Perfect, thanks.

96 + 80 haste rating / 15.76 = 11.2% haste

If the calculation is additive, total haste would be 30% + 11.2% = 41.2%

2.0 / 1.412 = 1.42 sec cast

Now with the multiplicative calculation.

2.0 / 1.3 / 1.112 = 1.38 sec cast

Looks like it works the same for spell haste as physical hastes.

EDIT: For clarity, the reason drums and gear haste stack additively is that they're both haste rating. All haste rating goes into a single multiplicand while any separate spells that grant a modifier are multiplicative on top of that. For you guys there's not many, but for example as a troll hunter I can get gear haste, drums, haste potion, troll berserking, bloodlust, quiver haste, serpent's swiftness, quick shots, and dragonspine trophy -- and only the first three are additive since they're rating, rather than percentage-based.

There's no extra bonus to LB over CL. They're both multiplicative with a 2.09 factor assuming you have the meta gem which we've all got. My guess is that it has to do with the fact that if you use CL you've got more spell casts per minute, and therefore can possibly suffer from more latency penalties per minute just because you've got to press more buttons.

Kasi:

That link you posted shows a simulator which is (even by the author's own admission) possibly incorrect in some way. In addition, there is no math to support what he's saying.

All buffs, Convection, Misery, etc are applied after the fact and are multiplicative damage bonuses, and therefore can be completely and utterly ignored because they just fall out of the equations. Lightning overload is just a 20% chance to proc 50% damage which also crits at your crit rate. In other words, it's ALSO just a completely multiplicative bonus on the damage you deal with the given spell, and as such can be ignored. This is why I flat out ignored all of those in my two very simple equations -- they can be ignored.

Even the argument that maybe you get more lightning capacitor procs is flawed. You can ALWAYS add CL to your rotation to reduce your average cast time, unless you're already at the global cooldown. And with a fixed crit rate, reducing your average cast time increases your procs, and therefore the capacitor's total damage done. Therefore, adding CL to your rotation will always increase your Lightning Capacitor's damage output in the long run. And based upon what I previously posted, by definition Chain Lightning is higher damage per second unless you're at the global cooldown or using an un-achievable amount of +damage or +haste. Therefore CL is always worth using.

The only plausible argument for the fact that a rotation including CL is lower dps is that you'll have higher latency effects due to more button presses, but he explicitly stated that his simulation ignored latency. Unless you can give me an actual reason why the simulation reports something different than what I'm saying, I don't buy it. There has to be a reason other than faith or 'so and so's simulation says so'. If our dps was hard to model, I'd agree with a simulation, but it's not. We have two spells. We're not shadow priests or hunters or warlocks or any other class with a lot of choices -- we have 2.

Come up with a reason for the discrepancy and I'd be inclined to believe you. Without that reasoning, the only course of action is to believe that the simulation is buggy.

All of that said, it's not like Noobshock is even stacking haste so your suggestions arguably don't even apply. The question of how he's doing so much more still stands. After looking more carefully I also realize he was flasked, using oil, and food whereas I was using merely elixirs and food without oil. His fights are also shorter than in my guild, leading to longer bloodlust uptime and therefore shorter average cast time and higher dps. There are too many differences there to accurately say, I'd wager. This week I'll actually flask up and see how it goes, assuming I raid.

I wouldn't mind if LB spam was higher dps, I just want to understand WHY. 'It just is' doesn't sit right with me.

Posting a WWS from our recent ZA clear. While these encounters are straight forward, with 3 pure tank/spank, i thought the DPS benchmarks would be of interest. No boomkin unfortunately, but had a Spriest.

Wow Web Stats

I broke 1900dps on Halazzi because I could Chain Lightning his pet. I did break 1800 dps on Akil'zon the Eagle boss which involves some movement, and I would compare him Teron Gorefiend with the raid dmg causing the typical spell pushback.

I don't see a reason for the ZA bosses to have any lower resist rates etc, so I thought this was a semi valid benchmark.

