Cataclysm Fire Mage Compendium - Page 43

dar3652 · 11/15/11, 5:33 PM

I had figured the same thing, but tested on PTR with 1996 / 2004 / 2005 / 2006 haste and 2005 is indeed correct.

Here is the response I got from a simcraft dev about how exactly the math/rounding works:

"What you're missing is the rounding - read the issue I linked. Unhasted, the tick period is 1 second and the duration is 10 seconds, so at 1996 haste the period between ticks is:

1 / (1.05 * 1.03 * (1 + 1996 / 128.05716 / 100)) = 0.79995448

... which rounds to 800 milliseconds exactly. Divide 10 seconds by 800 milliseconds and you get 12.5 exactly, which rounds to the nearest even integer, which is 12. In other words, you need the period down to at least 799 milliseconds. (Because 10 / 0.799 = 12.5156446, which rounds up to 13.) At 2004 haste it's still not there:

1 / (1.05 * 1.03 * (1 + 2004 / 128.05716 / 100)) = 0.799522356 ~= 0.800

... but at 2005 haste it is:

1 / (1.05 * 1.03 * (1 + 2005 / 128.05716 / 100)) = 0.799468373 ~= 0.799"

• Zakalwe · 11/15/11, 5:38 PM

Originally Posted by Buundox

EDIT: Instead of working backwards from the 2005 value, I decided to look over the concepts closer during my lunch break. And since I haven't found a table listing what haste value I'd need for a third extra tick while having 3/3 NP, Dark Intent, and 5% raid buff, I figured I'd look into it here:

To get 13 ticks, you'd need Combustion to tick every 0.8 seconds. Which would require 1/0.800499 = 1.249220799 haste, or 24.9220799% haste. (I'm using the same logic in previous posts... correct me if I'm wrong).

To account for all of the outside haste buffs, I'd use:

1.24922079 = (1.03)*(1.05)*(1.03)*(Haste Rating from gear)
1.12143 -> 12.1438% haste from gear = 12.1438*128.057

Which solves to a required haste rating of 1555.1, or 1556 on gear.

Please advise to any inaccuracies.

No, when the final tick amount is exactly X.5, it is rounded to the closest *even* integer. The closest even integer to 12.5 is 12, so if Combustion ticks every 0.8 seconds it will result in 12 ticks. In other words, you need to get down to 0.799 seconds, which means 1/0.799499 = 25.07833% haste, which means 12.12284117% from gear if you have DI, which is 1573.06905 rating, meaning 1574 rating is needed.

Originally Posted by Hinalover

Maje I'm not sure if 2005 is the correct number, and I still say the 1996 number is correct. Unless my math is off which I really doubt.

100% haste = 12805.7006835937500

to get 13 ticks, we need 25% haste raid buffed
0.25 = ((1.03) * (1.05) * (1 + base haste rating)) - 1
0.1558021266759130836800739713361 = base haste rating or 15.58%
12805.7006835937500 * 0.1558021266759130836800739713361 = 1995.1554000791002080443828016644

Let's look at 2005:
2005 / 12805.7006835937500 = 0.15657089366212824450475337998538 or 15.657%

Good job completely ignoring the rounding issues.

Hinalover · 11/15/11, 5:43 PM

Originally Posted by Zakalwe

No, when the final tick amount is exactly X.5, it is rounded to the closest *even* integer. The closest even integer to 12.5 is 12, so if Combustion ticks every 0.8 seconds it will result in 12 ticks. In other words, you need to get down to 0.799 seconds, which means 1/0.799499 = 25.07833% haste, which means 12.12284117% from gear if you have DI, which is 1573.06905 rating, meaning 1574 rating is needed.

Good job completely ignoring the rounding issues.

Part of the problem is that you really didn't explain it well. You gave a general explanation using the 87.5% instead of say the +3 ticks to combustion but once Dar gave a concrete example, was I able to follow the logic on where you were going. I just could not wrap my brain around the idea of the 1601 and 1600 millisecond idea.

Edit:
So for Combustion, in order to reach the +3 tick, you need to bring the tick down to 0.7994999999999999 or:

1.2507817385866166353971232020013 = ((1.03) * (1.05) * (1+ haste rating))
1.1565249547726459874222128543701 = 1 + haste rating
0.1565249547726459874222128543701 = haste rating or 15.65%
12805.7006835937500 * 0.1565249547726459874222128543701 = 2004.4117203315535227460991136105 or 2005

Buundox · 11/15/11, 6:48 PM

Originally Posted by Zakalwe

No, when the final tick amount is exactly X.5, it is rounded to the closest *even* integer.

Thanks. I seemed to have overlooked that part in your previous post.

Kyrika · 11/15/11, 8:20 PM

Hi guys
I've reading last posts about haste...
With 2pT13 we will have between 0 and 500 more haste. This haste will be lost when combustion expires, so I guess we will be able to get to 10 stacks (+500 haste), use combustion, and when it expires we have more than a minute (even with 4pT13) to get 10 stacks again.
So, we could say that the "soft cap" to get an extra tick of combustion would be 2005 - 500 = 1505, because every time you use combustion, it is affected by the +500 haste from the tier.
Am I right? ^^

• Zakalwe · 11/16/11, 12:14 AM

Useful spreadsheet:

https://docs.google.com/spreadsheet/...URKcmdibFFTeVE

Kyrika · 11/17/11, 3:26 PM

Originally Posted by Zakalwe

Useful spreadsheet:

https://docs.google.com/spreadsheet/...URKcmdibFFTeVE

The link doesn't work for me (it says something about permission to see the content)

Exoduz · 11/22/11, 7:49 AM

is this right?

