The latest version based on the Patch 3.3 PTR is now available for download:

ZAP! 1.1.0 for patch 3.3 (PTR)
ZAP! 1.1.0 for patch 3.3 (PTR) - compatibility version

Changes - Version 1.1.0:

• Implemented patch 3.3 changes:
  • Elemental Mastery now grants haste instead of crit
  • 4t9 now only increases LvB damage by 10%, down from 20%
• Revised fight segmenting model to make it simpler to include new EM
• You can now choose whether or not to use CL during Heroism and EM
• "Rotation Delay" setting changed. It now allows the user to add a period of no-casting in between each LvB cast. LBs will be sacrificed first, then CLs.
• CastDelay now properly accounted for during Heroism
• Statistics:
  • Mana usage more accurate
  • Heroism and EM properly accounted for when figuring out totals
• Judgement of Wisdom and the Reign of the Unliving trinkets now use variable PPM based on the user's stats and options
• "Advanced Settings" now contain the options to factor in miss chance and partial resists (which doesn't affect EP, but it was a frequently requested feature)
• Sandbox's haste calculator looks better and has more options
• Haste from relics is included in the buff display until I can figure out a better way to make it obvious (people tend to forget about relic haste when playing around with numbers)
• Various other minor cosmetic changes.

The default download link will continue to point to the last release until patch 3.3 goes live.

As much as I try to check my work before releasing it, there were a lot of changes this time around which increases the likelihood that I'll have missed something. If you notice anything that seems to be awry, please do let me know.

Known issues:
For fight lengths shorter than the cooldown of Elemental Mastery, the full 15 seconds of the effect will not be accounted for. This is because EM is included as a 15/cooldown fraction of the total fight duration to smooth out its effect. However, the DEP values for potions - if set to be used during EM - do force the full 15 second duration for the effect no matter how long the fight.

The way you have written that leads to a slight concern. My understanding is that the damage multipliers are added at different stages of the calc and therefore it doesn't work out the same as a direct 10% reduction to the muliplier. It might just be the way you have worded it so I will check the spreadsheet when I get a chance to confirm it one way or the other.

My understanding is that the 10% DoT is a straight 10% multiplier on the final damage caused by the spell whereas the 20% damage is additive to concussion and call of flame modifiers making it actually slightly less than 20% overall.


[EDIT] I have checked the spreadsheet and your LvB damage calc looks like this:

=(C132+(E132*sptotal))*(1+concussion+callofflame+t9_4)*damagebuffs

This is correct for the old bonus but with the new T9 bonus I believe it should look like this:

=(C132+(E132*sptotal))*(1+concussion+callofflame)*damagebuffs*(1+t9_4)

Having read the tooltips I agree that the above does sound accurate. However, when we were first discussing the change from upfront damage to a DoT, none of us (that I recall) noticed that it implied a change in the real-terms value of the effect - we just assumed it was moving from DD to DoT and the damage was the same. I don't suppose there's any source of confirmation for what you've surmised other than the tooltips? It should be pretty simple to verify on the PTR.

Unfortunately I don't have any source other than Tufy's explanation (on WoW-Europe) that it is the case which is why I phrased it using 'I believe' and 'my understanding' rather than anything more definate. However, the calculation mechanics should be the same for any 'Ignite' style DoT and Ignite does indeed work that way as does our old T8 bonus. Tufy is very good on the theory crafting front and seems to spend half his life on the PTR and I'm sure that if you PM'd him he could point you in the right direction for the proof you need.

Thinking logically, the second equation with the external 10% multiplier is correct, as it's the last step in the process (ie: 10% of total damage is being turned into a dot).

Sorry for being late, didn't see the PM sooner.

The new T9-4 functions exactly the same as any other Ignite-type dot: it takes the actual damage of the spell, multiplies it with fraction of the ability (in our set bonus' case 10%) and then equally distributes the damage over all the ticks. So, to put it in formula:

Tick damage = ([Actual damage of spell]*0.1)/3

Let's put this to test: here's a quick PTR screenie:



I've taken off most of my gear to get a resist faster, hence the low figures. 4632 damage was caused to the target, this is the figure taken for the calculation (note that the resisted part is ignored). 10% of this is 4632 * 0.1 = 463.2 damage, which is then split in 3 ticks, each dealing 463.2 / 3 = 154.4 damage.