MY sim results did not show that LB was better than CL LB LB LB rotation it showed that hasted LB spam was about as good as non hasted CL LB LB LB. I DID NOT RUN ANY TESTS OF HASTED CL LB LB LB.
This is what I ran:
Set 1:

Damage: 1429 without both trinkets. Could we use TLC and Hex Shrunken Head as the two trinkets? HSH should be 88 passive damage which would bring 1429 up to 1517, and TLC well I'm not sure. This 1429 value does not include 85 dmg totem as well.
Crit: 503
Hit: 153
Haste: 0

Set 2:

Damage: (1334 without both trinkets, 1422 including 88 from HSH. Also does not include dmg or haste from totems equip slot.)
Crit: 403
Hit: 134
Haste: 209

So same thing as before.
Set 1 : LB spam with Ancestral Totem
Set 1: LB spam with Skycall totem
Set 1: LBx3, CL with Ancestral Totem
Set 2: LB spam with Ancestral Totem
Set 2: LB spam with Skycall Totem

Stop saying ridiculous things like my sim said LB was more DPS than CL. I never said that.

That simulation wasn't the only one that was done with the values I gave. Daidalos had one too and came up with similar numbers. I too was surprised at the numbers.

But lets see with the numbers I gave for top of the line gear with no haste. 1429+85+88+80 (lets go with crusade card instead of TLC, I don't want to deal with the complexity that TLC adds)

1682 damage. .794x1682 = 1335 + 611 = 1946 damage on an average LB, discounting misery and crit rate.
786+.64x1682 = 1862 damage on an average CL.

Add in there now that LB will be 5% more because of 4p T6 which makes it 2043 vs 1862.

Now assuming no latency at all. We get this.

2043 dmg/2 seconds vs (2043x3)+1862 per 7.5 seconds.

This comes out to 1021.5 vs 1066 dmg per second.

Now we showed from before that we can lose 95 spell damage and 100 crit rating to get 209 haste rating. Or 1.133 haste as Tsook showed. Take 2 second LB and divide by that and we get 1.76 second LBs. We also now get 1.32 second CLs.

Now lets go back to this calculation.

1587 damage = 611+1587x.794= 1871 average LB x 1.05 = 1964 LB
1587x.64 + 786 = 1801 average CL.

Now we have LB spam at 1964/1.76 = 1116 damage/second
If we do LBx3, CL we get 1964x3 + 1801 = 7693 spell damage. Over how much time though? CL still has a 6 second cooldown. The GCD is still 1.5 seconds. You finish casting CL and you have .18 on gcd to go. You cast 3 LBs and almost have .75 seconds to go before you can cast CL again. You really didn't gain any time here. This is why some people have gone to a 4LB/CL rotation. But with a 3/1 rotation you get 1025 dmg/second.

If you go to a 4/1 rotation that its 4x1964 + 1801 = 9657 damage over (4x1.76)+1.5 seconds. This is an 8.54 second cycle. This works out to in dps to 1130 dmg/second.

Thus where the gap was once 45 damage it is now 14 damage. Also note the dps in both cases, LB spam or LB/CL rotation is going up.

Lets try one more thing. Switch totems. Go from 85 damage to 79 haste rating. This one to me is important because in the last one we got more haste by dropping dmg and crit (which I've not factored in) this will only factor in dmg vs haste.

1502 damage = 1894 average LB dmg
1747 average CL damage.

Haste rating: 209+79 = 1.1827 haste or 2/1.1827 = 1.69 speed LBs.

1894/1.69 = 1120 damage/second.

So it keeps inching up. Interesting it showed that going from the Teron totem to the badges totem was a dps upgrade just spamming LB, even though its 79 haste rating vs 85 spell damage.

Now at this point we have lost 4.5% crit though. That's not the best thing ever. Factoring in CSD it probably works out to about losing 5% crit damage. Did we gain anything in response to losing that?

So we have this. Currently in game I have 401 crit rating. This is giving me a crit rate of 40% with talents and totem figured in. Funnily enough this is about the same crit rating as my non crit gear set. Thus the choices are 44.5% with the crit/dmg setup vs 40% without.

Dmg/Crit Gear:
1021 dmg/sec LB spam @ 44.5% crit rate = 1475
1066 dmg/sec LB3/CL @ 44.5% crit rate = 1540

Haste Gear with Teron Totem:
1116 dmg/sec LB spam @ 40% crit rate = 1562
1130 dmg/sec LB4/CL @ 40% crit rate = 1582

Haste gear with Haste Totem:
1120 dmg/sec LB spam @ 40% crit rate = 1568

Now we have figured in crit. We haven't figured in Misery but you are correct in that there is no need to do this. Also no need to figure out LO since it is just basically a flat chance and works out to just 10% more damage. One could figure in something like TLC and try to figure out how the balance of a 4.5% higher crit rate works against a faster spell cycle. The thing we are seeing though is that as haste effects (and ratings) go up, LB is overtaking CL. Mainly due to the fact that as you bring LB down closer and closer to 1.5 second cast time, it is the superior spell due to the GCD and that it does hit harder.

