19	5	516	9.669	11
1490	13.19	5	530	9.669	11
1552	13.19	5	551	9.669	11

Fire Mages in Cataclysm need to manage their mana carefully.

Failure to do this will result in going OOM early, often - and poor DPS performance. This section will go into thorough details on Mana Management. This is one of the most important parts of this post that readers need to come away with an understanding of.

Fireball is the high DPS nuke, yet unsustainable to cast for long periods of time. You'll OOM yourself after several minutes if you try, long before the encounter ends. Scorch can be cast indefinetely due to its free mana cost, but it's lower DPS. So you can't just cast Scorch forever either. So what do you do? The answer is: Use both. Somewhere between these two extremes you need to find the balance for a given encounter. If you get the balance right - you'll make your mana the necessary duration and avoid going OOM - and still deliver maximum DPS possible.

This is the basic idea: Get used to casting Scorch and Fireball. But there's a few finer details on how best to actually go about this. The differing Fire Mage rotations can be described as the following:

Name	Notes
Burn Rotation	The Burn rotation refers to chain casting Fireball. This is our highest DPS rotation for single target, but is not sustainable for long. You want to use this rotation during Boss Weakened Phases During powerful on-use or passive effects (Trinket Procs, Engineering Gloves, Tailoring Capes) Molten Fury range After Spellstealing powerful Buffs During Encounter-specific +damage mechanics (Eg: While standing in a friendly void-zone which gives you +100% damage for the duration) To burn off excess mana before an encounter ends
Hover Rotation	This refers to the rotation used to keep mana levels hovering at a constant value. We call this our Mana Reserve. We use the Mana Reserve to switch into the Burn Rotation at the opportune time. A Hover Rotation makes extensive use of "Scorch Weaving". Scorch is cast frequently in rotation alongside Fireball. For example: 2 Fireballs to every 5 Scorches.
OOM Rotation	This is the rotation you can cast while completely OOM. Scorch and Pyroblast! on Hot Streak procs.
Movement Rotation	This is the rotation you can cast 100% while moving. Scorch, Pyroblast!, Living Bomb, Flamestrike, Blastwave.

To play like a Fire Mage, some good advice is to start thinking like an Arcane Mage. Mana is DPS. More Mana means more DPS. Unused Mana means wasted DPS.

With more mana is available, you can cast your Burn Rotation for longer and more frequently. This means we should be using every mana resource available. Use Mana Gems on cooldown. Use Evocation. Accept innervates from any friendly Druids. Every ounce of mana you get will let you cast your Burn Rotation a bit longer.

Your goal is to end each fight with exactly zero mana. Unused mana is wasted DPS - you should have started casting Fireball earlier to burn it off.

Readers might be thinking, "Can I just cast my Burn Rotation until i'm OOM, then use Scorch to regen to full again - then alternate between the two?" The answer is you can most certainly try this, but it's not possible nor optimal in many situations. Here are several scenarios which won't allow you to do this:

- Myriads of Encounter specific mechanics. (Running a debuff out, AOE'ing adds, running to stack up or spread out, jumping between platforms, avoiding obstacles, Boss weakened phases)
- Movement
- Adds may spawn, requiring Burn Rotation immediately (You don't want to be stuck on the Scorch regen side at the time)
- A Boss might apply a powerful spellstealable buff at random times. You want to have a Mana Reserve ready to immediately switch to Fireball.

Encounter mechanics will force you to constantly swap between Burn and Hover rotations. You cant plan around simply having the perfect rotation mapped out beforehand, which would only work on a "Patchwerk style fight". Lets walk through a Hypothetical 5 minute Patchwerk style fight. We'll try to map our Mana and Rotations out perfectly beforehand:

Order	Notes
	Hypothetical example only: A 5 minute Patchwerk encounter with no adds, no movement, no trinkets worn, no glove/cape enchants, no weakened phases. Do not interpret the table below as anything other than a hyptothetical example for what would be an unrealistic set of conditions for an encounter.
1	Start the fight with a "Burn Rotation" until you're 12500 mana short
2	Use the first Mana Gem to go back to 100% mana
3	Continue the "Burn Rotation" until 35% mana (Approxinately 1.5 minutes)
4	Use Evocation to go back to 100% mana
5	Resume the "Burn Rotation" phase until around 35% again. Use the second Mana Gem as soon as off cooldown
6	Use the "Hover Rotation" via heavy Scorch Weaving to keep mana levels constant. Use your 3rd mana gem as soon as it's off cooldown to bring us up to 50% mana. (We're hovering at 50% mana to save up enough mana for the incoming Molten Fury Range, which we expect to last 1 minute).
7	As soon as the boss reaches Molten Fury range (35% health) switch back to the "Burn Rotation".
8	Boss dies exactly as we run OOM.

