If I'm interpreting Drysc's post correctly (which I hope I am) this means that haste will allow us to stomp totems faster
I am pretty sure you are right, since Totems are a spell. I am certain Life Tap will be able to be hasted.
Millions of words are written annually purporting to tell how to beat the races, whereas the best possible advice on the subject is found in the three monosyllables: 'Do not try.'
As far as I can tell, the only way for the GCD to be affected is all or nothing. I can't imagine them making spell haste affect the GCD only on spells that have a cast time, it would wreak havoc on cast rotations completely unnecessarily, and I see this change as motivated to bring these DoT classes back into competitiveness with the blast classes, not as a boost to us (despite if it helps us.) Affliction locks and Shadowpriests have fallen behind in damage as gear is stacking more and more "secondary" stats rather than large increases of spell damage, by far their best scaling stat.
I'm loving the idea of totem stomping going faster, even if its something we won't notice. We spend about 5% of a fight with no movement dropping totems, which puts a damper on our total DPS time. I didn't factor this into my inferences on the effects of this change, but it looks like haste will be far and away our best scaling stat 1:1 vs. crit or spell damage.
Basically I took our average damage in a 3:1 LB/CL cycle, with 1000 SD, 40% crit and 0 haste, added in a 20% chance to add half damage per cast (LO) divided it by the cast time of the cycle, and then compared the effects of increasing haste/SD/crit without any consideration for the GCD (or mana regen/consumption, which is why I didn't use Bink's sheet for this) for a pure damage comparison, and got
These numbers vary a bit as each stat is altered, but I found no conceivable numbers that could be obtained without intentionally messing up your gear or spec that would leave haste anything but number 1. I need to try adding in some mana component, as this is the inevitable consequence of DPS increases that do not decrease DPM, but I have a hard time believing that with a shadowpriest this would be a significant problem, we might have to use more Super Mana Pots, but who knows.
This completely changes spell rotations with CL when under a large amount of spell haste from items/trinktes/drums/Bloodlust. I am going to get the cloth haste bracers crafted asap as well as try to grab the next Zhar'doom. Wish I had picked one up earlier, now all those mages and warlocs wil want their grubby paws on one. Might have to go farm for another Quags eye (since I sold my old one) and try swapping that in over Hex head. This really does change my entire attitude towards large spell haste procs as a profitable DPS increase.
The Zhar'doom and crafted bracers might be good buys, assuming that these changes go live as is, but I wouldn't break you bank on it atm. As to the Quags eye, the numbers on it don't work out, it was still way back on the list of top trinkets, but the Skull of Guldan moved into an easy number one (again, assuming the changes go live.)
One of the weird kinks in this that give me cause for concern is that when comparing the DPS increases of the stats post-patch, the HEALING neck from the Eagle in ZA [Brooch of Nature's Mercy] moves into the number DPS 1 neck in the game for us, in pure terms of damage output, which makes me thing the change will be altered a tad, Blizzard doesn't want to skew the definition of healing/damage gear.
I'd like some assistance in remodeling TLC with consideration for the no GCD spell haste change, and see how it is affected by it. One thing to bear in mind with these changes, is that we are usually trading spell haste for crit, which has an affect on our longevity as well, and also a very direct effect on the use of TLC.
A friend of mine has helped me work out the formula for the Sextant in accordance with your current cast time and crit rating. It goes as follows:
(15/(cast time in sec/(crit %*0.2)+45))*190
credit goes to Tisiphone of my guild
Using that formula I also came up with the formula for the skycall totem:
(10/(2/(1+(haste rating/1577))/0.15))*100
Please check if that's accurate (the results seem good) and I suggest that it could be implemented into the Shamstats spreadsheet.
Last edited by Gaguusi : 02/02/08 at 7:24 PM.
Reason: orthography
Am I the only that feels the 2 piece bonus from T4 is under-valued by a lot of ele shaman? It would seem that 100 more spell damage for your group would make 2 parts of cyclone impossible to replace until you can hit the T6 4 piece bonus.
That being said i've seen very few shamans who actually do stick with 2 pieces of T4 up until they get the T6 bonus so is there something I'm missing here?
Sorry if this has come up before but a search for tier/bonus didn't return anything about T4 and maybe I'm just being dense but it doesn't seem the search function will accept a search for T4 or 2 piece.
Well the T5 bonuses are crap so yeah if you can get the best pieces of T4 then yeah it is worth using. The best pieces of T4 are chest, head and shoulders. So given at that gear level you have the LW set, then go head/shoulders. Howver if you pvp, the s2 and s3 stuff are pretty much superior now, especially s3. I only ever got T4 gloves so I never got to work it out.
