Allright if we can now get back to the topic of Affliction theorycrafting it will mean I don't have to start infracting people for being off topic and ignoring a mod warning. This was a good and useful thread, let's get back to that discussion.
Someone just watched the new Clint Eastwood movie. Or saw the preview on demand. The massive difference between the two is the game of WoW has solid numbers for things. A gem is going to give you a definite number of certain stats. Same with an enchant or a piece of gear. The thing they're doing is calculating how much those stats can give you in a dps value. Nothing a computer with baseball can do. Two totally different worlds.
regardng haste plateus, despite people using the phrase extra dot tick that isnt what they actually mean (or if it is, they are wrong ). its a shorthand way of saying 'extra Dot tick every 18 seconds (for agony as an example). just wanted to say that cos i feel that a lot of people dont get that....
more on topic.... i for one am going to continue gemming pure INT in red, and then INT hybrids (as opposed to mastery hybrids in all slots) purely because while in one situation haste > mastery, and in another situation mastery>haste INT will ALWAYS be very strong. ill admit i may just be stuck in the old mindset of MORE PRIMARY STATS!, but hey...
and also it lets me keep playing demo every now and then.
The problem is this is no longer accurate for single target. When you go from 6600 to 6640 haste, your dot tick timer moves what, .01 seconds? The benefit is that 1 more tick per cast gets added on (when you won't be renewing it) and - the only part that really matters for single target - you reach the maximum duration per cast, up from the minimum. As you gain haste, your dot lasts for X seconds. X becomes lower and lower until you trigger a new tick, where X then becomes the highest number it can. Eeeventually, this leads to times you don't need to refresh your dots! An extra GCD in a fight, while minor, is an extra GCD, and that's DPS spent doing MG.
Hastes' OTHER benefits make that plateau relatively unimportant. You still gain time on haunt casts, MG ticks, DS ticks, corruption tick #s - for more soul shards, and in AoE situations, not to mention that damage per second from the dot does NOT increase faster at haste plateaus. Damage FROM corruption increased by a 100 haste increase is the same whether it crosses a haste cap or not, assuming you renew it before it fades.
Rogue Forum Moderator (formerly Mugajak)
Former Electric Sheep Recruitment Officer (They're Great!) - Kaeldraens
A good post bringing up something that I've forgot during this discussion. This goes further into explaining how haste benefits us outside of extra dot ticks.
I'll make a summary here soon regarding what we've discussed about haste to allow everyone to add what I've missed. I don't want to update the OP quite yet before getting these things confirmed.
---------- Post added 2013-01-09 at 03:27 PM ----------
So, a short summary regarding the haste topic:
- The statement that we don't benefit from haste plateaus was false. Explained by SirFlipper in this post.
- While we do benefit from extra dot ticks we won't always get all ticks, because of Pandemic. Consider it a way of clipping the dots. As an example let's say we have a dot with 16 ticks on a 30s duration and then if we push for next plateau for a total of 17 ticks (with same duration, to keep it simple) which we are refreshing at exactly 15s left because we can buff the dot with procs we are still only getting 8 ticks just like we would've at the previous haste plateau. (Feel free to add your thoughts on this)
- Haste still benefits us between haste plateaus because it will make our hardcasts (Haunt, UA and SoC) and mainly our channel spells (MG and DS) faster. Dots will also tick faster because their duration will be lower (and thus deal damage faster). Because of this we don't have to stop stacking haste after reaching certain haste plateaus.
- Mastery scales synergistically with haste, but it still hasn't been confirmed if we at some point would prefer getting more of either of the stats.
- With all this in mind, we will now again consider SimCraft as a solid source for stat weights. We also haven't had any good arguments to not trust it when it tells us to focus more on either haste or mastery.
I probably forgot something again and I might need some math to back this up and make it worth having it in the OP as solid statements for guidelines that we consider should be followed from now on.
From here on after having shed some light on this topic it would be cool to figure out the value and especially scaling of mastery. I'm not sure where to start but it feels like we haven't gotten far with it. Some people have claimed that it's what's suggested to stack for multi dot scenarios where we can't use SoC, but I've not seen any math for this yet. What exactly makes our mastery better is what I'd like to get on paper, so to speak.