Combustion	+1 tick	+2 tick	+3 tick	+4 tick
+Goblin +3/3 Netherwind Presence and 5% Raid Buff	-	679	1859	3027

Icaruz25 · 11/22/11, 10:08 AM

Originally Posted by Exoduz

is this right?

Combustion	+1 tick	+2 tick	+3 tick	+4 tick
+Goblin +3/3 Netherwind Presence and 5% Raid Buff	-	679	1859	3027

And +Di is it worth passing 2005 to 2845 as a goblin or 3002 as a troll and also the 3027, would be sacrificing a lot of crit doing so.

Ikcelaks · 11/22/11, 10:36 AM

Originally Posted by Maje

The numbers are correct, 4 ticks with 1181 haste rating.

On a similar subject, I'm not certain how helpful knowing the haste ratings for LB ticks is given that the DPS for Living Bomb follows the haste rating pretty closely regardless of the number of ticks. In laymen's terms having 1181 haste rating or 1182 doesn't mean a world of difference if any at all. The gain, if any, would probably be from more GCDs spent on other spells diminished by a smaller amount of explosions.

On the other hand, due to the nature of Combustion, being on cooldown, every extra ticks there is in fact a kink in the line and where haste rating value becomes discontinuous.

EDIT: it's arguable that since Pyroblast! dot works the same way Living Bomb does, but in a way is on CD (since the uptime isn't 100%) it might introduce some discontinuities in the haste value at the same points, the benefit however is marginal.

For Combustion however, reaching the first ticks is worth 10% of combustion value, the second 9.1%, third 8.3%, fourth 7.7% and so on. Assuming combustion is around 8% of the fire dps, those kinks are not insignificant, starting from 0.8% total dps down to 0.66% at 2005 haste rating.

With the buff to the LB explosion damage, hitting haste plateaus actually decreases the damage done by LB, but the saved GCDs make up for it. The explosion buff should help smooth out the value of haste.

Once we equip our 2-piece bonus, our haste will vary by upto 500 points, so I suspect that the Combustion breakpoints will be the ones that really matter.

The buffs to Fireball and Pyroblast in addition to the mechanics of the 2-piece bonus have drastically increased the scorch tax. Compared to the old days of 4.0.6, I plan to drop Improved Scorch, cast as few Scorches as possible, and regain mana through other means if necessary. The extreme buff to Pyroblast might open up the option of hard-casting to maintain the DoT. The viability of this tactic will depend on our HS frequency and the effect of the high mana cost, but I believe it's worth considering. I hope that it turns out to be optimal, because that would add some nice flavor to the otherwise ultra-simple single target rotation.

What is the current accuracy of simcraft with respect to ignite munching? I'd like to play around with some crit-stacking itemizations to see if they're viable, but I'm not sure how much I can trust the sims. The quadratic nature of T2 Hot Streak potentially makes extreme values of crit very valuable. (T1 Hot Streak scales linearly with haste and less than linearly with crit, but T2 Hot Streak scale quadratically with crit and not at all with haste.)

GnomerTerp · 11/23/11, 11:02 AM

For what it's worth - I couldn't drop improved scorch in my PTR build on madness of DW encounter (25 normal mode). The fight is just very long and I'd imagine the additional haste in T13 will dry up the mana reserves even faster. I didn't play around with armor switching at all.

• Zakalwe · 11/24/11, 8:25 AM

Originally Posted by Exoduz

is this right?

Combustion	+1 tick	+2 tick	+3 tick	+4 tick
+Goblin +3/3 Netherwind Presence and 5% Raid Buff	-	679	1859	3027

Nope, this is right:

Combustion	+1 tick	+2 tick	+3 tick	+4 tick
+Goblin +3/3 Netherwind Presence and 5% Raid Buff	-	678	1858	3027

Sothasil · 11/26/11, 1:26 PM

I've made a table for the haste-plateaus, I'm not 100% sure about the numbers, but Hinalover allready looked over it (thanks again to you).

Imageshack - hasteplateaus.jpg

Maybe if I'm bored enough I will make an interactive one of this.

• Zakalwe · 11/26/11, 7:51 PM

Originally Posted by Sothasil

I've made a table for the haste-plateaus, I'm not 100% sure about the numbers, but Hinalover allready looked over it (thanks again to you).

Imageshack - hasteplateaus.jpg

Maybe if I'm bored enough I will make an interactive one of this.

Most of those numbers are wrong. What sort of crazy math did you use to come up with them?

In fact, the only correct numbers I can find in that table are the ones for +3 ticks - every single number I've checked outside of that row are wrong. Makes me wonder if you started with that one row and extrapolated in some strange way.

Sothasil · 11/27/11, 7:01 AM

Originally Posted by Zakalwe

Most of those numbers are wrong. What sort of crazy math did you use to come up with them?

In fact, the only correct numbers I can find in that table are the ones for +3 ticks - every single number I've checked outside of that row are wrong. Makes me wonder if you started with that one row and extrapolated in some strange way.

Again, thanks for your help with this, Zakalwe

Now my numbers should be right, but can't hurt to check some of them :P

ImageShack� - Online Photo and Video Hosting