Cataclysm Fire Mage Compendium - Page 35

Charybdis · 06/08/11, 11:50 AM

The main reason for the odd haste cutoffs is the way DoTs scale and how Blizzard has set up ratings in general. Other DoT (and HoT) classes have noticed it takes 1 or 2 extra haste to fully gain the benefit of extra ticks. There's been no confirmation on it, but from a programming standpoint it's almost guaranteed to be rounding calculations that are causing the issue. Haste being multiplicative means they're trading accurate, more intensive code for simpler but less accurate numbers.

Because it's a rounding issue, haste calculations across the board will need to add 1 or 2 extra rating to be sure the game recognizes the character as having the needed haste.

Maje · 06/10/11, 12:39 PM

Did a couple of short tests with 378 ptr gear, both tests were done with only t3 HS selected, Molten Armor + Glyph.

Test1: 4t12 crit 30.13%
scorch-crit 796 (0.35066079295154185022026431718062)
scorch-hit 1474
Hot Streak 376 (AURA_APPLIED + AURA_REFRESH) (0.47236180904522613065326633165829)

With such an extremely high HS rate I decided to run a test without using 4t12, switched some gear around and managed to get the same crit%.

Test2: 2t12 crit 30.13%
scorch-crit 273 (0.34732824427480916030534351145038)
scorch-hit 513
Hot Streak 29 (AURA_APPLIED + AURA_REFRESH) (0.10622710622710622710622710622711)

Test2 was much shorter since I didn't have much time and just wanted to verify some things.

Conclusions:
1. 4t12 is probably T3hs = 4t12% + MAX(k(c), 0) (as Zaldinar suggested)
2. 47% is around 5% higher than the suggested value assuming the old k(c) formula and 30% for 4t12, which means that either or both are slightly wrong.
* Test2 shows that at least k(c) didn't change significantly, but since the test wasn't long enough, we can't really check the correct value exactly.

I did a bit of testing on 2t12, it seems to have a pretty high proc chance, and around 45s icd, it's up for 15 seconds and seems to cast Fireballs even if you don't have a glyph, the spell is Fireball - Spell - World of Warcraft which suggests it doesn't scale with spell power, it manages to cast 4 fireballs during the uptime. Averaging the numbers (assuming 0% crit) it's around 28k damage every ~50s. Though more tests are needed for the specifics.

Variable Pulse Lightning Capacitor - Item - World of Warcraft seems to behave as expected, ie. there is a 10% chance per stack to proc a Lightning Bolt - Spell - World of Warcraft which does damage times the number of charges you had, it uses your crit rate, doesn't scale with spellpower and all the bling aside simply put does Lightning Bolt - Spell - World of Warcraft damage every time you crit while obeying the 2.5s rule.

EDIT: about the capacitor, the proc is extremely simple mechanic which they just played a bit with (not sure why), but it's equivalent to having no stacks and proccing on every crit (2.5s) for the 1 stack amount, or can also be looked at having 1 proc every 10 crits for 10 times the 1 stack proc.

Elrin · 06/13/11, 7:07 AM

I did a bit of testing on 2t12, it seems to have a pretty high proc chance, and around 45s icd, it's up for 15 seconds and seems to cast Fireballs even if you don't have a glyph, the spell is Fireball - Spell - World of Warcraft which suggests it doesn't scale with spell power, it manages to cast 4 fireballs during the uptime. Averaging the numbers (assuming 0% crit) it's around 28k damage every ~50s. Though more tests are needed for the specifics.

Maje - by chance did you look at whether our intrinsic haste affects the FB cast time of the the mirror? Just wondered if there was a cut off haste at which they will cast a 5th fireball. If possible it would equate to a dps increase of about 140 (35k dmg over 50s VS 28k dmg over 50s)

Batlecruiser · 06/13/11, 7:27 AM

Originally Posted by Maje

I did a bit of testing on 2t12, it seems to have a pretty high proc chance, and around 45s icd, it's up for 15 seconds and seems to cast Fireballs even if you don't have a glyph, the spell is Fireball - Spell - World of Warcraft which suggests it doesn't scale with spell power, it manages to cast 4 fireballs during the uptime. Averaging the numbers (assuming 0% crit) it's around 28k damage every ~50s. Though more tests are needed for the specifics.

If the illu lasts 15secs and casts the mentioned FB in 2,5sec: Why does it manage only 4 casts? 15s / 2,5s seems to be 6 -.-

Hinalover · 06/13/11, 3:00 PM

Originally Posted by Batlecruiser

If the illu lasts 15secs and casts the mentioned FB in 2,5sec: Why does it manage only 4 casts? 15s / 2,5s seems to be 6 -.-

A few reasons is that there is a lag between when the mirror image finishes it's cast and the server recognizes that it's finished it's cast and the mirror image is told to start it's next cast. This equates to about 1 second gap between each cast. Frost Mages and Destruction Warlocks have this same problem on their Water Elementals and Imps. They usually negate this lag by macroing into their casts their pet's attack, forcing the pet to cast earlier than they normally do. However, there is no "pet cast" ability we can do to increase the casting speed of our mirror images.

The mirror image does starts casting a 5th Fireball, but despawns before it can finish the cast.

Althor · 06/14/11, 8:34 PM

I have been working on implementing Tier 12 set bonuses for all classes in SimulationCraft and here is what I've discovered so far for Mages from both testing on the PTR and data-mining.

