Some of the math looks good, some is far off. It really depends on the post. If my lightning bolt currently does 1076 damage for 1076 threat giving me 3.26 damage per mana, my new lightning bolt should do 1017 damage for 915.3 threat giving me 3.39 damage per mana.
If I'm threat limited, I can now do 11% more damage before generating too much threat.
If I'm mana limited, I can now do 4% more damage before running out of mana.
We're going to see how these changes pan out on the public test realms. Elemental Shaman will have plenty of time to test out the adjustments and provide their feedback.
Although the math isn't rugged enough to prove such, that comment really leads me to believe that the apparent DPS increase he referred to earlier was caused by a sustainability increase. I wonder if this misdirection was intentional or not. Food for thought, I suppose.
Sigh, I am seeing some issue with that math there. It's nice, but its not factoring in reduced damage on LO procs.
Lets just say your LB hits for 2000 on average. We know that crit and damage are statics in this calculation. Because they pretty much are. Factoring in crit before LO and crit after LO is pointless because its just a static increase. Also not going to count in damage coefficients and such because we're just modelling LO here. Old system will give you 1.05x2000 or 2100 damage.
New system.
Each LO proc has a 20% chance of giving you 50% extra damage LB.
So first LO. 20% chance of giving you a 1000 dmg LB. Should happen 20% of the time
Second LO has a 4% chance of giving you a 500 dmg LB. Should happen 4% of the time.
Third LO has a 0.8% chance of giving you a 250 dmg LB. Should happen 0.8% of the time.
Fourth LO has a .16% chance of giving you a 125 dmg LB. Should happen .16% of the time.
Going to do a couple more here, but at this point there is very little added on to the series.
Fifth LO has a 0.032% chance of giving you a 62.5 dmg LB. Should happen .032% of the time.
Not even going to bother for a 6th.
Adding this all together? (1000+0.2)+(500x0.04)+(250x0.008)+(125x.0016)+(62.5x.00032) = 200+20+2+.2+.02+etc..
Which is 222.2222 repeating.
So this comes out to a 2222/2000 which is well, a 1.1111111111111 percent increase on LB damage. Hence LO can be modelled as a straight 11.11~ % increase to LO, not 12.5%. Hence the numbers posted previously here are wrong.
You are of course working under the assumption that a second LO proc would be based off the spell that procced it, and not the last spell that the player cast. In the latter situation (where the 2nd LO bolt would be identical to the damage of the first LO proc), it is a 12.5% increase. We don't yet know which way it works.
Originally Posted by CheshireCat
Eh, my nostalgia goggles aren't as good as they used to be.
It's possible sure, but since LO procs on spell cast and the second spell cast has been confirmed by people here to be an independant cast (how can LO chain proc now if it doesn't consider the first LO proc to be an independant spell?) I am pretty sure the first LO proc will be considered as the base spell for the second LO proc.
Indeed, I hadn't thought of the second LO proc dealing yet again half damage. This is something that would have to be confirmed on the PTR, since it more than halves the effectiveness of recursive LO procs. For now, I'll leave my math the way it is.
If I'm not mistaken, you counted LO as a +10% dmg increase instead of a +12,5%, which it will be if it's still proccing of itself.
I don't ever recall seeing a double LO proc either. Then again its a 0.0025 chance of occuring with the current LO talent.
To put things into perspective I ran some quick and dirty math. This was using a personal dps spreadsheet. I wanted to see a detailed breakdown of our dps. Crits, hits, LO procs etc. Please let me know if there are any major errors please.
Assumptions:
Single Target
Fight Length 360 seconds or 6 min
Spell Hit Capped
Spell Dmg 1328 (raid buffed)
Crit 42% (to LB/CL with talents/totem raid buffed)
68 Haste rating (4.31% Haste)
.2 sec casting lag
Stand and nuke encounter
3 Lightning Bolt 1 Chain Lightning rotation
Elemental Mastery used on CL every cooldown
Assuming LO does not proc on itself
Unlimited Mana
To account for the 1% miss rate final DPS figures are multiplied by 99%
Partial resists not accounted for
Bloodlust not added
Patch 2.2 Damage DPS
LB Casts 144
LB Hits 84 146268 402.24
LB Crits 60 208954.29 574.62
LO Procs 7
LO Hits 5 8706.43 23.94
LO Crits 2 6965.14 19.15
CL Casts 36
CL Hits 20 34691 95.40
CL Crits 16 55506.29 152.64
LO Procs 1
LO Hits 1 1734.57 4.76
LO Crits 0
Total DPS 1272.73
_________________________________________________
Patch 2.3 Damage DPS
LB Casts 144
LB Hits 84 130332 358.41
LB Crits 60 186188.57 512.02
LO Procs 28
LO Hits 17 13188.36 36.26
LO Crits 11 17067.29 46.93
CL Casts 36
CL Hits 20 30897.14 84.96
CL Crits 16 49435.43 135.95
LO Procs 7
LO Hits 5 3862.14 10.61
LO Crits 2 3089.71 8.49
Total DPS 1193.63
In 2.2 LO accounts for 47.85 dps. 47.85 dps is 3.75% of 1272 dps.
