Efejel

Could I see the math behind this statement?

"Electronic communities build nothing. You wind up with nothing. We are dancing animals. How beautiful it is to get up and go out and do something." - Kurt Vonnegut

 

Lambach

You want to see math behind why increased dmg and reduction of partial resists increases our dps?

 

Maax

Obviously it increases the DPS of Starfire more than it does the DPS of Wrath. But if Wrath is higher DPS to begin with (and it certainly would be higher DPS if NG reduced the global cooldown), then Wrath could still win.

Devs: Our nerfs will block out the sun!
Druids: Then we will tank in the shade.

 

Nerfy

It increases starfire damage because starfire is arcane damage. Wrath is nature damage and isn't affected by it.

 

Maax

Yes I know. My point is that if Wrath DPS is higher than Starfire DPS (without CoS) by enough then CoS isn't the sole determinant of which is more DPS.

Devs: Our nerfs will block out the sun!
Druids: Then we will tank in the shade.

 

Ten

Okay, I've read through the whole thread, so I'm not going to ask anything that has been asked and answered previously.

My first question is regarding the [Ashtongue Talisman of Equilibrium]. Throughout this thread, it seems like it's considered useless based on the cast time of starfire and the fact that you'd only get two casts off before the buff disappeared. What if, however, you could stack enough haste to get 4x starfires in 8 seconds reliably? This would, in theory, keep the buff up nearly 100% of the time. Would it still be worthless then?
Based on stuff I've read here, NG is applied before spell haste, correct? If so, you'd need 392.5 haste rating in order to reduce your post-NG cast time from 2.5s down to 2.0s. This amount of haste is currently unattainable (assuming you're still keeping 4piece t5/t6).
If not (or if they change it to work this way), you'd need 314 haste rating to get starfire down to 2.0 seconds. This is surprisingly close to an attainable value. While wearing 4piece (all but gloves) t5/t6, you can get up to a maximum of 308 haste rating ([Bracers of Nimble Thought], [Brooch of Nature's Mercy], [Botanist's Gloves of Growth], [Belt of the Crescent Moon], [Footpads of Madness], [Shroud of the Highborne], 2x[Ring of Ancient Knowledge], [Zhar'doom, Greatstaff of the Devourer]), which is 308 haste rating (19.6% haste).
I suppose my question here is just "Is that trinket worthwhile if you can keep it up >75% of the time?" The +dmg seems pretty hard to beat.


Next up, to gear for all haste or not to? Using this spreadsheet, the highest possible dps I've been able to obtain was with 4t6, the abovementioned haste gear, Illidan trinket, and AToE (estimated with 75% uptime for 112.5 +dmg). Since I'm predicting this for my future raiding stats, I've included other stats that I anticipate having (~230 mp5 from a shadow priest, elemental shaman buffs, 100% JoW uptime from the ret paladin, 13% CoS, 50ms latency, treants doing 6k dmg per cast, 15 minute fight, 3 innervates, no potions). This gave me an average dps of ~1850 over the 15 minute fight without going oom doing the MFx2,SFx11 rotation. Based on the fact that I've tried to get higher dps out of the spreadsheet (which is awesome work btw), is it worth just going all out for haste and ignoring crit?
Relevant projected stats with that gear (buffed):
619 int
405 spirit
25.43% crit (not counting starfire talent or set bonus)
16.36% hit
1623 arcane damage
243 mp5
My guess is there's a series of haste thresholds that are optimal to reach and anything between them is useless, but I'm not too familiar with this since I only started looking at it a few days ago.


Is [Quagmirran's Eye] good enough to keep all through end game content? It seems like the 45s internal cooldown would make it fairly less desirable than other trinkets.

Essentially, I'm in a position where I can get whatever I want for gear, but I'm not sure what the best possible gear is.

 

Naboo9

I got my T5 helm last night and am seriously considering socketing it with the chaotic skyfire diamond as I already meet the socket requirements. I was looking around various parts of the spreadsheet trying to figure out how to change the parameters for this diamond from exactly 2 blue gems to at least 2 blue gems, and can't figure out how.

