The outcomes of these changes seem to be dummed down to a few simple things. Feel free to tell me if I'm wrong, I'm just trying to make a short summary:

-Shamanism will change the EP values to make spell power more valuable and as mentioned before, takes chain lightning out of the single target rotation.

-The change to Elemental Oath ups the value of crit by a decent amount, at least to the point of always having Elemental Focus up which would be more than 33% at least.

-The changes to Elemental Mastery make it worth much much more and a great cooldown to keep using. My question is whether or not to use it as much as possible to keep it always on cooldown or to save it for heroism or somewhere in between.

-AoE was buffed. Will we make a rotation of Fire Nova and Magma or just magma? Either of them combined with Chain Lightning should be a decent viable AoE. As soon as the PTR finishes installing I'll find a way to test this.

-This looks like the elemental build at least right now. There is currently a bit of debate over whether Flametongue or Lava are better Glyphs

If there's anything more the skimming reader should know please feel free to tell me, I'll add it to the post.

Shamanism does not make CL obsolete in a single target rotation. CL will still be a higher DPSC spell than LB just not by so much and there is doubt as to whether it is now worth the mana cost. However, the biggest reason by far to use CL in a rotation is to ensure you get as close as you can to that 8 secs of fillers between LvB casts. Here it could still be invaluable depending upon your haste rating.

Short of having 100% crit, there is no way to ensure that EO is always up. Due to the poor scaling of crit generally for us and the large number of crit buffs/debuffs available in raid, it doesn't appear that it is worthwhile stacking crit just for the EO effect - it just makes crit less undesirable than it was previosuly.

When to use your EM will depend to a great deal on the fight. If the boss has a mini enrage where you would usually pop bloodlust and nuke down hard (e.g. Maexxna) then I would save the EM/BL combo for then. If it doesn't then there is no reason why you shouldn't adjust the timing of your BL to bring it in line with when your EM is ready.

It is likely that the changes to to Magma won't buff it to the degree that raid DPS will increase by you using that rather than ToW. There may still be times to use it like when your other AoEers tend to get too much aggro from their AoE or where the Adds spawn in a specific locations and you can have it 'precast' ready for the mobs. If you are going to use it make sure your resto shammy is aware and has his FT totem down first! Personally, I tend to stick to Magma alone and leave Fire Nova out altogether. Partly this is because it's a pain juggling the two, partly it's because the mobs tend to die too often before my Nova decides to explode. I haven't done the calcs to determine which would actually produce the most damage though.

I just can't see reason why elemental want drop magma over tow. If there is resto shaman dropping flametongue then why wouldn't that resto shaman drop magma totem. Resto magma + ToW is better than Elemental magma + resto flametongue. And Blizzard has to fix Magma totem aoe cap or else it's pointless. It's cap now 1500damage per tick which is same value than at 70lvl.

Im guessing u confused EO with EF, but EO has already a 100% uptime due to LaB(Lab crits every 8 secs, EO has a 15 sec duration). I'll agree with it not being worth stacking crit though, and with the EM change crit is gonna go through the roof

As for EM being timed with bloodlust i have noticed that on its own u will have to w8 for the new bloodlust as long as the fight has space for a second one. however, if u factor in the EM glyph u will be dead on time to get an extra EM sqeezed in between the 2 bl's and its prolly this that gives extra weight to picking EM as one of the glyphs to use

Thanks for the quality information listed here. I've been wasting my time trying to get useful information on the wow forums, but they mostly seem to be rants fest. I know this post really adds nothing to this current thread, but I just felt compelled to say something. Just wanted you to know the silent readers of this forum are appreciative!

Yes - sorry I did indeed mean the buff from EO that is active when EF is up. My mistake.

With regards to the timing of the BL with the EM I guess I have it a bit easier than most because as our Raid Leader I get to decide when the BL gets popped anyway. For this reason I won't be doing any maths myself on whether the buff to damage on having BL/EM synched outweighs the nerf from taking a glyph that is weaker overall. It does seem to be a little bit contorted as far as ways to increase DPS go though. Might be best to just talk to your RL and get him to allow you that little bit of flexibility needed in timing the BL rather than nerfing your overall DPS for the sake of 30 secs.

