Whether it's worth the money for any individual person is up to many factors, from how much you value your dps to how many spare dollars you have. It's a definite upgrade from a pve perspective. I would never change race from night elf, personally, and remain hopeful that Blizzard will at some point offer a small bone to the melee specialist race of ancient elven warriors, but even without that night elf offers things that other races do not in other areas (generally not ones related to raid dps, obviously).
I'll also offer an anecdote, because I have a story for everything: On our guild first kill of Heroic Chimaeron (last week), the fight ended with me evasion tanking (glyphed evasion just in case), and one of Chim's big punches with a *miss*, and the others were dodges. With evasion up, Chim can still hit me- but did that 2% night elf miss save us? Well, maybe. The hit table would have had it be a dodge I'm pretty sure, but maybe one of those dodges would have been a hit! Well, I can hope anyway :P
im confused right now, by reading all this math .. is it better to rupture 4cp+ or 1+??
His claim is that 1+ is likely better, because low CP ruptures last for so long (because of the base duration + glyph) and because ruthlessness does not care how many CPs you used, it returns one 60% of the time... and he believes this makes up for the two rewarding factors for high CP rupture- greater rupture damage per tick, and more energy returned via relentless.
In his 1 CP case, he only needs to consume 3.33 CP, using his averaged out system, over the course of his 100 second run.
In his 5 CP case, he needs to consume 22 CP, using his averaged out system, over the course of the same run.
In the 1 CP case, he consumes 8.33 Ruptures * 25 energy /rupture = 208.25 Energy, and it is free 1/5 of the time (relentless), for 166.6 energy cost over his 100 seconds.
In the 5 CP case, he consumes 5 Ruptures, and all of them are free (relentless), for 0 energy.
He then says, ok, well, we can spend that energy on mutilates, and that makes like 7.57 CP, which you can subtract from the 22, and still comes to the conclusion that costing about 11-12 ish combo points over the 1+ rupture case makes up for the damage from the three mutilates and the higher rupture damage, because they can be turned into envenom uptime.
I'm not sure if I find it compelling over currently accepted theory (4+ rupture is best, but rupture at any combo point level if rupture falls off, because missing ticks costs you more dps than the little rupture will). 166.6 energy over 100 seconds is close to 10% of your total energy. We shouldn't stop there: we should work out how much extra envenom uptime we have, then bring in actual damage values for high and low ruptures and those extra mutilates, and see which is better. My impression would be that simcraft takes this into account already, but that's just a guess.
I think a major problem with this sort of analysis is that its entirely focused on energy/cp generation, when in reality the exact damage done by various moves is pretty important. For instance, when last I ran the numbers, the damage directly done by VW ticks accounted for roughly as much total damage as the energy return, and I see no accounting for such things here.
I'm not saying the conclusions are necessarily wrong - it's been a while since I've run numbers on it - but I do recall that the differences between ruptures of various sizes is pretty small - we're talking a couple of percent. So I confess myself skeptical of any analysis that generates a clear differentiation while omitting major aspects of the model, as it makes me suspect that the difference observed owes more to what was omitted from the model than to any actual difference.
What you say is true, and the calculation is only based on CP/energy regen and does not take many things into consideration.
I think it's still safe to say for example: in a fight you do an envenom, you get 1CP (due Ruthlessness) and your rupture is about to fall off, it's better to reapply the rupture with 1CP instead of doing a mutilate to get 4CP rupture up.
I think it's still safe to say for example: in a fight you do an envenom, you get 1CP (due Ruthlessness) and your rupture is about to fall off, it's better to reapply the rupture with 1CP instead of doing a mutilate to get 4CP rupture up.
This was my conclusion as well, since RNG can sometimes leave you dry of CP and energy after envenom. If rupture drops shortly after, I've found it can take longer than the two second threshold to refresh at 4-5 CPs. If we're to maximize envenom and rupture up-time, sacrifices must be made on occasion.