Efejel, could you please go over this when you have a chance?

 

Mitakulix

Nice spreadsheet.


Can you please put in "Skullfish soup" into buff food selection? (+20 crit)

 

Allanon

Yeah i'm in this position too (anyone farming content is) and there seems to be no definitive answer. Everyone says t5 > all, useless for me because I cant get, so shooting for 4xt6. But trinkets dont work in the spreadsheet and the meaning of haste is a really good question. One problem with the haste data you put is that without a spriest i think you will go OoM really fast, especially @2sec starfire. Dunno about you but i dont get spriest 100% of the time. There needs to be some kind of definitive gear guide for endgame moonkins and there simply isnt.

Dunno if this has been posted before, or if it's accurate, but here is something of a trinket guide:
SimulationCraft/Trinkets/Druid - Shadowpriest.com Wiki

 

Ten

In general, I can expect to always have a shadow priest and an elemental shaman in my group.

I made a separate list for maxing crit gear (36.5% crit, 1636 +dmg buffed). The projected max dps (MF/SF rotation) for the same 15 minute fight was 1713. Examining the mana consumption, there really doesn't appear to be much difference. Using the same 15 minute fight but subtracting all mana regen except , maximizing crit will make you go oom after about 7 minutes, and maximizing haste will make you go oom after about 6. This means you're running out of mana ~14% faster under those conditions. Maximizing crit only gets you an additional 11% crit at the cost of 60 int and some other negligible stats, meaning you start out with ~1000 more mana in haste gear.

What's fairly interesting about these two sets of gear is that as the fight time decreases, the dps for the two sets evens out. On a 3 minute fight (Gorefiend, Akama), the dps numbers are identical with optimal rotations, and the time till oom (no consumables or regen of any kind) is within 10 seconds. Essentially, testing these without assuming maximum mana regen is useless, because that favors gearing for crit far too much.

Going back for maximum mana regen (JoW, 230mp5 spriest, innervating), the 3 minute fight is vastly different. Neither selection runs oom, and the haste set maintains 1863 dps while the crit set only manages 1698. This is a 9.7% dps difference. An 8 minute fight shows the same results: haste gear outputs 1875dps, crit does 1710. A 9.65% difference.

The results are approximately the same for fights of all lengths. Gearing for maximum haste (308 rating) gives you a ~10% dps increase over gearing for crit. An elemental shaman provides the spell hit necessary to cap the haste geared moonkin, meaning the only difference is crit vs haste since +dmg is approximately the same.

Thus, it SEEMS like haste is always better than crit according to the spreadsheet.



EDIT: I also checked out t5 vs t6 on the spreadsheet. With the crit set, substituting t5 for all the t6 gear gives the same dps for an 8 minute fight. For the haste set, however, it yields 1887 dps, a 12 dps increase. The crit gear's dps with t5 remains about the same for all fight durations, but the haste gear's dps with t5 does increase a little as fights get longer.

 

Laqwanda

I'd like to hear the viablity of wrath spam with haste gear considering the new changes to GCD with spell haste in 2.4?

 

Efejel

A picture is worth a thousand words:

"Electronic communities build nothing. You wind up with nothing. We are dancing animals. How beautiful it is to get up and go out and do something." - Kurt Vonnegut

 

Nysus

I'm currently in the process of gearing up for a Balance respec and have a question about the spreadsheet.

How does the DPS portrayed in the spreadsheet match up to a real time game environment? Obviously this will differ from person to person and fight to fight, but generally how much less dps does everyone get compared to what the spreadsheet says?

 

Rosoo

A page or two back you can see a few people throwing up their WWS parses. For mine, I used a MF SFx4 rotation as well as keeping up IFF. The spreadsheet is not an incredibly accurate portrayal of exactly how much you'll be able to do, as every fight has something about it that forces you to not cast for a moment. I found it to be very accurate though when it comes to casting rotations and which will work best depending on how much you get pushed back, how much you have to move etc.

