Daidalos

Can you clean that up maybe label some stuff? Trying to read it gave me a headache. I'm pretty sure you are using different assumptions or messed up on the math but I can't tell from what you posted.

EDIT: here is an example calculation
603 LB base dmg.
concussion multiplier = 1.05
3/3.5 = .857
2.5/3.5 = .714
LO chance in 2.2: (.05 +.0025+....) = 1.052
LO chance in 2.3: (.2+.04+.008+...) = 1.25

603*1.05 = 633.15
LB hit +dmg with old modifier: .857 x 1250 *1.05 = 1124.55
LB hit +dmg with new modifier: .714 x 1250 *1.05 = 937.65


2.2 avg LB dmg with LO: (633+1124) *1.052 = 1848
2.3 avg LB dmg with LO: (633+937) + .25*(316+937) = 1570+313 = 1883
note: I'm assuming LO can proc itself in 2.3 there is currently some debate about this.
If it cannot proc itself.
2.3 avg LB dmg with LO: (633+937) + .2*(316+937) = 1570+313 = 1820.6

 

sleepcontrol

Hmmmm seems you are right, I can't do math today. I am redoing the calculation now,

 

Hidden

Read my post above? I edited in your two errors.

 

Chemdog

I didn't go thru your math, but I can tell you right off the bat some things you've done wrong. 33% crit is not 240 hit and 80 crit. That's 25% crit. 33% crit is 200 hit and 100 crit. You calculated the ratio of LO correctly in the pre-2.3 example (even though there would be 16 not 12 according to your example), but did everything incorrectly in the 2.3 example.

With all the edits people just gave you, maybe you'll get better results now.

 

Hidden

He did the crit percentage wrong in both examples and since crit % doesn't have an influence on the results that's not the main error.

 

Benjaimn

I thought that it was proven that LO can not proc of it self (http://elitistjerks.com/510836-post611.html)? or did I hear that wrong.

 

Chemdog

But it does influence the results from the LO crits, which were done incorrectly on the second example.

And no self proccing has been semi-confirmed. It appears a few people have agreed upon it, but maybe Blizzard intended it and didn't notice the down ranking bug on LB following LO. Then they saw what they thought was a double LO proc and just assumed it was working as intended.

 

Benjaimn

Perhaps, I personally think how spell damage is applied to LO procs fits in well with the blue posts saying it will be a slight dps buff and I forget who it was but the blue poster responded to a thread about the LO math and said "some of your math is right some is wrong", how +dmg applies to LO could of been what we got "wrong".

 

Caldar

I was under the impression that the current theory for 2.3 LO is:

It cannot proc itself and it has a hidden internal CD (which would also hold true for it not being able to proc itself.)

Or maybe I've just read to many different threads on the same topic.



Edit: I thought I saw a test someone did that showed you can get the stated 20% proc only if you space out your casts, but if you spam cast (as we do) it only ends up being like a 12% chance to proc on average.

 

sleepcontrol

I deleted the origional post, because after looking closely at it, there were a bunch of mistakes. Please ignore the first post of calculations. Here is a second attempt with the changes you posted.


Assumptions 33% Crit, 10 minute fight, no lag
Pre 2.3
200 Hits, 100 Crits, 10 LO Hits, 5 LO Crits
(Base Dam*Casts(200 Hits+10 LO Hits)+Base dam*Casts(100 Crits, 5 LO Crits)*2(Crit))*1.05
(600*210+600*105*2)*1.05=264,600
Spell Dam*Casting Coef * (200 Hits+100 Crits*2 Crit Multiplier+10 LO Hits+5 LO Crits*2 Crit Multiplier)*1.05 (concussion)
Spell Dam*(3/3.5)*420*1.05=Spell Dam*378

Post 2.3
200 Hits, 100 Crits, 40 LO Hits, 20 LO Crits
Base Dam*400(200 Hits+100 Crits*2 Crit Multiplier)+1/2 Base Dam*80(40 LO Hits+20 LO Crits*2 Crit Multipler)*1.05 (concussion)
600*400+300*80=277,200
Spell Dam*Casting Coef * (200 Hits+100 Crits*2 Crit Multiplier+40 LO Hits+20 LO Crits*2 Crit Multiplier)*1.05 (concussion)
Spell Dam*(2.5/3.5)*480*1.05= Spell Dam*(2.5/3.5)*504

264,600+Spell*378=277,200+Spell*(2.5/3.5)*504
Spell *378=12,600+Spell*(2.5/3.5)*504
Spell*18=12,600
Spell = 700

 

Daidalos

Seems correct I think assuming LO can't proc itself.
I think I ran the numbers assuming 2.2 can chain proc but 2.3 cannot and I got ~650 spell dmg as the intersect. Above which 2.3 is lower dps.

