With the changes to shadowform allowing DoTs to scale with haste it will depend on how much haste you have:
* Assuming zero haste (SWP ticks every 3 seconds) then it will be approximately 65 mp/5
* Assuming 25% haste (SWP ticks every 2.5 seconds) then it will be approximately 80 mp/5
Actually 25% haste on the PTR atm lowers the overall debuff length by 25%, which means SW:P decreases from 18s to 13.5s, creating 2.25s DoT ticks. Or if you'd rather think of it as 50% haste = 1.5s DoT ticks.
Of course DP is bugged on the test server, but SW:P and DP behave this way.
Also something I would consider a bug is the fact that when MF refreshes SW:P it resets it to the original 18s mark instead of your proper hasted amount. Let's hope they change this, otherwise we might actually have to consider the overall usefulness of that talent :P
Actually 25% haste on the PTR atm lowers the overall debuff length by 25%, which means SW:P decreases from 18s to 13.5s, creating 2.25s DoT ticks. Or if you'd rather think of it as 50% haste = 1.5s DoT ticks.
Of course DP is bugged on the test server, but SW:P and DP behave this way.
Also something I would consider a bug is the fact that when MF refreshes SW:P it resets it to the original 18s mark instead of your proper hasted amount. Let's hope they change this, otherwise we might actually have to consider the overall usefulness of that talent :P
Please test things properly before claiming stuff.
First of all it 25% haste doesnt lower the overall debuff length by 25%. What would happen if you reach 100% haste then?
100% haste will give you 1.5s ticks, not 50% haste.
Haste affecting DP does work. The bug is that the duration of DP stays at 24 seconds so what happens is you end up getting more ticks out of a single DP when you raise your haste. With 695 haste I'm getting 10 ticks per DP instead of the normal 8.
SW:P refreshes through MF also work properly. The bug is that the duration of SW:P gets reset to 18 seconds instead of the hasted duration of a fresh SW:P. The dot still correctly ticks faster, however.
SW:P refreshes through MF also work properly. The bug is that the duration of SW:P gets reset to 18 seconds instead of the hasted duration of a fresh SW:P. The dot still correctly ticks faster, however.
This is indeed how it works. Testing via combatlog can of course be a crapshoot on the PTR with fluctuating latency, but with 550 haste, the expected tick interval would be ~2.5 seconds, and I was seeing a range of 2.4 to 2.7 in my log even after refreshing with Mind Flay. It never reset to 3.0, even though the SWP duration did get set to the base 18 seconds.
Berserking brought the ticks down to a 2.0-2.2 second range and it seems they stayed there through Mind Flay refreshes of the same SWP.
Berserking brought the ticks down to a 2.0-2.2 second range and it seems they stayed there through Mind Flay refreshes
of the same SWP.
Hmmm....this would imply that we'd be able to "roll" a Heroism'd/Bloodlust'd SW:P then yes? Interesting.
I'll have to take a look at making minor changes to SimulationCraft to model both persistent haste buffs as well as DP (and refreshed SW: P) lasting the pre-hasted duration and having more ticks.
Please test things properly before claiming stuff.
First of all it 25% haste doesnt lower the overall debuff length by 25%. What would happen if you reach 100% haste then?
100% haste will give you 1.5s ticks, not 50% haste.
Haste affecting DP does work. The bug is that the duration of DP stays at 24 seconds so what happens is you end up getting more ticks out of a single DP when you raise your haste. With 695 haste I'm getting 10 ticks per DP instead of the normal 8.
SW:P refreshes through MF also work properly. The bug is that the duration of SW:P gets reset to 18 seconds instead of the hasted duration of a fresh SW:P. The dot still correctly ticks faster, however.
Actually I did test it - a lot this morning on the dummy
890 haste rating (T8 MB bonus) = 27% haste
My tested SW:P was just over 13s long when cast this way (timed it via needtoknow and a stopwatch repeatedly)
.27 x 18s = 4.86. . .subtract from 18 and you get. . 13.14s
I also did this at two lesser amounts with normal gear setups and the math matched my timing tests.
I'm not saying this scales up to 100% - maybe at or near 50% you run into DR. But for these levels it seems to be the current relation. (Also I know this is not how typical haste works for cast-times - this is a new realm for haste tho and I'm keeping an open mind based on data on how it works).
What I'm seeing on the PTR at the moment with Devouring Plague is this:
It seems to first calculate the number of ticks it should cast:
num_ticks = round( 8 * (1 + Haste%) )
Then, the *reported* duration of the DoT is fixed at 24 seconds. i.e. if your DoT timer queries the UI it'll say at the start that the DoT has 24 seconds left.
However the *actual* duration is: num_ticks * ( 3 / (1 + Haste% ) )
So for example at 204 haste rating == 6.22% haste:
i.e. The DoT will both have it's last tick and will SPELL_AURA_REMOVED after 22.59 seconds from when it was applied (with variance for latency etc of course). So your DoT timer will suddenly see the DoT finish before it had finished ticking down to 0.0.