However heroism does complicate things because it only takes a very small amount of haste rating to have heroism drop your LB below 1.5 seconds. However this is countered by the fact that using CL under heroism is pretty stupid. Even with no haste a CL under heroism is wasting .35 seconds off the GCD. Lets say we did have that high haste rating with heroism up. You have LBs at 1.3 seconds vs CL at .976 seconds. At that point its pretty silly to use CL. They both are "taking" 1.5 seconds. LB hits harder, thus use it. The only counter I can think to this is that if you add some amount of latency to each cast, it might be considered better to have 2 second casts seperated by .2 second intervals as compared to a 1.7 second cast seperated by that. At that point it is really up to the latency and speed your connection/computer runs at.

This all leads me to confirm my thought that going with haste/dmg gear is the best options we have now. Regardless if you lb spam or use the rotations. Also if you do use the skycall totem for the haste it gives using LB spam, do switch to Teron totem before heroism to get the most out of it. Other factors like movement and such might make you just want to use the teron totem anyway, given that the differences between the 2 is fairly minor.

Well after the math I worked out here, you might want to see how LBx4, CL works with the haste set. I figure the numbers will be better than LBx3/CL if the math I worked out was any good at all. :p

I'm going to see about doing a graph of DPS, LB & CL vs haste rating (again for X damage/y crit) over the weekend (work is too busy atm to slack off doing TC stuff).

I run 1.73 bolts, and Ive had better DPS results from 4-1 than pure LB. I just picked up my Skull of Guuldan this week, and with skull up Im pushing 1.58. We only raid two nights a week to drop MH/BT, so I wont get to really test out the difference till Tuesday. Im still gonna stick to 4-1, and keep with the pure LB spam on Lust or Skull use.

I've searched the thread and didn't find an answer to this so forgive me if it has been mentioned but I was curious as to how good you guys see the trinket [Fathom-Brooch of the Tidewalker] that drops from Fathom-Lord Karathress in SSC.

On the face of it, the trinket appears to be pretty awesome, essentially giving a chance for Lightning Bolt to give mana back instead of costing it and Chain Lightning to cost half mana (according to the wowhead chat it's a 15% chance with a 45s internal cooldown).

Is there some reason that it's not listed on page 1 of this thread? Is it, in fact, nowhere near as good as it may seem or is mana longevity simply not enough of an issue by that stage of progression to make you lose the other trinkets that are available to you?

It may be viable if you are starved for spriests, but even then probably not. I can count the number of times I havent had an spriest in group in the past six months on one hand so Ive never really given it much thought. Back when I was doing t5 content I had the 4pc t4 OP bonus, and dps trinkets were clearly superior. That and it did not drop once before we phased out the zone for t6.

I'm sorry if this has been asked - I couldn't find a clear answer in all 31 pages of this thread. I use Pawn to compare gear and help to determine if I can upgrade or not. After reading this whole thread (that took a long time!), I found one post that gave clear stat weights, but it was for pvp. Can someone refer me to a post that shows the pve mid-raid stat weights in a number format that I can plug into Pawn, please?

Thanks for the reply. I believe you are refering to these numbers from the first post:

unfortunately, I can't plug these numbers into Pawn as is, and it isn't a complete set of ratings. What I'm looking for is something more like this:

spell damage: 1.5
crit rating: 1
spell haste: 1.25
intellect: 2
etc, etc.

There's more stats than just spell damage, crit, and haste. I need the stat weights for those as well.

EDIT:
An example of what I'm looking for is in this post (unfortunately that one is for PvP). I'm not asking for the string unless someone wants to provide it, but the stat weights are embedded in the string.

Those depend on your gear and buffs. I suggest you use the spreadsheets in Bink's sig to determine your personalized values.

Well 80 int is one crit. The amount of mana and mp5 gain from that int may or may not be usefull to you. But even if you have 2 spreists in your group and can spam CL all day without worrying about mana the crit is a dps boost. So to determine int -> crit rating ....