In 2.3 LO accounts for 102.30 dps. 102.30dps is 8.57% 1193.63 dps. The LO change is a considerable jump in threat free DPS, but it is still overshadowed by the coefficient nerf.
1272.73-1193.63 = 79.10 dps or a 6.22% decrease in dps at this gear level. That calculation falls in line what other people are reporting.
With the exception of bosses that have some sort of threat reset or knockback (threat reduction) catching up to the MT is very difficult to do after the first minute. Finding that fine line of dpsing early but not too early that you burst past the MT is something that took some practice. This LO change will alleviate that and be better for the initial 30-45 seconds of the fight. As an elemental shaman I found to be at the highest risk of pulling aggro by dpsing early.
I do find it odd that these changes to overall dps are being made even when elemental shaman aren't consistently very high on damage meters (This applies to BT/Hyjal). I may be doing something wrong, but this from my experience. If we were dominating pure dps classes then I would see the need for the coefficient change.
I wonder if a talent to regain some of the lost +dmg would be out of line. Maybe empowered lightning bolts or something.
Spell Dmg 1328
Crit 42
Hit % 16
Haste 68
Cast Lag 0.2
CL6 LB12
Min 734 Min 563
Max 838 Max 643
Avg 786 Avg 603
Cast Time 1.5 Cast Time 2 1.913
Cooldown 6
CL 2.2 2.3 LB 2.2 2.3
DmgMod 0.714 0.571 DmgMod 0.857 0.714
LB 2.2 2.3 CL 2.2 2.3
Hit Dmg 1741.28 1551.57 Hit Dmg 1734.57 1544.85
Crit Dmg 3482.57 3103.14 Crit Dmg 3469.14 3089.71
I wouldn't mind giving up a bit of damage (100dps at 40%crit and 1400dmg according to the spreadshead with 5k mana every minute(very easy to get)) for trainable manatide.
30mp5s just isn't enough to warrant 100dps.
I don't need more mana, I need more dps to compete with our other casters.
Honestly, if Blizzard wants to tweak the 2.3 change a bit, the finest control they have is to adjust the % damage an LO bolt does. A 60% damage bolt in place of a 50% damage bolt might be enough to 'fix' PvE Ele Shaman DPS.
At this stage, I am assuming that LO cannot proc itself, but this is one of the things I hope we will be able to test on the PTR.
LO has been able to proc of itself all the time, I see this quite often. Once or twice a day, it is not that uncommon. I am 100% sure it is like this (atleast before patch 2.2 doubt it has been changed). Don't have any screens or anything fom it. Even got 4 procs from one bolt one time. Also, the chance for a Chain Lightning to Proc LO is not bigger just because it hits more targets...
Originally Posted by Kasi
Question: Why would LO with the new changes be a 12.5% increase?
Currently LO is 1.05.
My original thinking of just LO proccing once is 1.10%.
There is no way for Omen, that I know of, to determine LO procs from regular lightning bolt casts(or LC procs for that matter.) Maybe they could hack something together by figuring out the expected damage of a lightning bolt, but I'd imagine that would probably fail to work properly under a variety of circumstances.
If I'm not mistaken, I read somewhere that Blizzard would implement a built in threatmeter.
But if they fix it so that there is some kinda message in the combatlog, chances are they will also not consider LO another identical spell any more. Instead it will turn into a speciall spell, like Windfury.
Originally Posted by Bustadawg
Have had this happen some times and yes, it rocks! I'm really hoping LO will still be able to proc off itself cause that will be awesome, but I have a feeling this is going to "fixed". Also, liek someone posted previously, I wouldn't be surprised if Blizzard implemented some kind of hidden CD on LO.
I'm just hoping you are wrong, but I have the same worries.
Originally Posted by Kasi
It's possible sure, but since LO procs on spell cast and the second spell cast has been confirmed by people here to be an independant cast (how can LO chain proc now if it doesn't consider the first LO proc to be an independant spell?) I am pretty sure the first LO proc will be considered as the base spell for the second LO proc.
We don't even know if LO can proc of itself after the patch. Since they are reworking it to not be an identical copy of Lightning Bolt. And LO can only proc of Lightning Bolt and Chain Lightning, not other spells.
Originally Posted by Quasi
.2 sec casting lag
Aren't everyone using /stopcasting macro together with Quartz or a similar addon? I doubt you will have this much lag then. 0.05 would probably be a closer value.
The interesting thing here is the mana cost change. From what he has posted, the cost is dropping to 300 for LB, not the 275 we've been assuming.
I've added that in, and made a change to how LO proc chances are calculated, so that it will give a chance for LO to proc (20%), LO to proc a second LO (4%), that to proc a third LO (0.8%) and a fourth (0.16%)
(If anyone has some decent math on how to boil x + x^2 + x^3.... into a simpler calculation, let me know.)
With those changes, and Ames' stats, we're looking at a slight buff for 3/1 in 2.1 for 3 minutes over 3/0 in 2.2. After that, 2.2 LB spam > all.
I can put out 1200dps when I'm not constantly moving or kiting, and the only fight where threat is an issue so far is gurtogg, but that's more fight specific than anything else. Is the 2nd half of BT noticeably more threat sensitive?