 

Ambiorix

It seems like your uptime estimates are off. To simplify, let's just assume you only cast starfires. Then the "uptime" of the trinket you will be interested in is the percent of your starfires that are affected by the trinket buff. With the (extremely unrealistic) 4x starfires within every 8 second period, whenever you cast a starfire, the buff is up as long as at one of the previous 4 starfires proced, which has probability only 1-(1-.25)^4 ~= 68%, giving about 102 spell damage on average.

With a more realistic 3x starfires in 8 seconds, it's only 1-(1-.25)^3 ~= 57%, giving about 86 spell damage on average.

Even that amount of haste seems a little high for most people though (I won't claim to have tried it out in the spreadsheet, but I'd be surprised if stacking haste/healing items is ever good for dps...). It seems like it shouldn't be too hard though to stack enough haste that you only need to crit with one starfire to be able to fit three in 8 seconds, in which case assuming a 30% crit rate, the chance the buff is up will be 1-(1-.25)^2+.25*((1-.25)^2)*(1-(1-.3)^3)) ~= 53%, giving about 79 damage on average.

Similarly with two crits required to fit three starfires in, it's 47%, or about 70 damage, which seems pretty weak for a BT trinket. As you suggest, there are a bunch of different haste cutoffs where the value of this trinket jumps based on how many starfires you can fit in with various numbers of crits, but it seems like at any reasonable haste level this trinket is far from amazing (though once again, I've oversimplified by considering only starfire, not counting heroism, etc.).

 

Carisana

Next week, you guys should see your first 2k dps Boomkin. Madrox finally got his Zhardoom and had a nice Hyjal. He would have broken 2k on Rage last night but the mt died (don't ask not really sure lol). He still needs a Skull of Guldan so there is still even room for improvement over that.

But did pull off 1,942 on Anetheron
WWS Loading...

He still did mid 1800's on Rage even with the debacle and 1,830 on Kaz.

 

Moginheden

2.4 has 2 notes I noticed:
# Idol of the Unseen Moon: The cooldown added in a previous patch has been removed.
# Spell Haste: Spell haste now reduces the global cooldown on spells, down to a minimum of 1 second. This change does not apply to melee and ranged abilities.

Since haste will not change the GCD from idol swaping from what I can see. Is the unseen moon idol worth using again? Will the extra GCD time from swaping in this idol still kill it? Will wrath no longer be worth swapping idols for when using both wrath and starfire?

I know it was originaly suggested to have 2 starfire macros, one for unseen moon one for ivory moongoddess. Pesonally I find in real battles I don't know at the time I cast starfire if a moonfire, another starfire, or some other spell will be next becuase I'm not just sitting there concentrating on my spell rotation, other things are happening in the fight. Do others find the same or doI just need to concentrate more?

 

Celdhyrean

They are also upping several other idols, amongst which :
* Idol of the Raven Goddess: The bonus healing, critical strike rating, and spell critical strike rating from this item have been increased. In addition, it no longer sometimes fails to apply the healing bonus.

Was somebody already able to get info on this ?

 

◊ Lorewanderer

44 heal (up from 20) , 20 crit rating (up from 9)

 

Margot

I'm sure the new caster Meta will be lower dps than Chaotic, but it would be nice to have it modelled anyways:

http://static.mmo-champion.com/mmoc/...ry/newmeta.jpg

 

Benita

@Ambiorix

The biggest disadvantage this trinket had was that it did not affect the SF it procs of, if it wore off before. Do you actually calculate this part in or did that change in some patch since 2.1? I never bothered to test it after that.
Basically you can get 2 or 3 SF on a double crit in a fresh procs duration, which meant without massive use of crit or haste rating its effect was worse than quags eye even on chain casting SF.

 

◊ Lorewanderer

I've got to say I'm excited about the new loot from Sunwell. 8 pieces of t6 means we either get a really amazing (>4) piece bonus, (which honestly I doubt) or can get the 4pc on both t5 and t6.