But given you have T4 head/shoulders, going from T4 to T5 nets you 20 spell damage and 13 spell crit from the pieces alone. Obviously the T4 bonus counteracts the 20 spell damage. Thus it is 13 spell crit. However T4 socket colors and socket bonuses are much better too. 5/4 Spell damage with all gems being yellow as compared to T5 which is 4 hit/3 crit with 2 yellows and a blue. Given you socket with 5/4 gems and a 6/5 sta/dmg (and a 9 dmg gem on the t5 hit, worthless spell hit bonus) gem the difference is 24 damage and 12 crit for T4 compared to 19 dmg and 7 spell crit. Thus yes if you have 2 piece T4 where the pieces are shoulders/head, don't break them up til you get one of those pieces as a T6 piece.
At that place I'd likely do it like this. If you can get T6 2 piece or T4 2 piece get 2 piece T6. If you can get both thats great. If you can get 4p T6 forget about T4 altogether.
That is exactly the reason I'm still wearing two pieces, until I can contribute the damage in my own output, I can't justify removing the 100 spell damage from the raid.
This is further complicated when you look at what 1 point of spell damage does for other classes compared to us, we scale rather poorly per point due to lack of empowerment talents, so the 20 damage to the 2-3 mages in your group, and the 20 for the cyclic nature of damage to mana to the shadowpriest makes it quite hard to remove, and all but impossible to model. I chose to stick with the head and shoulders myself, as they seemed to be the smallest "downgrades," but I'm looking forward to breaking it the second I get the head off Archimonde, which I hope to be in the next couple weeks.
A friend of mine has helped me work out the formula for the Sextant in accordance with your current cast time and crit rating. It goes as follows:
(15/(cast time in sec/(crit %*0.2)+45))*190
credit goes to Tisiphone of my guild
Using that formula I also came up with the formula for the skycall totem:
(10/(2/(1+(haste rating/1577))/0.15))*100
Please check if that's accurate (the results seem good) and I suggest that it could be implemented into the Shamstats spreadsheet.
Ok, I can understand how the Sextant calc works, but I can't make sense of the one for Skycall...
Well, I was actually thinking that an explanation could be useful :P
It basically works the same as the Sextant calc, but the result is haste rather than spelldmg obviously.
(10/(2/(1+(haste rating/1577))/0.15))*100
10 seconds is the uptime of the buff, the bold part after that is simply the haste formula I got from Wowwiki, the 0.15 is the proc chance of the totem and the 100 in the end is the bonus it provides.
I put together some sets using Bink's spreadsheet, and did some math. The first set was basically full skyshatter + MH/BT epics, second set was a full haste set, 2xBand of Anchient Knowledge, Quag's eye, Skull of guldan, no Zhar'doom though.
Well, I was actually thinking that an explanation could be useful :P
It basically works the same as the Sextant calc, but the result is haste rather than spelldmg obviously.
(10/(2/(1+(haste rating/1577))/0.15))*100
10 seconds is the uptime of the buff, the bold part after that is simply the haste formula I got from Wowwiki, the 0.15 is the proc chance of the totem and the 100 in the end is the bonus it provides.
Hope this helps.
Ah, so the haste rating is referring to the currently equipped haste rating, so I could break that back to a simple current haste percentage.
10/2/.15 = 33%
From the sims run previously, we're looking at a 66% uptime without taking LO into account (I've been multiplying 0.66 by 1.2 to take into account the chance that LO will proc it), but your math suggests a lower uptime than the one I currently use, although it does take into account faster cast times. The only other problem I can see from that is it will generate a circular reference if I use it for the "average haste" value that I currently do.
When I've been calculating uptimes, I take the duration, multiply by the effect, and divide by the cooldown.
If you want to play around with other stuff, try working out a good method for calculating mp5 from the Ashtounge exalted trinket.
I've added that Sextant (and thus Nexus-Horn as well) formula in, and it seems to reduce the +dmg value slightly, which seems fairly good.
I'm also toying with re-doing mp5 values (and thus haste) by looking at fight durations, mana pool, and mana use (yes, I know, this is the third time, but if I'm not happy with something I'll keep tinkering with it until it works, damnit). I'm still keen on the idea of accounting for mana use, but it's a matter of making sure it's handled correctly.
Ah, so the haste rating is referring to the currently equipped haste rating, so I could break that back to a simple current haste percentage.