I think a major problem with this sort of analysis is that its entirely focused on energy/cp generation, when in reality the exact damage done by various moves is pretty important. For instance, when last I ran the numbers, the damage directly done by VW ticks accounted for roughly as much total damage as the energy return, and I see no accounting for such things here.
I'm not saying the conclusions are necessarily wrong - it's been a while since I've run numbers on it - but I do recall that the differences between ruptures of various sizes is pretty small - we're talking a couple of percent. So I confess myself skeptical of any analysis that generates a clear differentiation while omitting major aspects of the model, as it makes me suspect that the difference observed owes more to what was omitted from the model than to any actual difference.
Does Rupture damage affect the damage of Venomous Wounds though? It may and I just don't know, but the tooltip put me under the impression the Rupture's damage based on CPs didn't affect Venomous Wounds damage.
Does Rupture damage affect the damage of Venomous Wounds though? It may and I just don't know, but the tooltip put me under the impression the Rupture's damage based on CPs didn't affect Venomous Wounds damage.
The Rupture itself does not affect the damage of VW, it only gives it a 30% chance to proc. VW is a poison, and thus increased by Mastery, I'm unsure if it scales with AP, though.
Yes. As I summarized it here in unfortunately innuendo-laden fashion: uptime is more important than size, but bigger is better.
I'm interested in how you calculated, it will be probably more advanced than mine. But my findings are actually saying the opposite of yours: Rupture uptime is most important, but smaller is better.
So, I dug up my scratchwork on this - its based on a rather specific set of gear and some numbers pulled out of Shadowcraft, and I have no information about what that gear set may have been - I would assume it was near-BIS T11, but I can't say for certain. So the answers may be a bit different for any given person than I found here, but here's the sort of thought process that went into it, and at the very least it demonstrates the magnitude of the differences we're discussing.
The basic idea here is that under a certain set of idealized assumptions, we're going to compute the marginal value of a point of energy and a combo point, and then assess the quality of our cycle and various decisions by measuring inefficiencies relative to that theoretical ideal. This was an approach I originally developed for Combat, and I've done somewhat of a crude job adapting it to Mutilate - if someone would like to clean up some of the rough patches, that might be good.
So, first: what is our "ideal" cycle as Mutilate? Well, ideally, 1) All Envenoms are overlapping 2) No combo points are wasted 3) Rupture of whatever the most efficient size is will be maintained with perfect uptime and 4) All surplus combo points will go to Envenom. Note that this isn't actually possible - for instance, if we conclude that 5 point ruptures are the way to go we're assuming that we're doing *all* 5pt ruptures with *no* downtime and *never* wasting combo points (by, say, getting a SF crit at 3 CP) - totally impossible. But, of course, we're not claiming this is possible - just that its a theoretical cycle we can compare to. There are some problems with this assumption as Mutilate, of course - it worked better for Combat than it does here - but it gives the right idea).
Numbers pulled from Shadowcraft: a Mutilate gives 2.700 CPs, does 16403.00 damage (on average), and costs 55.61 energy (counting cost of reapplying due to dodges). Hence we gain .04855 CP per energy spent on Mutilate, and Mutilate converts energy to damage at a rate of 294.95 damage per energy.
At this point we should really figure out the optimal size of envenom and compare to that, but I don't appear to have scratchwork handy for that so as a hack we're going to use the average size envenom we actually see (assuming 4+ finishers), which is 4.5390 CP. This does, on average, 28690.54 direct damage and gives an extra .32096 poison procs or the ensuing 5.5390 seconds, which results in an extra 7577.01 damage. The net energy expenditure (after Relentless Strikes) comes to 12.663 energy, and the CP expended (accounting for both Ruthlessness and the combo points we *didn't* earn by spending those 12.663 energy on Mutilate) is 4.5537. Hence, we gained an extra 36267.55 - (12.663 * 294.95) = 32532.71 damage by spending that energy on Envenom instead of Mutilate, and thus gained an extra 7144.24 damage per combo point expended. The total value of 1 point of energy is thus 294.95 (the amount of damage we'd gain by spending that energy on Mutilate) plus .04855 (the number of CPs we'd get from spending the energy on Mutilate) times 7144.24 (the amount of damage each combo point so gained gives us), for a total of 641.77 damage per energy.