I wish item feedback was turned on for PTR. I think another drop in the lack of dps bucket is the misitemization of our tier gloves and the lack of a DPS related 2 piece bonus. At least 5 other tier sets I've glanced at have a 2 piece that contributes to DPS even if only situationally.

The problem with the comparison with Enhancement is that crit is a better stat for enhancement than haste but the position is reversed for elemental. So, given the choice between crit and haste gear, the enhancement shaman has the clear choice of crit improving his DPS more and increasing his flurry up time. The elemental though must the choice between increased DPS from EF up time from crit and increased DPS from haste.

To my knowledge, the latest simcraft model includes all the latest elemental changes including the EO/EF synergy and still shows haste as 34.5% better point for point than crit given Naxx 10 level gear.

Here's some of my analysis of the forthcoming changes:

Buffed Stats
removed averaging of trinket procs and bloodlust and set bonuses to provide a more generic baseline.
Spell power: 2447
Crit: 44.21%
Haste: 476

Applied to a priority cast rotation until 1 million damage (based on current damage).
Rotation: 16xFS, 29xLvB, 47xCL, 110xLB
Hasted Rotation: 13xFS, 26xLvB, 38xCL, 127xLB

I then analyse the result of adding +10 spell power, then +10 crit to that rotation.
For haste I'm using a value of +1000 haste and allowing the priority cast rotation to change. This is to average the value of haste over a number of rotation changes, and cover the scope of trinket procs.

Glyphs
Analysis of losing either Flametongue or Lava Burst glyph.
Based on Patch 3.0.8 modelling for 1 million damage
Rotation: 14xFS, 25xLvB, 40xCL, 97xLB


My Findings:
* Overall patch gives us a fraction under 14% dps increase.
* More importantly it significantly increases the scaling value of spell power for us.
* Crit gets a bit of a boost in importance but still firmly our worst stat.
* Haste improves but much less so than spell power.
* CL still provides better dps than LB but the difference is much less than it was. CL still has a place in rotations but the penalty for skipping it will now be less severe.


Notes:
Modelling for elemental oath changes is based on averaging the clearcasting up-time over a base damage increase of 10%. Uptime does not take into account the difference in crit chance between lightning and non-lightning spells. Currently based on the worst-case non-lightning chance so there may be a slight slight boost also from that.

Posted for general interest and for seeing how this agrees / disagrees with any other analysis out there.

You seem to be missing the extra 6 seconds on Flame Shock from the glyph, for starters.

Is there anyway you could calculate that without CL in the rotation and the 4% dmg increase of LB added in there? Im curious to see how much DPS we lose if we drop CL from the rotation while adding the new glyph in.

**EDIT**

Could you also post the mana consumption with all the numbers you have above too? Im interested to see what kind of mana efficiency we have with and without CL in the rotation. Thanks!

Should have specified what glyphs I was including:
Glyph of Flametongue weapon (rolled into buffed stats)
Glyph of Lava (+10% LvB dam)
Glyph of Flame Shock (+2 ticks of dot damage, FS not consumed by LvB).

Flame shock damage calculated in two parts - direct damage and dot damage.
Calculated as follows (3.0.8):

FS Direct: 525 + (2447 * 0.225) = 1148.87. Adjusted for chance to crit gives Average Damage: 1656.79
FS Dot: (1376/6) + (2447 * 0.155) = 651.58 (x6 ticks + 0 chance to crit) = 3909.5
The sum of these gives Avg FS Dam = 5566.28 (4462.02 dps)

Calculations seemed to match when this was for current values - but the 1376 is not tooltip observed but calculated based on the projected 60% dot damage.


What I did do wrong in the analysis above though, was forget to roll back the difference in CL cooldown - So the values for current were generated based on a 3.0.8 rotation with 4.5 sec CL cooldowns, instead of 3.5 sec.

Also, as pointed out above I don't yet have the new LB glyph factored in.

Am I correct in assuming the new proc from Elemental Oath would be additive to the new Lightning Bolt Glyph, or would it be multiplicative?

Thanks.

Multiplicative.