So here, with the addition of +1 haste rating you gain an extra tick of DP with your cast however the actual duration has now increased to 25.411 seconds which is *more* than the reported duration. It also means that your DoT timer will expire before the last tick and if you were waiting to recast the DoT as soon as it finished you'll clip your last tick and thus miss out on it.
In terms of the DPS gain of num_ticks/duration:
At 204 haste rating: 8/22.59 == 0.35414
vs
At 205 haste rating: 9/25.41 == 0.35419
i.e. In terms of just num_ticks/duration the DPS increase is smooth.
However of course, the longer duration between recasts at 205 haste rating means less GCDs spent casting it and thus more GCDs available to cast other spells. It also improves the mana efficiency. Mind you, with current DoT Timers it also increases the chances you'll clip your last tick.
That formula is not the one used for determining duration of the dots, it works exactly like the formula for cast-times.
Duration of the dot / (1 + Haste%)
I have 21.23% haste, let's take VT as an example with T9 set bonus
21 / (1+0.2123) = 17.3 - this is what i saw ingame on PTR
Your formula: 21 - 0.2123 * 21 ~ 16.54 - which is incorrect
Ok I stand corrected - I ran a series of tests tonight after raiding at various haste values, and tried to refine my timing a bit more with the debuff length stopping when the debuff fades rather than when the final tick "pops" and they matched almost perfectly with your formula. I suppose the addon data lag is to blame for my earlier numbers, which coming from the test server environment I should have predicted - my apolgies for the prior confusion.
Originally Posted by Althor
So here, with the addition of +1 haste rating you gain an extra tick of DP with your cast however the actual duration has now increased to 25.411 seconds which is *more* than the reported duration. It also means that your DoT timer will expire before the last tick and if you were waiting to recast the DoT as soon as it finished you'll clip your last tick and thus miss out on it.
I noticed this with DP tonight too - in fact the debuff will just "hang" for that extra second or so until the final tick is resolved - quite annoying! I hope DP is just a work in progress seeing that the talent change did not make it in this build as well.
*edit* When increasing my haste value from 650 -> 890 however, I gain a 10th tick and the debuff resolves properly at 24s.
From testing on the PTR I've noticed that (finally) spellpower derived from Spirit as well as the 2pc T8 set bonus now correctly increase the damage of Improved Devouring Plague.
Haven't been able to test the VT backlash damage yet.
Anybody else incur a very sporadic loss/gain of base stats in comparison between Live and PTR ( charterer copy in case you were wondering )? I am trying to compare personal numbers on test dummies but its kind of hard when you lose ~130Int/Spirit and gain 80 Strength and 160 Agility.
EDIT: Ok, somehow I managed to skip a page in my reading, was mentioned and I just simply missed it . My bad XD
Last edited by Hyperbolicious : 10/15/09 at 7:26 AM.
Althor, I know you ran simcraft for 3 hasted dots to show us our 'dps values', is it possible for you to run simcraft again with the additional announces changes (spirit scaling and imp DP) and calculate the new scaling coefficients? Very interested in seing the new values.
Althor, I know you ran simcraft for 3 hasted dots to show us our 'dps values', is it possible for you to run simcraft again with the additional announces changes (spirit scaling and imp DP) and calculate the new scaling coefficients? Very interested in seing the new values.
so far all i am seeing is a 200-300 dps increase over live
as well as from word of mouth i heard using full haste gems over the current full SP gems would net a significant increase , well i ran 5 parses with full spell power gems , then ran them with full haste gems then ran it again with spell power / haste gems.
well , from what i saw full haste and sp/haste come out just about the same and only very very marginally up from full SP gems.
My question is this. Has anyone actually confirmed it with Sim craft or parsing which gem set up will innevitably be best.
Not talking T10 4 piece theory craft but with the current premade character gear. Kinda hard to test it's accuracy with the stats being messed up and imp DP not being there.
Just curious. I have seen althors posts from the Simcrafts but just looking at that shit gives me a headache.
My question is this. Has anyone actually confirmed it with Sim craft or parsing which gem set up will innevitably be best.
Not talking T10 4 piece theory craft but with the current premade character gear. Kinda hard to test it's accuracy with the stats being messed up and imp DP not being there.
A page back or so, Althor posted the new parsed stat scaling factors, and with BiS T9 gear it shows Haste overtaking SP. So based on that, it would be a net gain in DPS to stack haste gems. I am sure those numbers will even out once we see the new tier of gear and such, though.
this is what i was looking for
Int=0.26 Spi=0.56 SP=1.00 Hit=1.42 Crit=0.84 Haste=1.31
what that basically means is a 20 haste gem is almost identical to a 23 spell power gem
meaning a 23 sp gem is 23 pp and a 20haste gem is 26.2
which is why i am seeing almost no increase in test dummy parsing when switching out haste gems for spell power gems. It also means that when we are able to figure out what the soft caps are for haste and where we start seeing the diminishing returns we can stack on sp gems or sp/haste gems with very little stat value loss.