Thus: we now know the damage value of a point of energy, and the damage value of a combo point; we can thus compute the real damage gain of doing, say, a 1pt rupture. A 1 pt rupture spends .4 CP and 20 energy, which works out to 15693.12 damage of lost opportunity cost (i.e., the energy and CP would do 15693.12 damage if we spent them on Mutilates and Envenoms instead), and gives 3.6 VW ticks; each VW tick does 7416.78 damage directly, plus 6417.71 damage from the energy it gives; the Rupture itself does an additional 4521.96 damage. Hence, the total net damage gain from a 1pt rupture is 26700.41 (direct VW damage) + 23103.76 (VW energy) + 4521.96 (rupture) - 15693.12 (opportunity cost) = 38633.01 net damage done. Since a 1pt rupture lasts 12 seconds, this works out to 3219.42 damage per second gained by keeping rupture up continuously with 1pt ruptures.
Repeating this last calculation for ruptures of various sizes, we get the following table:
Size
Net Damage Per Second
1
3219.42
2
3282.39
3
3322.46
4
3351.69
5
3375.32
That is: a 5pt rupture is roughly 4.8% more efficient in terms of damage per time than a 1pt rupture. As a more useful way of looking it: while a 1pt rupture is up, you are doing 155.9 damage per second less than you optimally would; over the 12 second duration of the rupture, this costs you 1870.82 damage. This is equivalent to the amount of damage you lose by having rupture down entirely for 1870.82 / 3375.32 = .554 seconds. Hence the conclusion: it is better to use larger ruptures than smaller ruptures, but its better to use a 1pt rupture than to wait even a single GCD to apply a larger one.
Thank you for your explanation Aldriana, but there are a few steps where my brain can't keep up with your speed.
I hope there aren't any mistakes in my conclusions.
Mutilate:
CPs: 2.700
Dmg: 16403.00
Energy cost: 55.61
1 Energy = 294.96 Dmg + 0.04855 CPs
Envenom size: 4.5390
CP gain from Ruthlessness: .6
Energy-equivalent CP: .6148
Net CP expendature: 4.5390 - .6 + .6148 = 4.5537 (discrepancy is due to rounding).
Basically: the 4.5537 number already includes the .6148 CP from energy not spent on Mutilates.
Then 32532.47 / 4.5537 = 7144.24 (again, rounding) as above.
I wanted to share a few general thoughts about gemming. The easiest way to describe it, is with a simple example:
Stat
EP
Agility
2.6
Yellow Hit Rating
1.75
Spell Hit Rating
1.4
Mastery Rating
1.3
Haste Rating
1.2
Expertise Rating
1.1
Crit Rating
0.9
White Hit Rating
0.75
MH DPS
3.36
OH DPS
1.04
Yellow
Spell
White
3/3 Precision
241 (2%)
1127 (11%)
2523 (21%)
Let's use these values from the introductionary post. You get [Wind Dancer's Legguards] and have 2 options, straight up agility or matching both socket colors for the socket bonus.
Using the above values, we get:
80 Agility = 80*2.6 EP = 208 EP
or
60 Agility + 20 Hit + 20 Mastery = 60*2.6EP+20*1.4EP+20*1.3EP = 210 EP
Seems like matching both socket colors is the way to go. Especially when people just replaced an item with hit on it, they tend to use the 'below-cap' value for hit rating (same goes for expertise rating as combat).
Let's assume our rogue A has 1107 hit rating and 2020 mastery after equipping the pants with two empty sockets.
He achieves 1127 hit and 2040 mastery after matching both socket colors.