On more info about the mechanic, there was a thread here on EJ in mage forums, but I can't find it right now. In any case, most of that theorycrafting is irrelevant for T9-4, since dot duration is shorter than Lava Burst's cooldown anyway.

I hope that helped.

It's likely that the dot will be getting double benefit from CoE. We cannot be certain this is how it functions but history indicates that this is how it works unless blizzard takes specific measures to prevent it.

I just checked this out on the training dummy with a warlock friend and the DoT doesn't double-dip CoE. It's exactly 10% of the damage the spell does upfront whether CoE's active or not. Ah well.

Hello,

I am really glad you have taken the time to help the community and work on this spreadsheet.

Did you have a chance already to look at some of the trinkets? I know alot of the loot is still unknown but i was curious to how they scale compared to some of the trinkets that are already in your spreadsheet.

[Dislodged Foreign Object]

[Sliver of Pure Ice]

[Muradin's Spyglass]

[Maghia's Misguided Quill]

[Nevermelting Ice Crystal]

Are some of the new ones that i would be interested in seeing some hit value and normal values off.
This is sounding really lazy but i have no clue how to do calculations on this so i'm coming to the ppl that do have the "Know-how".

Actually almost all the loot is available on the armory (except for Arthas, and I wasnt able to filter for trash drops); it's not been reported on mmo-champion et cetera yet for some reason.

The most interesting Trinket might be The World of Warcraft Armory

Phylactery of the Nameless Lich
Heroic
Binds when picked up
Unique
Trinket
Requires Level 80
Item Level 277
Equip: Improves critical strike rating by 172
Equip: Each time one of your spells deals periodic damage, you have a chance to gain 1032 spell power for 20 sec.

Depending on how the proc chance is etc.

Assuming 45s ICD and 10% proc chance, it would proc once every 75 sec with just the flame shock dot, which would result in > 250 average spellpower in addition to the crit rating, so it would be worth roughly 340 spellpower .
If it were to proc about once a minute, it would be worth > 400 spellpower.

As for the other trinkets, the Spyglass from hc is only worth about 270 spellpower.
The Dislodged Object is worth 155 haste + an average 555 spellpower for 20 secs. So assuming one proc per minute, it would be worth about 185 spellpower + 155 haste, so ~340 spellpower.


Using the same estimations, the heroic version (The World of Warcraft Armory) would be worth 170 haste + 220 spellpower => ~ 390 spellpower.

The healing trinkets are quite obviously bad for us, as should be the Ice Crystal considering how poorly we scale with crit.

I plan to have a new release out sometime in the next couple of weeks to include the new Icecrown trinkets as well as to add Fire Nova totem to the sandbox. I also just noticed that intellect buffs aren't being added to the MP5 total from Unrelenting Storm.

New version up:

Changes - Version 1.1.3:

• Reign of the Unliving/Dead trinkets are no longer treated as if they were affected by Elemental Fury.
• The Dislodged Foreign Objects trinkets' procs are now correctly showing their DEP values rather than just their spellpower values.
• Added DEP for Enchant Weapon - Black Magic, as measured against Mighty Spellpower.

Many thanks to Skajin for pointing out the trinket errors and for the evidence for the behaviour of Reign's proc.

Note about the DEP for Black Magic: this was something I did out of curiosity rather than because I wanted to include it in the spreadsheet, so it's displayed a bit awkwardly right now. The DEP value is the effect of swapping Mighty Spellpower to Black Magic - that is, losing 63 spellpower in exchange for the chance to proc 250 haste - so it will nearly always be negative. There is no option to "equip" the Black Magic enchant, nor do I see any reason to implement it.

Yet another new version:

Changes - Version 1.1.4:
- Changed the way Reign of the Unliving/Dead proc frequency is calculated as per this discussion. The trinket should now be ranked slightly lower.

This should put heroic Reign in the same sort of ballpark as heroic Spyglass, and promote Dislodged Foreign Object to second place behind Phylactery (which, of course, is still based upon wild assumptions).

Download

Download - Compatibility

Another update with a fix, a tweak and a new feature.