For our 5 minute fight, with all these unrealistic conditions attached, our strategy might work. But can we rely on doing this for the majority of encounters? No.

If this boss had a movement phase, adds that spawned, AOE that was required, a "Weakened Phase", and if we're using powerful On-use trinkets, glove/cape enchants - suddenly all our neat plans go out the window. Fire Mages need to react dynamically to the unique needs of each different encounter, and even to differing attempts on the same encounter. A static, pre-planned rotation must only be treated as a guide - be ready to deviate at any moment. Be ready to use the "Burn Rotation" for a specific opportune time, then ready to switch back to the "Hover Rotation" after. This will vary not only between different encounters, but even between different attempts on the same encounter.

There are two other ways you can attempt to approach mana at 85. Neither of these are optimal from a DPS standpoint, but we'll note them here for the sake of thoroughness:

Technique	Notes
Don't use Fireball, only cast Scorch!	Despite sounding like a ridiculous thing to do, it could actually provide up to ~90% of the DPS your normal Fireball/Scorch rotation would ordinarily deliver. Surprised at how strong it is? The big reason is Hot Streak: Scorch is a fast casting spell, and casting it more often means you get many more opportunities for Hot Streak Procs. Those Pyroblasts are big DPS. Ultimately, attempting this style of play will be a DPS loss, even if not quite as big as you might initially suspect.
Use Mage Armor	Using Mage Armor makes mana issues largely trivial . However it comes with two key problems. Firstly, you lose DPS mobility. Firestarter allows Scorch to be cast while moving, when Molten Armor is active. There are many Cataclysm encounters where you need to move, and move often. Using Mage Armor will prevent you from DPS'ing effectively during these periods. Second: It's lower DPS. Simulations show that using Molten Armor (With Scorch Weaving) is simply put, the better thing to do. You can attempt to use Mage Armor, but you're putting yourself at a disadvantage.

Edit - Someone asked about using Mana Pot, Rejuv Pot and Volcanic potion. Which do we use? I don't know myself yet. Hopefully someone can chime in with more insight into the correct answer. I'll edit it in to this (which in turn will be edited back into the OP) as necessary.

I'm writing the aforementioned post in the style of a community resource which should be easily understood by everyone who visits EJ. Many players will come here with the very basic questions, "Im going OOM, help! What am I doing wrong?" and "How am I meant to be managing mana?" It sounds like a very noob question, but it needs to be answered carefully - as many Mages are essentially experiencing mana growing-pains currently. Controlling Mana like this was, for years, "an Arcane thing to do". And not really required for much of WOTLK at all anyway, even if you were Arcane. It's a very new concept to many players.

Let me know if there are any glaring ommissions - or inaccurate/wrong recommendations.

It does look likely the Living Bomb Glyph may need to be more accurately re-classified as Situationally Recommended, instead of "Skip". Pending the fight(s) you're doing. The OP was written for Glyphs with a single target scenario in mind only, which probably wasn't a good way to go about it. Several fights have AOE or multi-DOT potential: Maloriak, Halfus, Tron Council, Ascendant Council etc. Control/DPS on Magmaw's Parasitic adds is also (arguably) more important than DPS on the boss itself, so I could understand an argument for it on that fight too.

I believe Vontre mentioned the Molten Armor was by far the strongest, and thus is never swapped out under any scenario. Pyroblast was next strongest by a margin, with Fireball therefore being the leftover you'd consider swapping out for LB. The recommendation on Vontre's thread was to do this if "Living Bomb damage was double that of Fireball".

After someone clarifies this, we'll update the Glyphs section into the OP to account for whether Glyph Choices may change for specific AOE/multi-dot encounters.