I'll be messing around with the spreadsheet to see how the new haste gems stack up, but I've got to say it looks promising.

 

Efejel

Version 0.80c released! >>>Link to FileFront.com<<<

v0.80a changes
-DPS Comparison sheet added
-Healing items (epics of iLvl 100 or greater, leather only for 8 main slots) added
--Please make me aware of any cloth items that must be added b/c of w/e reason
-Currently expected 2.4 items added
--Except healing items, sorry. But, hey, it's not like you can wear it today anyways!
-NO 2.4 MECHANICS CHANGES
--Will need better idea of the spirit change & the effect of NG on GCD before delving into this
-Weapons now support gems
-Possibly other changes--have been working on this/delaying so long I forget
v0.80b
-Fixed belt sockets, weapon sockets/gems
-Added auto-detect for Windhawk 3-pc. bonus, added detection for new T6 pieces to the auto-detect on 2- & 4-pc. bonuses
-Updated Raven idol to 20 crit rating (this means it's wrong for current in-game calcs, add a -11 crit to the custom to correct)
-Added the ability to see the total change to in-combat mp5 as a result of the 2.4 spirit changes
-Added Royal Cloak of the Sunstriders back (whoops on losing it between versions)
-Added more 2.4 cloth gear
--Set fight to demonstrate haste-stacking DPS
-Added [Skullfish Soup] to food buffs
-Added mana potion option for PvP/Major/Unstable mana pots
-Added DPS penalty for Major Dreamless Sleep potions
v0.80c
-Added addt'l 2.4 items, especially badge & Sunwell loots, and epic haste gems
-Updated the base in-combat mp5 calcs to account for the 2.4 spirit changes
INNERVATE CALCULATION NOT YET UPDATED
No other substantial changes, though hopefully some minor bugs I'm forgetting the specifics of have been fixed.

TO DO:
Implement 2.4 innervate calcualations.
Update GCD for 2.4 haste changes & display (somewhere) haste-modified "base" casting time of spells.
Add [Sorcerer's Alchemist Stone], [Grand Magister's Staff of Torrents].
Fix name of [Fused Nethergon Band]

"Electronic communities build nothing. You wind up with nothing. We are dancing animals. How beautiful it is to get up and go out and do something." - Kurt Vonnegut

 

Efejel

For my own notes:
Sheet uses one-roll, two-roll may be more accurate for spells

To-do list, indeed.

"Electronic communities build nothing. You wind up with nothing. We are dancing animals. How beautiful it is to get up and go out and do something." - Kurt Vonnegut

 

Murph

Hi Efejel,

Firstly, many thanks for all your effort in producing and maintaining this spreadsheet - I have been finding it extremely useful in tuning my equipment and techniques to squeeze just that little bit extra damage out during long boss fights. For reference, I'm 56/0/5 spec, currently working on finishing Karazhan and starting Zul Aman.

One small request is for you to add the Windhawk set bonus automatically. I've patched this in myself by adding a new row to the Set Bonus Calcs sheet, and here's the formula for column D for your convenience:

I'm then just adding it to the mp5 from gear by changing L7 on Gear Select to "=SUM(L8:L94)+IF(Wh3;8;0)", obviously naming the relevant cell in column C of the Set Bonus Calcs sheet as Wh3. I don't think it needs the ability to be manually selected, but I'm not entirely sure what your motivation for having that functionality is.

I think I have also found some bugs in the new rows of the Gear Select sheet added for weapon gems in v0.80a. Cells N84:N86 have ";12)" at the end of the forumla, which seems to be erroneously adding 36 nature damage to my gear. Changing to ";0)" produces the result I expect. Less importantly (for me), cells U84:W86 seem to erroneously refer to the chest item, not the weapon. Other than that, it appears to be produce the same results for me as v0.72c or v0.80beta.

Thanks again, and keep up the good work! :-)