Changes - Version 1.1.5:

• Trinkets can now be "equipped" using the menu below the Item Sets area: selecting a trinket will add its proc and only its proc to the user's stats/damage. Static bonuses should already be accounted for in the character sheet/armory displays.
• Changed the way Reign of the Unliving/Dead proc frequency is calculated again. Refer to this discussion for the reasoning.
• Swapped the "DEP (hit)" and "DEP (no hit)" labels in the Compare Things Thingy (they were the wrong way around!).

Trinket selection menu
I really hope the trinket selection thing is helpful and not confusing. Let me explain a little more clearly exactly what it does and doesn't do:

• It does not add the trinkets' basic stats: these should already be accounted for by the armory, your character sheet, the Wowhead profiler, etc.
• It does not model the proc any differently: they're just the same values as were already being reported in the trinket list. Averaging out a proc's effect is not the same as applying its full effect to only a part of the fight.
• Equipped damage procs are not accounted for in DEP calculations.

However:

• It does automatically add the proc's averaged effect to your stat buffs and your DPS total.
• It does account for the "trinket proc every ... seconds" setting and all other settings that affect trinkets' procs.
• It can be turned off entirely by selecting "(None)" for both slots.

The number following the selected trinket is just an information display and you don't need to do anything with it.

So that's it - nothing more and nothing less! If it's too confusing or if there are any problems with it, suggestions as to how to improve on it would be welcome.

Download

Download - Compatibility

Looks great however, I see that only Alchemy is represented as far as professions go. While I know that most of the other professions are represented on the paper doll, the Hyperspeed Accelerators from Engineering are not.

I know that I could (and will) just add in the amount of haste given from the glove enchant, but it would be nice to see a little box to check for them to count them in so I won't forget.

For quick reference Hyperspeed Accelerators. For anyone in the same boat, the Accelerators are worth 12x340/60=68 passive haste rating.
Thanks!

[EDIT]: Also, have you any plans to add in the Ashen Verdict Proc? Works out to 10*285/60 (ICD)=~47 spellpower

Should be fairly straightforward to put it in; proc chance appears to be 10% as well, so this would be sp=10*285/(60+averagecasttime/10%). I think it boils down to something around 40 sp assuming some normal (i.e. around 1000) haste numbers

Hello,

First of all you will have to excuse me if my theorycrafting knowledge is somewhat weak and I get something completely wrong. I took a recent interest in your Spreadsheet due to trying to work out some of these things for myself!

I have just a couple of questions.

Does ZAP model the Fire Elemental or Searing in any way ? I could not find any options to enable or disable it.

I am pretty clueless on this so I was wondering what the correct option to choose for partial resists is for a standard raid boss.

At my level of gear I've struggled to find something that models correctly since timing plays such an important part. So far this seems to be one of the more accurate things I have tried to model my damage so thanks a lot.

1. Not yet, but it's one of my main priorities to include in the next release. I'm currently working on rebuilding the entire DPS calculator to more accurately address some of the questions arising from ICC gear levels and the increasing use of fire DPS totems as norm rather than exception.

2. I don't actually know. The formula I used for partial resists by players on bosses is out of date and I haven't managed to find any proper testing on the mechanic since WotLK's release. In addition, the debate over whether bosses had 5 or 8 extra resistance per level never seemed to be satisfactorily resolved. I decided using an old formula which is probably wrong is better than nothing. The 5/level option gives 3.75% mitigation and the 8/level option gives 6% mitigation, which correspond to the conflicting figures discussed in the TTT article comments. If anyone has any better data on this I would, of course, be most interested to see it.

But having said all that, partial resists don't affect the relative weights of stats, glyphs etc. - which is why I've not been too worried about testing it myself.

I'm glad you're finding the spreadsheet helpful!

Well version 1.2.0 is finished, with its inner bits corresponding to what we expect the 3.3.2 changes to be.

Changes - version 1.2.0
Lots and lots. Seriously. Loads. Basically the entire guts of the spreadsheet have been rewritten, lots of new stuff added, some old stuff tweaked, etc.