80 int -> 1 % crit -> 22.1 crit rating.
3.62 int -> 1 crit rating.
soo lets normalize for the given example in terms of spell dmg.
3.62int -> 1 crit rating -> .6588 spell dmg

1 int = .181988 spell dmg or 1 int = .2 spell dmg with kings
1 crit rating = 0.6588 spell dmg
1 haste rating = 1.225 spell dmg

mp5 isn't all that good for us. I'd rather have crit than mp5 anyday.
Stam is nice but use your own personal weighting on that.

Bink,
Was wondering if with the gem selection, instead of having a row for each color maybe 1 Meta row and 4 gem rows. Also have a column for each equipment Slot at the end of the gem selection that you could pick how many of each of that gem are on each piece so you can get more exact stat numbers from gems. Then you could also do Meta verification based on the quantity of each gem with their color requirements.

Just a thought.

I don't have proof to back this up, only my experience. If you have mana issues, that is you have no dedicated Spriests for the caster group or he just sucks, then this trinket can be good or bad depending on what other trinkets you have. For example, if you have TLC and Sextant, and are having mana issues, the best way to remedy this is use the Totem of Pulsing Earth. This relic is far more MP5 while spam casting than the Fathom Brooch is. The Fathom Brooch has a internal 45sec cooldown, so even if it procs INSTANTLY as soon as the Cooldown is over, at best this gives you 37mp5. The Totem of Pulsing earth assuming spam Lbolts at a 2.0sec cast time is ~62.5mp5. Losing the +55dmg totem is not as bad as losing TLC, but might be comparable to the Sextant /shrug. Still larger mana gain from the Totem of Pulsing Earth.

The BT Rep reward trinket is the best mana gain trinket for Elemental shaman, however by the time you get the reputation from BT for the trinket, the mana woes should be a thing of the past. Kinda sucks like that.

Spell dmg trinkets get BETTER with haste and crit as well. Crit trinkets get better with spell dmg and haste. Haste trinkets get better with spell dmg and crit. I don't really see anything unique there. TCL is a great trinket but I don't see anything special about it ability to scale. Someone wanna post a scaling comparison of dif trinkets? Offhand I woudn't think it would scale any better.

To me just saying it is the same thing does not mean you can just replace it with another. In respected science, one does not simply replace one real correct variable, with an artificial oversimplified one just to save time. The impact of changing conditions to make things easier often invalidates findings, or atleast forces someone to follow up on the results knowing that they were skewed. Thats all. Also as a note, +dmg trinkets scale much differently because there are more variables that affect it (4pc T6, CSD, Stormstrike). In my mind this means TLC would scale differently, although I don't know in what direction, which is why I like simulations that include it. =)

I was mostly responding to the implication that TLC scaled with haste as somehow being distinctive in some way. I pointed out that pretty much everything scales with something. TLC mechanics get messy when including storm stike.. assuming there is a constant number of SS buffs up for a given time haste and TLC woudn't scale as well. I haven't updated my sim in awhile I guess I could work on modeling TLC with 2 SS every 10 s. I was planning on giving it a GUI and hosting it somewhere first ..hrm choices choices.

Now we're getting technical. What I might end up doing is putting another socket set selection on the front page, so if there's an item with BYY, and you throw RRR in there, you'll be able to specify that your gems are RRR.

Looking at it, I think that's the way to go, but I'll just need to tweak the socket bonus part, as it's referencing the item listing at the moment.

Plus I really need to get more graphing done for the main page

You might want to add the Pawn values to the main article, from the ENH post I found the weighted values most useful and having real-time values on loot makes it much easier for me to keep track of drops. It also helps when other Shaman ask me which is better X or Y from dungeon gear.

Pawn Addon Forum :: Index

Thanks, I'm sure those numbers are accurate, but I don't know how to use them in Pawn. Let me see if I can translate those into numbers for Pawn...tell me if I'm close. I guess we're using spell crit as the baseline to determine all the other numbers, correct?

spell crit: 1
spell damage: .6588
spell haste: .80703 (.6588*1.225)
intellect: 0.2762 (1/3.62)

I'm also guessing, based on other caster experience, that spell hit would be a high priority until you hit the cap (which should be fast). Am I close on this one?