10/2/.15 = 33%
From the sims run previously, we're looking at a 66% uptime without taking LO into account (I've been multiplying 0.66 by 1.2 to take into account the chance that LO will proc it), but your math suggests a lower uptime than the one I currently use, although it does take into account faster cast times. The only other problem I can see from that is it will generate a circular reference if I use it for the "average haste" value that I currently do.
When I've been calculating uptimes, I take the duration, multiply by the effect, and divide by the cooldown.
If you want to play around with other stuff, try working out a good method for calculating mp5 from the Ashtounge exalted trinket.
I noted this when I posted, but in the sims I ran with 100+ hours the up time for skycall was about 62%. I think the numbers with 66% were shorter runs and more subject to margin error.
Ah, so the haste rating is referring to the currently equipped haste rating, so I could break that back to a simple current haste percentage.
10/2/.15 = 33%
From the sims run previously, we're looking at a 66% uptime without taking LO into account (I've been multiplying 0.66 by 1.2 to take into account the chance that LO will proc it), but your math suggests a lower uptime than the one I currently use, although it does take into account faster cast times. The only other problem I can see from that is it will generate a circular reference if I use it for the "average haste" value that I currently do.
When I've been calculating uptimes, I take the duration, multiply by the effect, and divide by the cooldown.
If you want to play around with other stuff, try working out a good method for calculating mp5 from the Ashtounge exalted trinket.
I guess multiplying the result by 1.2 is a sound way of incorporating LO into the calculation, although I'm somewhat surprised by the result I get with my current haste (164 permanent haste), which yields an average of 99 haste on the totem.
I'm currently working on the formula for the Ashtongue trinket as per your request. Really glad I get the opportunity to help you with your project
I use a zero haste model, because that's where I'm sitting atm, and there should be a baseline like that, plus I would expect to see a gradual uptime increase, not an exponential one.
5 is the conversion to MP5, the 25.5 is simply the amount of mana multiplied by the proc chance and the 1.2 in the end is LO.
edit: about your comment on the average uptime of the Skycall totem: I get 75 haste without any current haste or LO being used, so I don't understand why you said that my uptime ends up being slightly lower...
I just noticed that my Skycall totem calculation is basically useless, as the proc uptime itself and the effect the haste gain has on the proc isn't part of my formula. I'll try to work it out correctly, but it seems pretty hard to do an accurate calculation.
5 is the conversion to MP5, the 25.5 is simply the amount of mana multiplied by the proc chance and the 1.2 in the end is LO.
edit: about your comment on the average uptime of the Skycall totem: I get 75 haste without any current haste or LO being used, so I don't understand why you said that my uptime ends up being slightly lower...
I'll go over it when I get to work, but it looks fairly good there. I'll most likely substitute it to be:
5*((Mana*Proc)/cast speed)*LO as I've already got the cast times calculated elsewhere.
One thing, can you send me a quick PM detailing how you came up with these formulae? It seems a fairly decent way of calculating returns/uptimes and I want to include it in a wiki entry discussing proc theorycraft.
Originally Posted by Gaguusi
I just noticed that my Skycall totem calculation is basically useless, as the proc uptime itself and the effect the haste gain has on the proc isn't part of my formula. I'll try to work it out correctly, but it seems pretty hard to do an accurate calculation.
Yeah, that's why I was relying on the sims to give us a better understanding of what the uptime is.
Ah, must have missed that bit.
It wasn't including the possibility of LO causing skycall procs was it?
Correct LO did not proc skycall in my sim.
Originally Posted by Gaguusi
I just noticed that my Skycall totem calculation is basically useless, as the proc uptime itself and the effect the haste gain has on the proc isn't part of my formula. I'll try to work it out correctly, but it seems pretty hard to do an accurate calculation.
Yeah I tried to work out some sort of recusrive math solution got a headache , gave up, and wrote a sim. (I'm fairly sure if you used some calc you can do it I just been out of school too long and I forget how)
"Patch 2.3
With patch 2.3 both LB and CL are being changed. The base cast time is being reduced by 0.5 seconds, along with a reduction in mana cost. This has the adverse effect of reducing the spell damage co-efficient from 0.857 to 0.714 for LB, and from 0.714 to 0.571 for CL."
This is still in the first post, while I believe (and others too) is wrong. Just a fex posts above someone does his math with a 0,794 coefficient.
Please correct it, it missleaded me at first too.
Where L = Lightning bolt damage, C = Chain Lightning damage, and T is the "break even" point
Do you think this works for figuring out whether to use a 3/1 rotation and wait for CL cooldowns, or a 4/1 rotation which uses an extra LB rather than waiting
Last edited by Binkenstein : 02/04/08 at 10:53 PM.
02/02/08, 12:34 PM
Similar Threads
