[ARCHIVE] 1.1.5 Madness/Balance Sorcerer DPS Compendium

[Uninformed]

very nice! Now let's see a new discussion about Sorcerer pvp using this assessment.

I am inclined to go Creeping Terror full 31 build for the stun. Do you consider that build superior to CL based builds for pvp or no?

[pingpong]

Thanks for this write up. I've been following a 0/13/28 path so far and it's good to get some confirmation on the DPCT of various spells.

[Bunnyfufu]

I'm curious of the breakdown of your math for your DPCT. For instance Creeping terror seems low considering it only costs 1 global CD to cast?

[Kaedis]

Quote:I am inclined to go Creeping Terror full 31 build for the stun. Do you consider that build superior to CL based builds for pvp or no?

This is primarily a PvE discussion thread. On the PvP side, I've honestly not considered it.

Quote:I'm curious of the breakdown of your math for your DPCT. For instance Creeping terror seems low considering it only costs 1 global CD to cast?

The math is pretty simple, DPCT = (Stdmin/max * StdHP(level) + FBonus * Coefficient * NumTicks * Modifiers) / CastTime (1.5 if instant)

It's coefficient is identical to Affliction's (0.031 StdHP min and max, 0.311 coefficient), but it's an unranked ability due to being skilled for, and thus uses the level 40 base HP (1380) rather than the level 50 one that Affliction does (1680). Thus by default it will always do 9.3 damage less per tick. Even if that wasn't the case, CT falls behind. By default, it lasts 6 ticks compared to Affliction's 5 ( = 20% more damage), but in a Madness build, Affliction has the benefit of +6 seconds duration (+40% to total damage) and +15% damage from Force Horrors, placing it at 34.2% better than Creeping Terror (1.4 * 1.15 / 1.2). For the 3/7/31 build, Affliction's DPCT is 37.7% higher than CT's, exactly on par when you include the lower base damage. Creeping Death and Deathmark affect both equally. Also, I didn't include any calculations into Affliction's DPCT for Lightning Barrage for builds that have it, since it's difficult to model properly, so that would further increase the cap between them.

CT's DPCT isn't necessarily low, though. The issue is that Force Lightning's DPCT is so ridiculously high. The calculations above, btw, didn't factor in DPS gain from Wrath, also difficult to model, further inflating the value of Force Lightning. The fact that that one spell is spammable, our most force efficient by far (actually being force-positive in practice due to base regen), procs Wrath, gains benefit from Lightning Barrage, and has a very high intrinsic DPCT is definitely problematic, as it makes it difficult to give us reason to cast anything else without making our dps output completely overpowered.

[Bunnyfufu]

Those calculations seem unnecessarily convoluted. Wouldn't it be easier to just take (total damage)/casting time. One GCD = 1.5s cast?

[Kaedis]

Total damage is given by Stdminmax * StdHP(level) + FBonus * Coefficient * NumTicks * Modifiers.

Edit: More detail to clarify. Stdminmax * StdHP(level) + FBonus * Coefficient gives the tooltip value for anything but DoT abilities. For DoTs, you must multiple by the number of ticks, as the XML data notates base damage and coefficient per tick for DoTs, not for total. Modifiers include talents, buffs, etc, and must be taken into account as well, as not all abilities are affected equally. This gives the final total of how much damage the ability will deal. Dividing by the effective cast time gives DPCT.

[Asmiroth]

With Force-Master gear, Draay's Legacy and M7-R3 relics my base stats are different than yours, though not by much. Notably Crit is 5% higher, at 41% - which I'm pretty sure comes from Coordination buff.

At that stat level, CT comes out above Force Lightning. Still, when modeled, it actually drops overall DPS since you're better served with DF ticks on the two first DoTs. So that removes 31 Madness from even being an option for PvE.

Due to the RNG aspect of the build, DPS numbers vary by 1-2% depending on your luck. Stat weights are transient due to curves but at this gear level it appears that all stats are fairly close together, except crit which seems to always come in lower.

[CaseyTheRetard]

(12-20-2011 02:23 PM)Kor Wrote:  From a pure DPCT standpoint, the ability priority is clear:

Crushing Darkness > Affliction > Death Field (w/ Deathmark) > Chain Lightning (if specced) > Force Lightning > Creeping Terror > Shock (w/ Chain Shock)

DPCT alone is not sufficient for determining ability priority in a system of multiple competing cooldowns. Resolving priority conflicts requires you to look at the actual DPS contribution from each ability to know which ability to delay casting so as to incur the smallest damage loss. The classic WoW example of this is the comparison of Curse of Doom with any other warlock dot: despite having the highest DPCT, Curse of Doom was always the lowest priority since losing 1.5 seconds of any other dot ticking was a greater damage loss than 1.5 seconds of CoD.

For this analysis specifically, I see that Affliction with 2/2 Exsanguinate lasts for 21 seconds vs. Death Field's 15 seconds. That gives DPS of 1890/21 = 90 for Affliction (1740/21=83 for the 0/17/24 build) vs. 1730/15=115 for Death Field (1550/15=103 for 0/17/24). For a first-order analysis (i.e., ignoring Affliction's possible contribution through Lightning Barrage) prioritizing Affliction over Death Field will cause a net loss of around (115-90)*1.5 = 37.5 (respectively (103-83)*1.5 = 30) damage when they conflict.

(12-21-2011 01:39 PM)Kor Wrote:  The math is pretty simple, DPCT = (Stdmin/max * StdHP(level) + FBonus * Coefficient * NumTicks * Modifiers) / CastTime (1.5 if instant)

Shouldn't that be DPCT = (Stdmin/max * StdHP(level) + FBonus * Coefficient) * NumTicks * Modifiers / CastTime ?

[Kaedis]

Quote:DPCT alone is not sufficient for determining ability priority in a system of multiple competing cooldowns. Resolving priority conflicts requires you to look at the actual DPS contribution from each ability to know which ability to delay casting so as to incur the smallest damage loss. The classic WoW example of this is the comparison of Curse of Doom with any other warlock dot: despite having the highest DPCT, Curse of Doom was always the lowest priority since losing 1.5 seconds of any other dot ticking was a greater damage loss than 1.5 seconds of CoD.

This is true, and I've not yet covered conflict resolution because I wanted to have solid sim data to back it, and I'm still working on the sim.

Quote:For this analysis specifically, I see that Affliction with 2/2 Exsanguinate lasts for 21 seconds vs. Death Field's 15 seconds. That gives DPS of 1890/21 = 90 for Affliction (1740/21=83 for the 0/17/24 build) vs. 1730/15=115 for Death Field (1550/15=103 for 0/17/24). For a first-order analysis (i.e., ignoring Affliction's possible contribution through Lightning Barrage) prioritizing Affliction over Death Field will cause a net loss of around (115-90)*1.5 = 37.5 (respectively (103-83)*1.5 = 30) damage when they conflict.

Actually, to be precise, it will cause a loss of (37.5*1.5) = 56.25 damage per conflict if Affliction is is prioritized. That is not equivalent to a 37.5 dps loss, it's just a 37.5 dps loss over that 1.5 seconds.

Quote:Shouldn't that be DPCT = (Stdmin/max * StdHP(level) + FBonus * Coefficient) * NumTicks * Modifiers / CastTime ?

My mistake. That's how I calculated it, just didn't notate it properly

[CaseyTheRetard]

(12-22-2011 12:56 PM)Kor Wrote:  Actually, to be precise, it will cause a loss of (37.5*1.5) = 56.25 damage per conflict if Affliction is is prioritized.

Check your units: that's 37.5 damage, DPS has already been multiplied by the GCD.