Cataclysm Fire Mage Compendium - Page 27

talchas · 03/18/11, 12:24 AM

Originally Posted by Kavan

I've been looking into Combustion lately and I just can't figure out how it works so if anyone has any insight I'd like to hear it.

I haven't tested it in 4.0.6, but my (not as well explained as I'd like) results from earlier are at here and here. The short version is that it just checks the existence of LB/pyro and then computes combust dot strength based on a separate, not quite identical set of coefficents. In addition, it applies mastery twice (but CM only on one of the mastery applications), and when summing the LB and pyro dots it seemed to increase by about 3%.

Kyr205 · 03/19/11, 9:13 AM

Originally Posted by Kavan

I've been looking into Combustion lately and I just can't figure out how it works so if anyone has any insight I'd like to hear it. The tests were performed with fire spec, no glyphs, no fire power, one point in critical mass and no debuffs. In each case I just combusted a living bomb with no procs.

1) 197 spell power, 22.4% flashburn

living bomb tick: 715
combustion tick: 288

2) 4884 spell power, 44.268% flashburn

living bomb tick: 2875
combustion tick: 1273

3) 5199 spell power, 44.268% flashburn

living bomb tick: 3012
combustion tick: 1332

In each case living bomb tick matches exactly the predicted theoretical value of (403 + 0.233 * spell power) * 1.25 * (1 + flashburn + 0.05).

Now from all I know combustion should have dealt one third of living bomb tick value since living bomb ticks once every 3 seconds. From some other tests I confirmed that combustion tick scales the same way with fire power and critical mass as the living bomb tick. So the only thing that I can't figure out is the base conversion. I can't explain it with flashburn or any other multiplier. It almost seems as it's using completely different base and spell coefficient for the purposes of computing combustion. The only problem is that when I tried to fit the coefficients I wasn't able to get anything consistent with data so far.

Haste affects LB as well. Given from the information, I can't tell whether or not that is the problem you are encountering because the only listed numbers are SP, Mastery, and tick values. Since LB would be ticking faster with haste, the dps value would increase even if you weren't getting another tick out of the LB. At least I believe so.

• Wizeowel · 03/20/11, 6:29 AM

Originally Posted by Kyr205

Haste affects LB as well. Given from the information, I can't tell whether or not that is the problem you are encountering because the only listed numbers are SP, Mastery, and tick values. Since LB would be ticking faster with haste, the dps value would increase even if you weren't getting another tick out of the LB. At least I believe so.

The dps value of LB itself might increase with haste, but if you want to assert that this also affects the value that is used for combustion then you need to provide some more evidence than "I believe so". As Talchas wrote, the value used by combustion for the Living Bomb ticks is demonstrably not based on the actual dps from that LB, so if you want to argue against that you need to show some combat logs and calculations which could support it.

shadowritual · 03/21/11, 6:55 PM

Playing a bit with Rawr I noticed that we could set the theorical damage reduction of ignites. What I would like to know is if we set that option to 0, can we use the new stats relative values as what it would be without ignites or is there any other mechanism that would change those values?

I know that in rawr, munching ignite is calculated as a direct damage reduction on crit so I was really wondering if I could use those stats as a preview of what is to come, if they manage to fully fix ignite munching.

Kavan · 03/21/11, 7:24 PM

Originally Posted by shadowritual

I know that in rawr, munching ignite is calculated as a direct damage reduction on crit so I was really wondering if I could use those stats as a preview of what is to come, if they manage to fully fix ignite munching.

Yes lowering that should be enough. You should also check the PTR option in advanced tab to disable ignite on dots and no gcd on combustion.

Mangara · 03/22/11, 1:15 PM

I was wondering about something Wizeowel said in the Simple Questions thread:

Originally Posted by Wizeowel

The actual priority in terms of damage per cast is: FO, Pyroblast!, LB, Fireball. However, there is an argument which says that if you get hotstreak at the same time as needing to refresh LB, then do the latter since simultaneous crits from a fireball+pyroblast! would munch one of the ignites.

In the absence of Ignite munching, is it always better to use the HS proc before refreshing LB? Due to haste scaling, a Living Bomb always lasts 10 GCDs for players between 12.5% and 37.5% haste, so the question becomes whether 1/10th Living bomb does more damage than we stand to lose by holding on to the HS proc. This depends on the average damage of LB, Pyro and the probability that the HS proc is refreshed during the GCD used to refresh LB.

The first two are easy to find from Rawr; in a near-BiS set with 8.5k spellpower, 13.7 mastery and 38.2% crit, Living Bomb does 29712 damage on average, while Pyroblast! does 36928 damage on average, including ignite. The refresh probability is harder to compute. If you are quick to react to the HS proc, you see the proc during the cast of the fireball after the one that procced HS and this is the only spell that lands during the next GCD. This means that T4 HS doesn't even come into play; we only need to take the T3 proc chance into account. At 38.2% crit raid-buffed, this gives a proc chance of 0.052, according to Shaewyn's formulas. This makes the expected damage loss from holding on to the HS proc 2714 damage, while delaying the LB costs us 2971 damage. So even in the absence of Ignite munching, you should refresh LB before using a HS proc if you are near BiS gear. Note that these calculations all assume that Glyph of Pyroblast is used and Glyph of Living Bomb is not used. Changing this would skew the results even more towards LB.

With lower gear like my own (mostly 359), the situation changes slightly. The expected damage loss of a Pyroblast! with full ignite becomes higher than that of delaying LB, but lower if the ignite is lost, making the situation less clear cut. To fully analyze this situation, we would need to know how likely it is that an ignite gets munched. However, I think that in general, reapplying Living Bomb before using Hot Streak procs is a good rule of thumb.