On CL being included in the rotation, that's all fine and dandy, but CL is unusable when there is more than 1 mob around for "oops aggro" or "oops broke CC" reasons. CL isn't exactly care-free.
I have overaggrod on Supremus (that was a massive massive critline tho, something like 10-12 crits in a row), Gurtogg, i have overaggrod on Teron, i have overaggrod on RoS (this is the fight i find most annoying, if i get a LC proc in p2 it completely messes up my stopcasting and i miss a bolt plus i have to wait about 20 seconds in p3 before i can dps or else ill overaggro), i have overaggrod on Illidan, i constantly overagro the infernals in the Anetheron fight, i'm pretty much aggro capped on Winterchill and Naj'entus. All the other fights in BT/Hyjal are so dynamic/specific that its near-impossible to overaggro or you have to have a serious deathwish if you want to be anywhere near tank aggro. By using only tranq totem in all the fights that issue is relieved a bit, but using tranq means all my group misses out on some spelldamage i'm supposed to be giving to them.
x + x^2 + x^3 + ... + x^n = x / (1-x) , assuming |x| < 1.
I'm pretty sure thats the limit of n goes to infinity.
(1-x)(1 + x + x^2 + x^3 +...+ x^(n-1))= 1 - x^n
so 1 + x + x^2 + x^3 +...+x^n= (1-x^n)/(1-x) or (x^n -1)/(x-1) but the first is more convenient for |x| < 1
Subtracting one from both sides gives
x + x^2 + x^3 +...+x^(n-1) = (1-x^n)/(1-x) -1 = (1-x^n -1(1-x)/(1-x) = x(1-x^(n-1))/(1-x)
Now if we take n goes to infinity then the x^(n-1) term goes to zero and we are left with x/(1-x)
I have overaggrod on Supremus (that was a massive massive critline tho, something like 10-12 crits in a row), Gurtogg, i have overaggrod on Teron, i have overaggrod on RoS (this is the fight i find most annoying, if i get a LC proc in p2 it completely messes up my stopcasting and i miss a bolt plus i have to wait about 20 seconds in p3 before i can dps or else ill overaggro), i have overaggrod on Illidan, i constantly overagro the infernals in the Anetheron fight, i'm pretty much aggro capped on Winterchill and Naj'entus. All the other fights in BT/Hyjal are so dynamic/specific that its near-impossible to overaggro or you have to have a serious deathwish if you want to be anywhere near tank aggro. By using only tranq totem in all the fights that issue is relieved a bit, but using tranq means all my group misses out on some spelldamage i'm supposed to be giving to them.
Supremus seems unlikely, like you said massive critline, and to top that off half the fight threat doesn't even matter. Gurtogg it's expected, as it's a heavily threat limited fight by design, so that doesn't point at any issues with the class. Salv and TA on p1 while waiting a few seconds for the 3 tanks to build aggro is usually good enough and actually helps spriests and locks in my group who are also crying for less threat.
I don't know about ROS or Illidan yet so can't help you there...
Naj is usually a non issue for me as the best way I've found to get a quick kill is usually to spot heal a couple of times to minimize healer panic mode before shield burst so I'm losing a little dmg. Infernals in anatheron I'm not sure how you play it but I've been DPSing anatheron for quite a few seconds before we turn to kill the infernal so that eliminates the aggro issue as the tank has a little buffer built up.
If threat seems to be such a big issue at your +dmg/haste though, why aren't you using subtlety to cloak? Every piece matters.
Well subtlety wont help that much, its only 2% and because of the multiplicative nature of -threat, its only actually 0.98% if i have tranq+salv. So it's kinda like -6 to -8 TPS \o/.
Just last week i nearly overaggrod with casting two spells, went from 6k threat to 20k threat in 3 seconds, while the tank was on somewhere like 16.4k. Major LO crittage. With tranq+salv. QQ
The real question though, is are you ready to take a nerf to your main spell coeff, and consequently to your scaling as the gear goes up and up, for some threat reduction.
Imo it's a bad trade, and I have a feeling that even at your gear level if you are somewhat threat limited, it probably isn't by much and you will not see an actual increase if DPS with the new mechanics.
Yeah, just didn't get the time to reenchant my cloak, was hoping for council one but it didn't drop, so probably going to do that tomorrow, still its pretty much a drop in the ocean :P. But i do have problems with threat, even with tranq down and i'd like that blizzard wouldn't fix this by nerfing our DPS. Our caster dps actually talked about this more than a month ago and quite a few mages and other people agreed that in pvp shaman burst was way too sick, but you couldn't just straight out nerf the damage because it would put the shamans usefulness in raids in doubt. And already then they suggested that somehow they should nerf the LB average damage but add nature damage to CoE (are shamans and druids the only ones who think of nature as an element?) or something in the lines of that.
I would be fine with Curse of Elements being nature damage.
I'd be even more happy with a talent in the Elemental tree that does something similar to casting offensive spells to spell damage or even better group spell haste to be similar to enhancement shaman's unleashed rage.
09/28/07, 5:19 PM
Similar Threads