Hi Bink and gang,

I noticed this weekend the Procodile mod was mostly fixed. I wanted to create a baseline to compare the haste proc totems. Blizzard had hotfixed the Skycall at least once in tBC, so I wanted some data to refer back to from time to time.

To my surprise, by default it started tracking my clear casting uptime. It seems to be a little under 60%. WWS shows my LB crit around 45.5%. This sounds quite close to what has been discussed here. Is this going to scale very well, or are we going to be hovering around 60% most of the time?

I think I need to write an essay on Clearcasting and how most mods/log analysis apps get the uptime wrong.

The problem with working out clearcasting uptimes is that it goes up and down a fair bit, and that it starts on spell hits, but is consumed on spell casts.
So what does this mean?
Firstly, the timestamps, and indeed order of entries in the combat log are not accurate. In a general overall view it can be mostly accurate for longer duration procs, but for something that can go up and disappear in 5 seconds, this sort of variability can affect how uptimes are calculated.

For example, I've had a talk with giansm (the guy behind StasisCL) to try to figure out how we can get better uptime figures.
Firstly, here is the log that he split off for Patchwerk http://stasisguild.org/cc.txt
The parsed results were as follows:

Ability	Total%	Hits	Crits	Ticks	AvHit	AvCrit	AvTick	% Crit	Avoid
Lightning Bolt (r14)	454,753	62.5	65	41		2951	6412		38.7%
Lava Burst	173,768	23.9		22			7898		100%
Flame Shock (r9)	70,394	9.7	6	7	70	1224	2501	650	53.8%
Lightning Bolt (r14)	29,236	4.0	10	5		1486	2874		33.3%

He ran a new script looking for Clearcasting, and got the following results:
So that looks about similar to the 60% parsed uptime.
However, when we look at the log we see some odd stuff.
If you read it correctly, I gain CC from the FS crit, refresh it on the LvB following, but it "fades" after the next non-crit LB. Logic tells us that the second LB should also be affected, but any log parsing will not see it as such. The log will say that 2/4 spells were affected, with a 75% uptime, but we know that 3 of the 4 were, at a 92% uptime. You can also see here that clearcasting fades before the last LB starts casting, even though we know it will be affecting that cast.

I've still to finish off the direct comparison between uptime calculations, but if I count the spells affected by clearcasting, it comes up as 80%, rather than the 60% uptime listed. This is why assuming that WWS/StasisCL uptimes are accurate for a short duration charged base buff like Clearcasting is a very bad thing. This is also why I believe that the 3.0.8 changes will be more of a buff than most people assume, because of the higher "perceived" uptime.

There's also a blog entry by Pewsey that covers similar material.

Firstly thanks to Bink for a genuinely useful analysis. Slightly unnerving of course - if the combat log is that mangled, it makes you wonder what else might not be working properly. Anyway, I guess we can consider the mystery solved, and get on with looking at what the buff might mean, rather than why TC wasn't equating with log analysis.

Personally I will be using the (1-crit)*(1-crit) method for calculating the benefit of the new clearcasting buff, (where crit = a weighted average of individual crit rates for all spells used). There are any number of more complicated ways to do it, but I am not convinced they will give much more in the way of accuracy.

Hmm i simulated the uptime of clearcasting and came to the result that it is nearly 82% at 55%crit ( raidbuffed and including Lavaburst&elemental mastery )
The programm simulated 25000 5min fights .. i think it can't be so wrong.... but we'll see when the patch is online.


Used simulationcraft - Google Code

See previous posts for formulae on calculating uptimes with a 100% crit chance on lava burst
Most things that the combat log shows are correct, it's just anything that's time sensitive, like this and looking at time between cast end/starts, is where it falls over. As for the uptime calc, you'll come in on the low side, but I always go on the "it's better to under-estimate than over-estimate" philosophy.
[quote=Kegsta;1032388]What do you guys think of this for a build next patch for both PvE and PvP.

Yeah, deadmon seems to be using a similar basis to me for calculating uptimes, that and with a sim the timestamps will be correct

I've pulled out some equivalency points from my own spreadsheets. I'm throwing this out there just to get some discussion going really.

These are a first attempt, and I haven't modelled the new Elemental Mastery yet. Shamanism and Elemental Oath are included though.