Using 1107 hit rating and 2020 mastery rating as the base values just seems wrong to me. He should have reforged to 1127 hit and 2000 mastery rating before calculating the EP values. After matching the socket colors and reforging again he would end up with 1127 hit and 2040 mastery rating but would only have gained 40 mastery rating.
This would mean that our calculation was flawed: 60 Agility + 40 Mastery = 60*2.6EP+40*1.3EP = 208 EP should be the right calculation for this scenario. Now that's only applicable to the best case, where you can transfer all the excess hit into the best non-cappable stat. In some cases you will end up somewhere between the two best non-cappable stats or even a few rating over a cap. Similir logic applies when we consider using a 20 Agility/20 Hit gem with a hit value over the cap. In this case hit would be reforged into mastery giving the gem an EP of 78 instead of 67.
The general rule of thumb should be updated to something like: 'When gemming/enchanting a cappable stat, always use the EP value of the best non-cappable stat.'
Cases where you can't reach the hit cap with reforging and end up way too high or low shouldn't be that common. It's not that big of a deal for assassination with mastery rating this close to spell hit rating but should be considered when optimizing.
You're approaching it the wrong way, in that because of reforging, you always get the better stats.
In short, the better the stats you start with, the better the outcome. So if you start with a good EP, even if you reach spell hit cap, you end up with a little less EP (because of reforging), but still better than the other options, because they start with less.
It is the same way to think about items really. We havent crossed the point where we have so much hit on gear that we're unable to reforge out of it. So I really dont think this is necessary.
Let's assume our rogue A has 1107 hit rating and 2020 mastery after equipping the pants with two empty sockets.
He achieves 1127 hit and 2040 mastery after matching both socket colors.
Using 1107 hit rating and 2020 mastery rating as the base values just seems wrong to me. He should have reforged to 1127 hit and 2000 mastery rating before calculating the EP values. After matching the socket colors and reforging again he would end up with 1127 hit and 2040 mastery rating but would only have gained 40 mastery rating.
You actually gain 20 agility as well for matching your sockets and you don't gain 40 mastery rating, you gain 20 and 20 from some other stat (it'll be worth the difference).
Let me back up and start by pointing out that you actually get the same dps for the particular example of these legs. In the case of matching your sockets you gain 60 agility, 20 spell hit, 20 mastery, and 20 mastery from 20 crit (I am going to assume you're always reforging from crit somewhere in your gear).
In the case you decide to reforge to hit cap before hand and then gem straight agility you gain 80 agility and 20 spell hit from 20 crit.
80(2.6) + 20(1.4-0.9) = 218 EP
Lets take a look at the same situation with a different piece of gear such as [Wind Dancer's Tunic]. Here we have a red and blue socket and a 20 agility bonus. Matching the sockets gives us 80 agility, 20 spell hit, and 20 mastery from 20 crit. Reforging to cap before gemming gives us 80 agility and 20 spell hit from 20 crit. You can already see without calculating that reforging beforehand is not the right thing to do.
I can't think of the best way to explain this but it may help to not see reforging and gemming as the same thing and instead view reforging as a way to convert lesser stats into better stats and gemming as a way to add stats (including agility). By reforging to cap, you void some opportunities to gain extra EP at least in the case of red and blue sockets with an agility bonus. It appears that for yellow and blue sockets with an agility bonus or single blue sockets with an agility bonus you end up with the same overall EP.
I think for practicality, most are going to gem first then reforge since you do not have to worry about missing potential gains from socket bonuses and the fact that you only need to juggle reforging to maintain caps instead of both reforging and gems.
Last edited by ieatpaperbag : 03/23/11 at 12:26 PM.
I think for practicality, most are going to gem first then reforge since you do not have to worry about missing potential gains from socket bonuses and the fact that you only need to juggle reforging to maintain caps instead of both reforging and gems.
Well, yeah. You'd end up losing stats if you reforged THEN gemmed, if you think about it. On the t11 legs for example, it's slightly better to run 20/20 gems if, and only if, the hit will go to spell hit. Then you can reforge the hit into mastery (or haste, if that's the case) if needed.
