Ah sorry for not having specified. No I played as fire through 4.2. Rawr would occasionally recommend one non-int gem last patch. Currently, about half of the gems suggested by Rawr are hybrid haste/hit + int. This thread recommends inferno rubies in every slot. I've been going with the recommendations posted here, but am wondering if it is perhaps no longer correct to do so.

Primitive Python modelling suggested that a 10-stack mean time would be approximately 36s, with one standard deviation being approximately -3/+7s (the distribution is obviously sharply biased by a minimum time at 100% proc chance). That took account of the self-interaction, and living bomb recasts, but only assumed a static chance per cast to proc Hot Streak. I no longer have the code, but it assumed of the order of 1900 base haste.

However in the "real world" combustion is a proc and not really so much of a cooldown. There is no guarantee that once you have stacked the 2pt13 to full you will get a suitable ignite in the next 0-20s at least.

I have been thinking about the optimal time to pop the first one as well, but to me it seems more important to make sure that 1) your ignite is over a certain threshold (taking prepot into account) and 2) you have gained enough stacks to satisfy the 2005 treshold or alternatively can compensate with a larger ignite (I use 10% for on the fly calculations, but 1/13 is more accurate I think).

The main conclusion isn't to wait for 10 stacks before using it, but rather, that for the first Combustion you have about 40s-50s window to use it and still have the same amount of activations in a fight.

It doesn't generally help with the second and the third ones, they will work the same way, ie. you'll still wait for a 'good enough' ignite to use it and will still have to gouge the fight duration to know how many you could optimally fit.

Speaking of, it would be good to have those rules or rather restrictions in simc. something in the order of: assuming a 6m fight the most I can fit is 5 activations, the average Ignite I expect in a fight being X and breakpoints in the fight duration past which you lose an activation if you wait to long. Although I understand it's probably not trivial.

Your T12 pieces are all iLvl 378, I wouldn't hesitate to replace legs/chest with 397 tier. I have 391 tier T12 except for head, and both Rawr/SIMC suggested swapping to 2+2 would be close to a breakeven for me; I went ahead and switched as it will make my remaining tier pieces obvious upgrades and to get some experience with the T13 2P bonus.

Rawr is also influential. Kavan reported today that T13 modelling could get some improvement due to the bonus being considered averaged and not at the top stack of it. However, I do not know the extend of that influence - it might be minimal.

Considering the discussion on the timing of Combustion and the interaction with the 2pc/4pc-bonus of T13, it has been suggested using Arcane Blast in fire specc to stack up the hastebuff. I havent found anything on that in here yet, so I thought I'd bring it up. A few thoughts I had on this:

- The benefit is obviously the 100% ramp-up chance on a very short casttime. It should take maybe arround 12-15 seconds to fully stack up the 2pc, thus gaining the full benefit of the 4pc.

- It could allow reliably reforging haste to the 1505-value. Right now, albeit just by guesswork instead of math, I decided not to get rid of my haste in order to not sabotage my first combustion, which would otherwise be applied after the first few casts without the full buffstack (also, having so little haste makes me feel slow. And fat. And like my parents never loved me.)

- The question of getting a strong combustion obviously remains. This could be a double-edged sword. Starting off with AB would most likely be a DPS-loss in itself (by how much?), but it would also interact with trinket-proccs etc. This could be a gamble in both directions, having trinkets procc too early, but also getting full effect of all buffs when starting to go for the combustion. Could be less of an issue without trinkets like Volcano or Theralion.

- If it turns out to be useful for the first combustion, there also remains the question of repeating this every time the buff runs out. Again, without proper math, it would seem that the comparatively low value of haste except for the Combustion-plateau would make this inadvisable, as long as we reliably get 10 stacks before the next combustion-cast. With low values of haste, this might change in terms of stat-value?

- Finally, using one AB to prevent your stack from falling off seems like a decent solution to extremly bad RNG.

All these points are mere thoughts. I understand that right now Rawr cannot be used to model these issues clearly due to the way the 2pc is handled, so I can't really put forth any data on this. I'd appreciate any thoughts on the issue.

There is surely a negativ side of stacking Stolen Time with Arcane Blast as a Fire mage but there are a two arguments we can't deny:

-The already mentioned 100% ramp up of the Buff
-The faster casttime of AB

The outcome of this is the faster usage of the second Combustion. So the main question is if there does exist an amount of ArcaneBlast worth to be cast for stacking Stolen Time regarding the average lost of Ignite und Hot-Streak-Procs. Something like : cast 5 Arcane Blasts and then continue with Fireball. I am not able to simulate this but I hope the main Idea is clear.

Refreshing Stolen time with AB before it fall off is indeed a great idea.

Thank you for explaining the mage to me. I also recently noticed that AB wasnt a fire spell. Now, back on topic, I tried some very rough calculations:

Given a normal gearset, stacking up the buff should take around 13 seconds with AB. In this time as fire I would cast ~6 fireballs and a HS-Pyro, adding up to a generous value of 350.000 DMG. I would have triggered my Combustion at ~3 Stacks of my setbonus, resulting in a CD-reduction of 10 seconds.

Casting the 10 Arcane Blasts during this time would get me ~200.000 DMG. I would have 10 stacks of my setbonus, resulting in a CD-reduction of 50 seconds. Taking another few seconds to ramp um a combustion reduces this to effectively ~38 seconds.

This means I have to weigh the ~150.000 DMG lost against the value of having:
- Combustion available ~28 seconds earlier the second time.
- Reliably having stacking trinkets activated for Combustion (relevant in this content).
- Proccing trinkets might go either way, proccing the first 10 seconds (bad) or after (good).
- Having the 500 haste available during Combustion, gaining either 1 tick (if you dont have 2005 on gear) or freeing up 500 haste to be reforged (if you have them right now).
- Apart from Combustion, higher uptime of the bonus, slightly increasing the value of the 2pc.

This so far is radically speculative. My guess would be that it would likely be valuable if 28-30 seconds less CD on Combustion actually gain you another cast of Combustion, this obviously depending on fight length. Also, 500 statpoints freed up in a way seem decent. Stacking again after the initial Combustion would obviously be less effective, unless bad RNG doesnt let you stack the buff till the cooldown comes up again.

You can try something like: actions+=/combustion,if=buff.bloodlust.up

It depend if you sim your fight with a BL at start or in execution phase. And don't forget RNG.