2pc: It does indeed last for 15 seconds and have a 45second ICD. It derives no benefit from the Mage's haste nor crit. In a few hundred casts the only times I have seen it crit was when someone had put up Critical Mass on the target.
In terms of mechanics it appears to be pretty much identical to the Balance Druid and Warlock 2 piece bonuses. I performed extended testing on the Balance Druid version and on my connection I noticed an average latency between casts of 0.74 seconds with a standard deviation of 0.62 seconds.

4pc: For Frost Mages this increases the proc chance of Brain Freeze by 15%. For Fire Mages with Hot Streak I G15'd some Scorch tests.
In the first test I was wearing only 2pc Tier 12 with a paperdoll crit of 24.64%. I had 1094 scorch crits and 289 Hot Streaks (SPELL_AURA_APPLIED+SPELL_AURA_REFRESH) or a 26.4% conversion rate.
In the second test I wore 4pc Tier 12 with a paperdoll crit of 24.63%. I had 1927 scorch crits and 1112 Hot Streaks (SPELL_AURA_APPLIED+SPELL_AURA_REFRESH) or a 57.7% conversion rate.

Using the formula in this thread for the 2pc bonus I should expect a 27.73% Hot Streak proc chance and I got 26.4% (which seems near enough).

Using the formula on my 4pc test I should expect a 27.76% Hot Streak proc chance discounting whatever bonus the Tier 12 4pc gave. As I had a 57.7% conversion rate that is almost exactly 30% more than the expected amount.

So my testing backs up earlier testing that the Tier 12 4pc adds a flat 30% proc chance to Hot Streak.

Maje · 06/15/11, 4:30 AM

Latest version of rawr models the 4t12 for fire. I've ran a few tests on the valuation and it seems that if you have 4t11 heroic ie. 372 gear switching to normal 4t11 is a dps loss, it follows that breaking 4t11 for higher ilvl pieces isn't worth it either.

Basically the earliest you can expect to switch your 4t11 is when you start getting heroic t12 pieces.

Grimatoma · 06/19/11, 1:31 PM

Hi, I have a quick question on how I should AOE as a fire mage. I have played a Mage since the start of wow but stopped playing for a while (around the time living bomb came out) and when I came back the new way for Mages to aoe is to use living bomb and Impact it so that it can reach all the mobs.

So when I AOE now I open with pryo then lb then bw then fireblast which should be correct but the part that I am not sure about is what to do after this on a long fight? Do I just spam Flamestrike or do I restart my rotation with another pryo? Or do I reimpact a living bomb every time the previous one explodes and spam Flamestrike in between? Or something else.

I have been looking around on the forums for a while but I cant find any information about this.

Senovit · 06/19/11, 2:15 PM

Usually it depends on how long the enemies typically stay alive and how many there are, but on mobs with high hp you'll generally want to start off with your normal single target rotation on one, build up a good combustion and then spread that. Afterward you will want to keep blast wave and flamestrike dots on the ground, and afterward a good choice would be to either fireball or scorch and try to impact the ignites off. And of course, always keep living bomb on 3 targets.

If the enemies have low hp though, it may be better to simply spam flamestrike even though it's a lot more mana intensive, as the mobs might die before you can land a solid crit. Flamestrike spamming is rather low, but consistent damage, and spreading ignites is inconsistent high damage, which becomes more consistent if the mobs have much higher hp.

Grimatoma · 06/20/11, 12:18 AM

Originally Posted by Senovit

And of course, always keep living bomb on 3 targets.

Do you mean tabbing to put Living Bomb on 3 different targets or to have the Living Bomb impact off?

Lurama · 06/20/11, 12:06 PM

Originally Posted by Grimatoma

Do you mean tabbing to put Living Bomb on 3 different targets or to have the Living Bomb impact off?

Most likely an impact spread LB. Otherwise you end up spending an additional GCD.

okaymindflay · 06/20/11, 1:29 PM

The DPET of LB is higher than anything else you can do in that situation, so it is certainly worth it to get LB on three targets if no impact is up.

Blueobelisk · 06/20/11, 5:47 PM

Hey guys I have a question.

On some other forum I have been seeing people say that in 372 heroic gear fire is less RNG/more consistent as comared to lower leveled geared. Consistent enough to hit the top of the DPS charts.

My question is why such a seemingly small difference between gear make such a big impact on Fire being more stable? Does the added haste/crit make it so that there are much more casts or much more criticals to get hot streak to proc or ignite damage? Or is it still somewhat random and I'm just hearing the false information?

Lgs · 06/21/11, 12:09 PM

Originally Posted by Blueobelisk

...On some other forum I have been seeing people say...

Multiple simulations (RAWR, simulation craft) and collected damage logs (stateofdps) all show that fire doesn't scale as well as Arcane, and it still relies heavily on RNG that is independant of gear.

Videl · 06/22/11, 4:42 PM

Originally Posted by Blueobelisk

On some other forum I have been seeing people say that in 372 heroic gear fire is less RNG/more consistent as comared to lower leveled geared. Consistent enough to hit the top of the DPS charts.

More significantly, I think, your premise that consistency will lead to topping dps charts is just wrong. The opposite is actually true. Highly variable dps like fire mages have makes it more likely that, in spite of having middle of the road average dps, things like WoL will show some top parses that look very good. Some would argue that that is as much a problem with fire as anything, Top end parses make it look better than it really is.