That, and something else to take into consideration, is that after you reforge - And I've done this a couple times - If you just *cannot* get your hit to a nice place, which would be somewhere between 1322-1352 (cap being 1332, 2/3 pres), you can always pull the gem out after you reforge if your super worried about that 5 extra DPS.
Well, yeah. You'd end up losing stats if you reforged THEN gemmed, if you think about it. On the t11 legs for example, it's slightly better to run 20/20 gems if, and only if, the hit will go to spell hit. Then you can reforge the hit into mastery (or haste, if that's the case) if needed.
This was in reference to King-Slide's post of whether it was better to gem and then reforge to hit cap or to reforge to hit cap and then gem. If I am understanding what you're trying to say, reforging to hit cap, then gemming with 20/20 gems, and then reforging back down to hit cap; that's really the same thing as gemming then reforging to hit cap. The only way you would be able to use the following is if you were above spell hit cap with gems but no reforging.
This was in reference to King-Slide's post of whether it was better to gem and then reforge to hit cap or to reforge to hit cap and then gem. If I am understanding what you're trying to say, reforging to hit cap, then gemming with 20/20 gems, and then reforging back down to hit cap; that's really the same thing as gemming then reforging to hit cap. The only way you would be able to use the following is if you were above spell hit cap with gems but no reforging.
True. Though I don't think we'll see any of that anytime soon..... LOL. Every piece reforged to hit as it is. Good to know though.
I do have a question, and it's something I've been playing around with recently, haven't found an answer to it. I know a lot of dots we're changed in Cata so you..in essence..couldn't "clip" them anymore (read: Flameshock). Meaning when you reapply the dot, it doesn't reset tick times, it just resets the duration (I believe this is only to a point, it has to have less than X seconds or X% or some variable). Does Rupture do this? I notice a lot of times that I pop an envenom, hit Mut, and I'm pooling energy, then my rupture ticks down to 3 seconds but I'm almost energy capped. What would be the technical better thing to do?
A) Hit envenom, pray for a ruthlessness/mut crit to get back to 4cp quickly and refresh rupture >2s of it dropping
B) Clip rupture and don't let it drop period.
Also, one other scenario: You're at 2CP, you have ~60 energy, rupture just fell off. Do you:
A) Hit rupture immedietly
B) Waste one GCD to Mutilate without rupture being up and apply a 5CP rupture.
I guess this is more a question of is having Envenom buff more important than losing rupture for, at the most, 1 GCD, since you'd use any current points to toss back in. Obviously a high CP rupture will be better than a low CP, so you might be better off having it drop for one GCD and doing more damage, either way I'm sure the difference is very minor, but you never know, eh?
Let’s say that you are sitting on 4 cp’s, you have 3 seconds left on your rupture, and are sitting on 70 energy. Knowing that an envenom cost you 35 energy, and with 4 cp’s you’ll be refunded 20 energy by relentless strike (normalized – 20% chance to refund 25 energy per combo point spent, with 4 cps you will receive 25*(20%x4) = 20 energy), this scenario puts you at 70-35+20 = 55 energy.
A Mutilate costs 60 energy, taking the gcd’s it takes you to perform the envenom into account, and possible venomous vim procs, you will most likely be able to mutilate again straight after the envenom, as it takes you less than a second to gain the 5 energy missing. This will leave you with 2-4 cps from that mutilate on the target (based on a crit and/or ruthlessness), even before the 3 seconds of your initial rupture drops you will have sufficient amount of cps, whether or not you get a Ruthlessness proc or crit, you should reapply your rupture with whatever cp’s available, as a low cp rupture is better than no rupture.
However, this is very situational based on your energy, if you take the same example as above, but you were sitting on 50 energy, you might be better off pooling, because if you were to do the same, you would need to rely heavily on venomous vim procs, relentless strike and ruthlessness to gain the energy you need to not let rupture drop for a second or two.