• Wizeowel · 03/22/11, 8:55 PM

Originally Posted by Mangara

To fully analyze this situation, we would need to know how likely it is that an ignite gets munched.

Kavan asked me this the other day but I couldn't find any conclusive tests. I always believed that since fireball and pyroblast! have the same travel time and spell queuing makes the pyroblast truly instant, that a double crit will always munch the fireball's ignite. Imagine my surprise when after testing it quickly at the dummies just now I see that the pyroblast! lands first and it's that one which is munched. It's not really an easy thing to test since you need 4 crits in a row, so I didn't test long but here are two occassions the pyroblast! ignite seems to have been munched. My mastery should give 31.4% extra damage on dots.

[00:42:16.838] Wizeowel begins to cast Fireball
[00:42:19.134] Wizeowel's Hot Streak fades from Wizeowel
[00:42:19.134] Wizeowel casts Pyroblast! on Raider's Training Dummy
[00:42:19.561] Raider's Training Dummy afflicted by Pyroblast! from Wizeowel
[00:42:19.561] Wizeowel Pyroblast! Raider's Training Dummy *25656*
[00:42:19.986] Wizeowel Fireball Raider's Training Dummy *21582*
[00:42:20.387] Raider's Training Dummy afflicted by Ignite from Wizeowel
[00:42:20.387] Raider's Training Dummy afflicted by Critical Mass from Wizeowel
[00:42:20.387] Raider's Training Dummy's Ignite is refreshed by Wizeowel
[00:42:22.374] Wizeowel Ignite Raider's Training Dummy 5689 (~= 21582 * 0.2 * 1.314)

[01:02:43.833] Wizeowel begins to cast Fireball
[01:02:46.105] Wizeowel's Hot Streak fades from Wizeowel
[01:02:46.105] Wizeowel casts Pyroblast! on Raider's Training Dummy
[01:02:46.512] Raider's Training Dummy afflicted by Pyroblast! from Wizeowel
[01:02:46.512] Wizeowel Pyroblast! Raider's Training Dummy *25718*
[01:02:46.673] Wizeowel Fireball Raider's Training Dummy *21278*
[01:02:46.911] Raider's Training Dummy afflicted by Ignite from Wizeowel
[01:02:46.911] Raider's Training Dummy's Critical Mass is refreshed by Wizeowel
[01:02:46.911] Raider's Training Dummy's Ignite is refreshed by Wizeowel
[01:02:48.848] Raider's Training Dummy's Ignite is refreshed by Wizeowel
[01:02:48.881] Wizeowel Ignite Raider's Training Dummy 5609 (~= 21278 * 0.2 * 1.314)

Any corroborating or counter evidence would be welcome, not just for the above question but also to help with Rawr.mage ignite munching calculations.

eaglesrock · 03/23/11, 2:48 PM

Originally Posted by Bowchikabow

Are there any predictions on how the ignite change will affect FB fire vs FFB fire? I remember one of the reasons FB > FFB was because FFB dot was eating ignites. If the DoT is no longer affecting ignite, and the damage (after the FFB glyph) is comparable while having the same mana cost. Given that the FFB DoT, so far as I know, contributes to combustion damage, and can be spread through impact; would it make sense that we would see an increase in FFB use?

I would love to get in on the PTR and test this.

I would love to know this too. With high levels of gear mastery getting closer in relative value to other stats it seems like there should be a reachable breaking point where ffb > fb

talchas · 03/23/11, 3:29 PM

At the moment ffb's contribution to combust is a bad joke for some reason - 18 damage/tick combustion from a 3 stacik ffb dot ticking for 330. This might get fixed in 4.1, if someone wants to test on the PTR.

Zaldinar · 03/23/11, 7:56 PM

I start on night shift next week for three months, and as a result will have a fair amount of free time available for doing research. I can start looking into these things in more detail again.

Things needing to be isolated still:

- Ignite munching conditions (especially in light of the 4.1 changes?)
- Combustion generation (Still 'wonky')

Anything else to put on my list?

angayelle · 03/24/11, 8:13 AM

I'm sending you my beta version of CombustionHelper with ignite prediction and munching report. I couldn't use yours as it is but you may have a use for mine.

Rulaneda · 03/27/11, 9:46 AM

Is there any addon that helps me track the internal cooldown after Cauterize has been activated?

I couldn't find an answer to my question in this or any other threads here, or I am just too stupid to find one. Thanks in advance.

Edit: I just tried Forte which tracks trinket CDs very nicely but fails to track Cauterize.

fok · 03/27/11, 1:56 PM

You can try a `KorreCauterize` addon.

But I simply look at other skill on my panel bar, like combustion, it gives me approximate time, because I really don't need to die every minute.

Rulaneda · 03/27/11, 7:05 PM

Yeah, my guild is doing Valiona and this Thoralion and we use cauterize to eat blackouts and meteors. But sometimes you "die" and may get a blackout or meteor about 60 seconds later and today we had this situation once that I didn't know wether cauterize was up or not.

Oh, and I can't find your addon on the two usual websites I get addons from. Or is it possible to add this cooldown to Forte somehow? I just downloaded it a few hours ago and still try to figure out details.

Maje · 03/28/11, 12:08 PM

Originally Posted by Rulaneda

Is there any addon that helps me track the internal cooldown after Cauterize has been activated?

I couldn't find an answer to my question in this or any other threads here, or I am just too stupid to find one. Thanks in advance.

Edit: I just tried Forte which tracks trinket CDs very nicely but fails to track Cauterize.

I use ICDbars, the latest version can track Cauterize (debuff triggered ICDs).