New stuff

• DPS and DEP calculators entirely redesigned. They now pseudo-simulate a "rotation" (more on this below).
• Chain Lightning can now be used on cooldown if desired. You can still elect not to use it during EM/Heroism. Please note the setting for EM use applies to trolls'/engineers' abilities too.
• Fire totems:
  • Fire Elemental totem is now fully modelled based on data from the EnhSim project (thanks especially to Levva for this). As it gets some benefit from STR/AGI and AP, there are now stat/buff options for this. Please note that shortcomings in the model are my own fault and not EnhSim's.
  • Searing and Magma totem are also fully modelled.
  • "Placement time" advanced option increases the amount of time spent "casting" fire totems, e.g. if you have to run into melee range and back.
  • "Time on target" can be reduced if you think your fire totem(s) will be out of range or not attacking for a portion of the fight.
• You can now select your race and profession bonuses in the buffs menu.
• Hyperspeed Accelerators and Berserking now fully modelled.
  • Uptime for these and Elemental Mastery can now be toggled between "actual" and "smooth". The latter models uptime as duration/cooldown*fightlength and the former determines exactly how many seconds they would be active for if you used them on cooldown every cooldown given the fight duration.
• Limited "Cleave"-style AoE modelling
  • Selecting Chain Lightning, Fire Nova, Magma Totem or Fire Elemental Totem plus more than one target will generate additional damage based on number of targets.
  • The model is rudimentary at present and does not properly affect Clearcasting or spell priorities.
  • Debuffs selected will be applied to all targets.
• You can now see a "snapshot" of the "rotations" the spreadsheet is using to work out your DPS. I figured this may be helpful for new players, as one of the most helpful things to me early in WotLK was going under the hood of SEIC and seeing how it constructed its "rotation".

Changed stuff

• Enhanced movement modelling. More on this below, but here are the main points:
  • "Movement per LvB" advanced setting works similar to the old "Rotation delay", only now the calculator can cast spells (frost shock, fire nova or both) during "movement" if allowed by the options.
  • Spell damage can now be universally modified by the "Deflate spells" advanced option. You can use this to reduce the damage done by all spells of a certain type.
    • Reducing Lightning spells affects both LB and CL and will also reduce the number of Lightning spells factored in to 2t10 EM cooldown calculations.
    • Reducing "AoE" spells affects the primary hit of CL as well as subsequent ones. Sorry.
• Some default options have changed:
  • Miss chance and partial resists - now turned ON by default. Now also apply to DEP calculations.
  • "Cast delay" is now an Advanced Option. Still set to 0 by default. I hope to model latency and reaction time better in a future build.
  • Trinket proc setting is now at 55 seconds instead of 60. I may do a proper uptime model for every trinket later.
  • "Rotation delay" is redesigned into the new movement stuff.
• Trinket, racial, profession and potion DEPs are now on a separate tab.
• Reign of the Dead/Unliving trinkets are now included in DEP calculations, if selected.
• Haste DEP no longer has freaky dips and peaks at certain values. Instead the spreadsheet performs two separate calculations for haste DEP and selects the most reasonable one to use. This should eliminate most "haste anomalies" from DEP, though not from DPS.

Stuff that's gone which might come back later

• Black Magic DEP
• Encounter statistics
• eDPS and DPCT

The encounter stats were really useful to compare against WoL reports when I was writing the first version of the spreadsheet, but now I've got other ways of verifying the formulator and the stats are a lot of work to write. They may make a reappearance later. DPCT I just forgot and can't be bothered adding right now.


Known issues (will be fixed before 3.3.2 is released)

• Thunderstorm's MP5 and -DPS are calculated but not integrated.

The new "movement" stuff
Obviously a big limitation of theorycraft is calculating the incalculable - factors which can seriously affect our DPS and spell use etc, but which vary between encounters, players, and so on. We often think "ah, but movement" whenever we TC something and have to "guesstimate" how it'll affect what we're thinking about.

I've thought long and hard, and had some detailed discussions with certain folk, about how to model movement better in formulation.

The bad news: I'm pretty sure it's impossible to fully model the effects of movement in a formulation without entirely defeating the point of formulation itself.

The good news: I've dramatically improved the abstract, formulaic methods which the spreadsheet provides to "mimic" the effects of movement if not movement itself.

Movement per LvB: This inserts a delay in seconds after every LvB cast the calculator performs. These can be filled with shocks/Fire Novas (cooldowns and settings permitted) or with blank, no-damage GCDs. This can be used to represent very high proportions of movement which change spell use significantly.

Deflation menu: This is a bit more abstract. Basically if you think that you'll be casting, say, 30% fewer Lightning Bolts because of movement, just deflate LB by 30%. This also reduces the number of casts counted for EM's cooldown with 2t10, and it fully affects DEP, so it should really provide a good model of what happens with reduced LB usage. The same options are available for LvBs, shocks and AoE spells if you can think of any reason why you'd want to deflate their damage.