In the other scenario, where you are talking about clipping rupture, I will have to say that this is not always possible, if your previous rupture is applied with 4 cp’s you will not be able to overwrite it with another 4cp’rupture. And if you have to wait for those 3 seconds, you will most likely cap on energy and also lose envenom uptime, which is a dps loss. And if you previous rupture is a 3cp one, you will still be better off envenoming and then get whatever cps you can get before the rupture drops completely, and then reapply it with whatever you got (even if it’s just one, yes , this is not optimal dpe, but it’s still better than a few seconds with no rupture)
As for your other question, I think that is answered by this, rupture with whatever cps you have available, it’s not mandatory to use a 4 cp rotation on your rupture, as the it’s the venomous wound procs you are after, which hits for the same amount regardless of the rupture size, the damage from rupture itself is not hurting your dps by a lot if it’s a low instead of a high rupture. This is also elaborated by Aldriana in this post.
Below is a table of the damage lost for various actions.
Action >>>>> Estimated Damage Loss
Leaving Rupture down for 1 second >>>>> 3375
Since I cannot use a smaller rupture over a bigger one (A 4 CP Rupture while a 5 CP one is running, for instance) and I play with 200ish MS; it means I'll be losing a lot of damage everytime I have to wait my rupture to fall to apply a new one?
Slightly off-topic, I noticed a weird bug with Cold Blood a while ago and it's still around. If you don't wait a GCD and cast Cold Blood right after using mutilate / backstab (~200 ms later), then Cold Blood buff fades from your character instantly. I checked logs and it doesn't make that Mutilate or the following Envenom critical, it just generates 25 energy and vanishes from your character within 20 ms.
I play with 50-ish global latency so it shouldn't be a factor. I usually avoid using Cold Blood in a manner like this, but when I get impatient, this bug occurs consistently.
As Sculduggery said, you'd be in combat before you were in position to use that macro almost every time, and therefore, unable to pot again during the fight. You might try this macro instead just before the fight starts.
#showtooltip Potion of the Tol'vir
/use Potion of the Tol'vir
/cast Stealth
This way you get the full benefit from Overkill, you can open with Garrote, and you only lose a second or 2 of your prepot. Be sure you're not in stealth when you cast it as that would defeat the purpose.
#showtooltip Potion of the Tol'vir
/cast Potion of the Tol'vir
/cast Stealth
/cast [target=focus] Tricks of the Trade
/cast Sprint
This depends on strategy and fight. You can change the tricks of the trade to how ever you like tricksing the tank. When 4.1 hits this will be used even more because tricks will be 100yrds instead of 20yrds. Some fights in cataclysm you will not be close enough to the tank at the start(alakir depending on strategy).
#showtooltip Potion of the Tol'vir
/cast Potion of the Tol'vir
/cast Stealth
/cast [target=focus] Tricks of the Trade
/cast Sprint
I can see the benefit of adding Sprint, but not the Tricks. If you're tricksing the tank, you can easily do it in advance of potting. And if you're tricksing a dps partner, you'll want to wait a few seconds for the tank to get aggro and your partner to build up his buffs. I don't know if there's an optimal time for waiting. I'm pretty busy right at the start of a fight gcd wise. I'd say I hit Tricks about 8 seconds in. My partners poisons will be stacked and his trinkets active at that point, so the TotT does more for him than it would at the beginning of the fight.
Aldriana, wish I could do this myself but I have no idea how without running tests with my current gear setup on the dummy (which is so accurate). Do we know the damage difference between a 4CP and a 5CP Envenom? I'm looking to see about the 4pc t11, whether when it's active you'd almost be better off gaining an extra CP to use Envenom with, as the loss of CP's from Mutilate would be made up for with an extra CP on Envenom. Or possibly it's better off to use Shiv instead to gain that last CP, as I know the loss from Mutilate is almost 20k damage. Btw - Did you do these tests in a BiS 372 set, or a 359 set?
03/14/11, 5:50 PM
Similar Threads