 

• Binkenstein

Quick thinking gives me the following:

Base LO chance = 20%
If LO cannot proc, then the multi proc chance of 25% is actually 5% chance that a normal LB is LO'd.

Thusly, base damage ends up being 7.5% extra from LO, +dmg is 20% as per normal, rather than 12.5% & 25% like we expected. Unless this problem is fixed, we're looking at a 2-3% dps nerf, which is oddly enough the exact opposite of the buff we thought we had.

 

Chemdog

Hopefully a blue post will comment on the findings.

BTW, Bink, you might want to add that Water Shield will not trigger the 5 sec rule. As anyone tested to see if it triggers the global cooldown? I would imagine it does, but just wondering.

 

Daidalos

GCD yes
global no
(if it costs 0 mana it doesn't trigger. You can rank 1 HW with ssc relic without triggering 5sr)

 

Telkster

srry its a bit off the current subject:

i currently threw up a couple links to

SimulationCraft/Trinkets/Shaman - Shadowpriest.com Wiki

and their corresponding pages for locks / spriests / mages / etc. on our guild site. But when our top ele shammy looked it over she was very surprised to see the lightning capacitor was ranked higher than her silver crescent and was wondering why. Does is have something to do with proc'ing LO? thx

Boo creepy foot doctor! Hooray beer!

 

Daidalos

This is fairly well known. Fast cast times plus LO plus high critrate makes the Lightning Cap the best trinket for an ele shaman. The same goes for arcane mages as well (I always thought it odd our BT mages farmed kara everyweek for that trinket when you would think there would be somthing better in BT/hyjal. There isn't though.).

I'm fairly sure even without LO it would still be the best ele trinket just not AS good. Also with 2.3 and more frequent LO procs the trinket will be even better. Lightning cap isn't the best for every caster its only the best if you have high crit rate and frequent hits (Arcane missle hits, arcane blast spam, lightning bolts etc)

 

Chemdog

Silver Crescent is a straight ~69dmg ((43+(155*20/120))*.714).
TLC is highly dependent on your crit rate, but assuming a 33% crit rate, it should proc about once every 10 casts on average. That's 750/10=75dmg. I'm sure my math on TLC is a bit off (it tends to get complicated including LOs and such) but that would be a rough guesstimate. If you include LO procs, it's even better.

 

Kaideq

I think it's highly dependent on having TLC as from my initial few hours of testing this afternoon,I took a huge blow dmg wise and the dmg became a lot more flucky.

Over 15 2minute boom attempts I dropped 100 ish dps on avarage, I'd say a fair bit more than 2-3%.

Euro PVE realm just went down, but I will do some more extensive testing tomorrow and try to get some people to tank a few raid boss's for me to see how it works out on those.

 

Kasi

Looks like tests are showing that LO has something we've all seen before. A 3 second hidden cooldown. So in chain casting LBs the proc chance for LO currently is not 20%, it seems to be about 13%. This is just making me depressed. Are elemental shamans strong enough that this is seriously an issue?

Definitely going to keep working on my enhancement gear. These changes are ridiculous.

 

Jini

According to Thottbot World of Warcraft: Lightning Overload it is a one second 'cooldown'.

This debuff would also reduce the damage of any other lightning bolt that hit within one second of the last lightning overload proc as well.

I have no idea why the debuff has two charges or if it is actually used. Is there a mod that can be used to show hidden debuffs? If there perhaps a visible debuff associated with LO now? I can't get on the PTR yet as my transfer is still in queue.

 

Kasi

Actually according to that it has a duration of 1 second and a cooldown of 0 seconds. What that means I don't know though.

 

Daidalos

If this information is correct and they are going to keep a hidden CD I think it would need to be around a 30% chance or higher depending on the CD for LO to see any real gains. I don't really understand the reasoning behind the stealth nerf. I'm guessing it was to reduce burst for pvp but didn't take into account that ele pve dps was hardly top notch before.

 

Kasi

Its looking like right now from the test realms that LO is proccing around 13-14% of the time (and never chain proccing) I don't even want to do the math on that, but I'm pretty sure that its a massive dps nerf. Not 2-3%. More like 10%.

Hmm, going back from my calculations on the other thread.

Old LB/LO at 1250 spell damage: 1992 LB damage.
New LB/LO at 1250 spell damage with 14% proc rate (cause of 3 second hidden cooldown): 1874 LB damage.

Thus this is a dps downgrade of around 6%.

Funny enough this might be an upgrade for CL, given that you have 3 targets considering apparently each can proc LO. But for single target DPS this hurts a lot.

 

PSGarak

I'm still sanguine that this is a bug, and not intended behavior. Time to start flooding the PTR forums and in-game feedback features with bug reports until you get a definite answer.

 

Benjaimn

This would account for LO dmg LBs some people get. And I agree that this must be a bug, Blizzard intended us to get a slight dps buff so hopefully we will get this all worked out before it goes live.