Last edited by Bonkura; 2013-01-09 at 07:14 AM.
good point mugajak - hadnt thought of it like that! ok, so expanding on that, haste plateus may or may not be important (im inclined to believe not more often) but their importance will vary depending on fight length. also, standard breakpoint calculations use base duration, which i guess is wrong for us due to pandemic? the biggest benefit i see for a breakpoint is an additional tick of highly buffed DoTs, and we will always try to get the max duration of those buffed DoTs if possible, so our calculations should use 1.5x base duration instead?
just spitballing here...
The point that has been made is that if you always reapply your dot with the same duration left you will also get a full duration of that dot. We just have that wiggle room to play with due to Pandemic making no guarantee of benefiting of extra ticks, but most times we are more likely to get them than not.
Yes.
No. The number of ticks is completely irrelevant (unless you are in a multidotting scenario where you do not refresh). The important thing about extra ticks is the amount of GCDs we gain because we have to refresh our DoTs less often because they last longer. I demonstrated some pages ago that Pandemic doesn't change the amount of GCDs we gain. If you keep the duration the same "to keep it simple"
Yes.
One very important thing to keep in mind: To calculate the scale factors simulationcraft adds 1 000 of the specific rating and then divides the difference in DPS by 1 000. With all the Haste breakpoints you will always have another breakpoint within 1 000 Haste rating, which then distorts the Haste rating scale factor (by a factor of (breakpoint_value/1000)). This is very significant when you are comparing two relatively close stats like Haste and Mastery. So basically, the scale factor Simulationcraft provides is only really valid if you can get over the next Haste breakpoint
I have not yet seen any math that supports Haste as the best stat for multidotting.
No, because calculations ingame are based on the base duration and then add Pandemic to that in the end, not the other way around.
But the "same duration left" happens at different "time elapsed" of the DoT (and thereby more or less often) which can be seen in my post about 4 pages ago.
Check the scale factor section in every simulationcraft report, for example [url="http://www.flippworld.de/simc_report_s.html"]this one[/quote] where it says "Scale Deltas" in the scale factor table.
we also need to keep in mind that the haste breakpoints will also allow us to get more out of our haunt, if we have enough haste to get another tick from our dots or malefic grasp while our haunt is up, with the 4717 haste breakpoint for example we get around 4,9ish ticks or more precisely 4 ticks out of corruption when haunt is up and 8,88 or 8 bonus dmg corruption ticks from malefic grasp, this is just some napkin math ive made, but we get about 4 ticks of each our dots at 4717 haste when haunt is up, being able to make it 5 is a sizable dps increase single target when haunt is up, and can be another 5k+ dps when haunt is up and assuming with about 70% uptime on haunt, it can be another 3,5kish dps increase overall. ofc this isnt game breaking, and is just some math on the top of my head, so it prolly isnt very precise but it is worth factoring in i think.
Last edited by mmoca748dddcc2; 2013-01-11 at 02:31 AM.
afaik i never say haste would be more useful than mastery for "getting more out of haunt", what i did say that it was worth factoring this in when it comes to haste breakpoints. i dont see where i in my post ever mentioned anything about mastery, in fact the discussion was about haste breakpoints.
In theory that might be true, but considering Haunt lands at random times during our DoTs durations and our MG channel I highly doubt this is realistically managable by anyone during an actual encounter.
I dont think you can have a dicsussion about haste without taking mastery into account, afterall that haste is coming from somewhere right? And furthermore my real point is that haste plateus changes realistically nothing about the amount you get out of haunt, because it does not increase the dps of the dots. It basically just boils down to "more haste means you get more dps out of haunt, just as you do when you gain any other stat".
Last edited by Micke; 2013-01-11 at 09:40 AM.
by that reasoning, the opposite is also true, you cant have a discussion about mastery without taking haste into account.
anyways, you're still wrong tho when you say "haste plateus changes realistically nothing about the amount you get out of haunt, because it does not increase the dps of the dots" as haste does increase the dps of dots, math proves that, ofc i dont know if it was just sarcasm, and im inclined to believe that or you have no concept of how haste works.
i know that, but you can take steps to counter it, such as going into melee to greatly reduce the travel time of haunt and make it more predictable but yeah, you do have a point.
Last edited by mmoca748dddcc2; 2013-01-11 at 11:15 AM.