...... how in the hell can a pyroblast glyph be better than fireball. you need to get fireball crits to even proc pyroblast - so by increasing your fireball crit you are increasing how many pyroblasts you are getting, by lowering your fireball crit you are essentially lowering the amounts of pyroblasts you proc anyways. There is no logical reason I would ever use pyroblast over fireball glyph wise. It just doesn't make any sense. I would LOVE to see a good argument for that, if there is one.

You don't really need to adopt that tone. There's other factors as well you're not accounting for. For example, on a fight where you make heavy use AOE (or a fight where you'll be extensively applying multiple LB) - you won't have so much mana available to cast Fireball or Burn Rotations as much. Therefore, you'll be casting more Scorches. What happens when you start casting lots of Scorches, because an encounter's AOE Needs mean you dont have much mana to cast Fireball? Hot Streaks and Pyroblasts go up considerably. You might, ironically, end up with more Pyroblasts cast - not less.

Wait until some of the more experienced theorycrafters chime in. What might make no sense to me or you, could probably be pretty easily explained by them.

It makes perfect sense. On fights where Living Bomb increases in value (Either heavy AOE fights, or where you are applying LB to multiple mobs) - you have less mana to cast Fireball or Burn Rotations. It doesn't matter whether you want to use Fireball the most. We're governed by what our mana will allow us to do, within the unique needs and constraints of encounter requirements, not what we'd personally ideally like. The more we need to AOE or apply multiple LB, the less mana we'll have available for Fireball. One might assume this means we'll see less Hot Streaks. But given that less mana means we'll be Scorching more in Hover Rotations, we ironically could be actually seeing more.

Astute readers might be thinking "Just Impact LB". While that might seem true at first glance, Blastwave is the only super-efficient way to do it. If you're fishing for Impact procs with Flamestrike, you'll be expending a lot of mana. If you're fishing for Procs with Scorch, then you'll be getting more Hot Streaks (And Pyroblasts). And finally, some fights need to have mobs tanked separated - where Impact isn't an option anyway.

Makes perfect sense. MA glyph gives you crit on everything, so it contributes to your DPS and HS proc rate regardless of what spell you're casting.

LB can potentially make up a large amount of your damage potential, or a small amount, depending on how many targets are available to you at the time. Thus it situationally can trump other spells and thus the glyph may trump others depending on the situation.

Depending on how many Fireballs you're able to cast, it may make sense to use Pyroblast over Fireball for glyphs. It isn't necessarily the spell you use the most, that is highly dependent on your mana situation and what priority list you are executing depending on fight specific mechanics. If your damage done list looks something along the lines of Scorch > Pyroblast > Fireball > LB, then your glyphs should be MA / Pyro / FB. If it looks like Scorch > LB > Pyroblast > Fireball, then MA / LB / Pyro may be the right answer. What proportions would make those selections optimal are a bit complicated to describe however.


The long and short of it is that a TC tool can tell you whether the math works out in a theoretical setting, or log parses can tell you in a practical one whether or not the proportions would seem to support it in a given scenario. The base concept is perfectly sound however, and makes perfect sense.

Has anyone considered replacing Fireball with Pyroblast as the main nuke in your rotation when Combustion is up? I'm not going to attempt the math but it's much more likely to score a single, large Pyroblast crit than attempting to crit 3 times consecutively with Hot Streak (2x to trigger Hot Streak, 1x for the Pyroblast! Ignite).

If not for Hot Streak, it seems like Pyroblast would be a better choice as a main nuke with higher dps. I'm not taking mana into account, but 1291 - 1631 damage in a 3.5 second cast comes out to [(1291+1639)/2]/3.5 = 418.6 dps, compared to Fireball's 405.6 dps and that isn't even including Pyroblast's DoT component (which has an added bonus by providing more fire for Combustion). It seems then, that waiting to use Combustion until you score a crit Pyroblast! by spamming Fireball is a waste of Combustion's uptime when you could just spam Pyroblast with it's normal cast time until it crits.



Edit: corrected # of times to proc Hot Streak

I was calculating fireball the same as I did the pyroblast dps [(892+1136)/2]/2.5.

I purposely left out the Dot portion of Pyroblast since, if youre spamming it, youre not making much use of it (although I guess you would get at least 1 tick during the cast time of your next Pyro if the first one did not crit.). Even without the DoT portion, the dps is higher.

I did not realize spell power scaled differently for different spells, but either way, according to WoWWiki, Fireball has a 100% coefficient while Pyro's is 115%. This page may be out of date though...