Live
Spellpower - 1.26
Crit - 0.60
Haste - 1.14

3.08
Spellpower - 1.48
Crit - 0.79
Haste - 1.31


The haste value is probably a little high, caused in part by this being modelled around my current gear level.

I think though that they firmly support what most people had worked out here already; a) crit will be a better stat, but still worse than haste and b) spellpower becomes probably even more attractive as our prime stat.

I'm happy to go through the methodology, but to be honest I am more interested in whether others are feeling weightings will move in this general direction than dissecting the TC,

For Clearcasting I'm going to try to put it in layman's terms for those who aren't mathematically inclined, feel free to correct me if I'm wrong.

We assume clearcasting is only removed by being consumed(2 spells).

We don't care about the percentage of time clearcasting is up, only the percentage of spells that it affects.

The TIME between lava burst and what not is irrelevant, we only care about the NUMBER of spells cast between lava bursts.

The 2 spells cast after lava burst are always affected by clearcasting.

For calculating how many spells cast between lava bursts:

Assume 15% Haste.

We have 9.275 seconds of casting time between lava bursts.

Lightning Bolt is a 1.7 second spell.
Flame shock is a 1.275 second spell.
Chain lightning is a 1.275 second spell.

You are going to have 2 CLs in every other rotation of lava burst and 1 CL in every other rotation.
You are going to have 1 flameshock in every other rotation (18 seconds)
This means you can get 2 1.275 second casts in lava burst cycle and the rest is 1.7 second lightning bolts.

9.275-(2*1.275)=6.725
6.725/1.7=3.955 or 4 casts.

On average you are going to have 6 casts between each lava burst and the lava burst itself so 7 total casts we are going to work the numbers for.

2 casts by the nature of lava burst will have a guaranteed elemental oath so we need only solve for the other 5 casts uptime.

You want to cast your Flame shock directly after the lava burst whenever you can because Flame shock benefits the most from spellpower outside of Lava Burst, and more importantly has a 5% lower crit rate so you don't want to lower the crit chance of the 2 spells you cast before lava burst.

Now what is the chance that one of the last 2 spells crit?

You can solve this by your crit rate. Lets solve for 50% crit in this instance.

To figure out the odds that you did NOT have a crit you simply multiply your chance at NOT having a crit(1 - your crit rate). With a 50% crit rate this is .5*.5= .25 or a 25% chance you did not crit the last 2 spells. With a 40% crit rate it is .6*.6 or a 36% chance you did not crit the last 2 spells.

So with a 50% crit rate we can solve the following cast sequence.


1st Cast 100% Uptime
2nd Cast 100% Uptime
3rd Cast 75%
4th Cast 75%
5th Cast 75%
6th Cast 75%
7th Cast 75% (Lava Burst)

Notice the first 2 will always be 100%, this means that our crit rate only affects 5 spells when calculating for clear casting and even with a 0% crit rate, Clearcasting will be active 2/7 or 28.5714% of the time no matter what.

This means we only need to solve for the last 5 spells and can thus multiply the not-active percent we calculated above by 5 and divide it by 7, then add 28.5714 to it.

So .75 * 5 / 7 = 53.5714 %
Add 28.5714% and you get 82.14% of spells affected.

For 40% Crit .64 * 5 / 7 = .45714%
Add 28.5714 and you get 74.28% of spells affected.

Now you can make this more complex by figuring in the 5% lower crit rate of lava burst and flame shock but you can roughly average them into your crit rate at 1.5 out of 7 spells average or just subtract 1% from your lightning crit rate to solve for the above.

So the formula I came up with as described above
(1-((1-(Crit-.01)*(1-(Crit-.01)))*5/7+.285714 Where Crit is your lightning crit rate.

Please point out any statistical errors, and yes I know the rotation can get very screwed up with Bloodlust and haste trinkets, but I think if you use Elemental Mastery to get the extra 20% crit during bloodlust it should roughly average out the extra casts you fit between the lava burst rotation.

Hope this helps people who get lost in all the formulas wrap their mind around this.

lightningoverloads also cause clearcasting on crits(but don't consume)... this will lift up the overall uptime a bit.