Keep in mind that the simcraft stat scaling factors are based on our current BiS which favors crit significantly over haste. We'll likely see more haste on our gear if redoing our BiS list for 3.3 scaling factors which will lower haste on the scaling chart, and make it more or less at parity with SP. The only gemming changes I see are replacing SP+Crit with SP+Haste, using spi gems for every 6+ SP or Haste bonus, and making more yellow bonuses.
Profession-wise, engineering is going to become significantly more useful and, if DoTs "remember" haste the way they currently do crit, will probably become mandatory for stacking haste (Accelerators + Lust + PI + Speed Potion) in a really, really fast SWP. Likewise, if that's the case, SWP glyph will be much, much better than dispersion for raw MP5. With all 5 of those up, assuming 30% haste raid buffed (722 haste rating, easily reachable), we'll get 105% haste (unless I'm incorrect in assuming that Lust and PI scale additively with haste from rating), pushing SWP ticks down to 1.46 seconds. An interesting side effect of this is that you get (glyphed) 132 MP5 to Dispersion's (glyphed) ~120 MP5 and will make choosing 3rd glyphs more interesting depending on the length and variety of fights in Icecrown.
With those changes do you think that the value of Shadow Word: Death will further diminish? It seems that with the super powered mind flay the only real situation when you would use it, is when moving.
SWD is already only used when moving, or as a killing blow (Anub scarabs, frost orbs, etc) to proc spirit tap. All things considered, my imp DP hits for greater than half of what SWD hits for (no t8 bonus) and it's going to be doubled in 3.3 + it scales better with SP (44.4% to 42.96) so even that's going to be more useful to use (if you're not clipping ticks of it and you can maintain the mana) when you're running out of fire or moving into position or whatever SWD would be useful for.
I bumped into this SimCraft yesterday and I'm a bit confused about the spriest numbers.
There is no mention of using 4p t10 bonus for that calc, yet the result is slightly higher than a earlier simcraft using t9 gear and t10 bonuses. I know there's been some additional buffs since, but I was under the impression that the buffs(disregarding t10) would only be in the regions of 1500 dps increase? However that simulation seems to suggest far greater gains from talent/glyph buffs alone.
I bumped into this SimCraft yesterday and I'm a bit confused about the spriest numbers.
There is no mention of using 4p t10 bonus for that calc, yet the result is slightly higher than a earlier simcraft using t9 gear and t10 bonuses. I know there's been some additional buffs since, but I was under the impression that the buffs(disregarding t10) would only be in the regions of 1500 dps increase? However that simulation seems to suggest far greater gains from talent/glyph buffs alone.
In 3.2.2 when run with optimal_raid=1 SimulationCraft sim's a Shadow Priest's DPS at 8600 in BiS T9.
When run from Raid_T9.simcraft it's a couple of hundred DPS higher because Raid_T9.simcraft makes use of a Demo lock with Demonic Pact and optimal_raid=1 just assumes Totem of Wrath.
That said, even just using the 8600 DPS figure, 8600 + 1500 == 10100. Which is pretty close to the 10100 in that url.
In 3.2.2 when run with optimal_raid=1 SimulationCraft sim's a Shadow Priest's DPS at 8600 in BiS T9.
When run from Raid_T9.simcraft it's a couple of hundred DPS higher because Raid_T9.simcraft makes use of a Demo lock with Demonic Pact and optimal_raid=1 just assumes Totem of Wrath.
That said, even just using the 8600 DPS figure, 8600 + 1500 == 10100. Which is pretty close to the 10100 in that url.
What exactly is it you're not sure about?
I don't actually play shadow(been "forced" to run dual healspecs since early Ulduar) so I guess I was simply making an assumption that the current DPS was lower than 8600.
I'm still a bit confused about how these new numbers relate to your earlier calcs here, seeing as how that seems to give around 500 dps increase for the setbonus the implication would be that the few additional buffs since have amounted to a further 700 DPS increase?
My earlier calcs were based off the proposed changes to allow our DoTs to scale with haste.
They didn't know the nitty-gritty details of how that was implemented (which was relevant for Devouring Plague).
Nor did those calcs know about the Improved Devouring Plague buff or the Glyph of Shadow buff.
The Glyph of Shadow buff alone adds another 143 spell power to the BiS set.
Since my second lot of results however I've made further changes to SimulationCraft fixing up some bugs that I've discovered, mostly to do with Devouring Plague.
This results in the 3.2.2 BiS Priest set dropping from 8600 DPS to 8504 DPS and the 3.3.0 set which at one point I was simming at 10060 DPS dropping to 9937 DPS.
(When run with optimal_raid=1).
10/14/09, 5:18 PM
Similar Threads