These settings are in the Advanced Options to the right of the buffs/debuffs menu. "Advanced options" aren't meant to be entirely unused, but at the same time it's important folk know what they're doing and why before they go fiddling.


New tabs
I know that one of the best things about SEIC's interface was that it was all on one tab, but I couldn't make everything fit and seem usable/friendly. I've moved the DEP displays for trinkets, potions and racials to the new Trinkets tab (along with new DEP for professions, though those are presently based on averaged stats rather than the new berserk/hyperspeed models). In future I'd like to dramatically expand the trinket information and add some new calcs.

The new Calculator tab has the same calculators for working out ratings etc that used to be on the Sandbox tab, as well as the Compare Things Thingy. Normalised EP is also relocated here. The haste calculator that nobody seemed to know about is also here now - it works out haste rating needed to achieve a specific haste % given different combinations of buffs.

The Snapshot tab is just where the user can view their "rotation" as decided by the spreadsheet. The spreadsheet "simulates" a maximum of 60 casts. After the first 40 casts, if the rotation is not naturally "repeatable" (accurate formulation depends on repeatable rotations), it starts to force a repeat by adjusting some stuff to get rid of (most) DPS anomalies. This should be, in general, a lot more accurate than the old formulator which had (for example) an artificially low Flame Shock DoT uptime.



New haste effects
Yes, Berserking and Hyperspeed Accelerators are fully modelled in their own little pseudo-sim with uptime models and everything.

These effects and EM also now model a complete "rotation" of at least the effect's duration, but if possible they model more to make sure the effect on the player's subsequent spellcasting sequence is taken into consideration. This means that they should be a lot more accurate than if they just modelled a whole rotation as if the buff were present the whole time, especially in the case of EM's instant-cast spell.


Downloads
Please note these are for patch 3.3.2 and the user guide is NOT yet updated. Hopefully this will be done to coincide with patch 3.3.2's release, at which time the changes will be finalised and the main links updated to the new version + any changes/fixes that may be necessary.

ZAP! 1.2.0
ZAP! 1.2.0 Compatibility version

I've been working on this nearly non-stop for two weeks and am just about on the verge of an existential crisis (cognito ergo sum means "I think therefore I am a sum", right?). It's more than likely there are things I've missed or that certain combinations will produce odd results. So please do report them if you find them.

However, the whole point of an exercise like this is to determine in abstract what we want to know in concrete. The answers may not always be what we expect. This does not mean they are "wrong". Unexpected results may or may not indicate bugs. There are a bunch of reasons why this may give different results to, say, SimulationCraft, which I will gladly talk about further if anyone does find the results to be a little different from what were anticipated.

I've tried very hard to make sure this is as accurate and powerful as possible, especially with regard to key questions of haste scaling and stuff, so bear that in mind. Thankee.

It's interesting to see directly how fight length impacts the DPS values of glyphs like EM and Fire Elemental. I may need to start carting around some extra glyphs for longer fights.

Thank you once again for all of your hard work on this project.

How close should i be able to get to ZAP's theorterical dps?

I used SEIC in the past and later i used Rawr because of the great user interface. But as Rawr isnt very good at trinkets i decided to check ZAP!.

It was a breeze to fill in the numbers and i find the comments on most fields very useful. But i do get a weird result. It says my max dps should be 11k. I probably dont ever have _all_ the buffs i said "Yes" to in this sheet, but 11k is nowhere near what i ever got in game. On straight forward fights like Patchwork (weekly) i am lucky to reach 9k, but most of the time i have to settle for 8.6k or less.

As Rawr tells me my max is 8.8k it feels that that number is closer to the 'truth'.

I cannot persuade Rawr to wear ilvl 264 T10 chest + helm as it insists it is a downgrade. I am trying to find proof with ZAP! that 2 x 264 T10 (helm + chest) is better than 3 x 245 T9 (chest, pants, shoulders) + 1 x 258 (helm). I have a feeling it will be a close call.

The reason is, that Rawr doesn't model the t10 set bonuses, yet.
I just don't have the time to do ist right now.
Experience shows, that 2t10 251 is already an upgrade from t9 245.