Frost DPS | Winter of Discontent [4.3] - Page 27

minytrooper · 12/25/11, 11:15 AM

Originally Posted by DaveA50

I just wish my raid would hero post-35% instead of on pull for 10H. It helps a lot for me, but not the overall raid

I'm not sure how many mages/shamans you have in your 10m, but you could try what our healers do for 25H. Hit your Heroic Will when they are counting down heroism, if you are in the other realm when it goes off you shouldn't get it. Then at 35% have a second mage/shaman hit hero again (select pets will also benefit from a second hero). I'm not sure how big of a dps gain it will be, seeing as you will be phase/stunned for 5secs at the beginning of the fight with trinkets etc. proced, but it may be worth a shot. We just have healers do this as it is way more beneficial for them to get it later in the fight with increased damage taken.

CortDK · 12/25/11, 3:35 PM

well we're getting somewhere. You're now agreeing that 2xHB is superior damage to 1x OBL.

Still, you continue to over-value OBL by not factoring in overwritten 2pc procs, 1x obl is (0.45 + (0.45*0.6)) at most is 0.72, but in reality a large number will be overridden. You are fudging the numbers to make OBL stronger than it actually is. In your own log there were 20% overridden. So that's a 20% reduction of the HB portion of OBLs damage, and resulting reduction in RP generation. There's also the matter of spell misses on rime which aren't refunded.

You are also under-valuing HB by neglecting to attribute dps to the increased RP generation, which is higher than the amount you state given the 2pc overrides.

On your follow up speculation, I disagree.

Originally Posted by Lamperouqe

No, we (me and Cort) did on this page, besides, the word "inferior" was never used. We simply agreed on it not being the recommended single target priority. Using the standard priority I have managed 49.3k on LFR ultraxion as well.

We have agreed on no such thing, you are misinterpreting my position on single target. I do recommend using as many HBs as you are able, and gaming unholy runes if you have focus free, as it maths and sims out to be a worthy dps increase. I stated as such in my very first post if you want to go back and read.

Lamperouqe · 12/25/11, 5:08 PM

Originally Posted by CortDK

well we're getting somewhere. You're now agreeing that 2xHB is superior damage to 1x OBL.

No, I am not. Stop interpreting data the way you want to. HB x2 would marginally beat OB without taking KM into account and overlooking the fact that it's not a melee attack.

Once you add KM to the picture, "masterfrost" does not reach the amount of KM's consumed by OB that the standard priority does. You have admitted this before. Every KM moved from OB to FS is a bigger loss in damage than HB x2 is a gain over OB without KM. Also adding the unpredictable nature of KM, latency, player reaction time, spell queueing, and so on, the standard priority will end up consuming KM with OB more often.

HB not being a melee attack also slightly devalues it. Anything that only procs off of melee attacks has no chance of proccing with it and it can miss unless geared for spell hit cap while sacrificing other stats.

Originally Posted by CortDK

Still, you continue to over-value OBL by not factoring in overwritten 2pc procs

I am beginning to doubt your reading comprehension abilities. Replying to you will just end up with me repeating things I've already said multiple times. HB x2 versus OB while low on resources includes space to utilize rime charges, when high enough on resources that overwriting rime starts to become a problem, the "masterfrost" priority will be obliterating as well. Though yes, there is a higher chance of overwriting rime outside of these situations with the standard priority, it is not even remotely close to as big as you make it to be.

Originally Posted by CortDK

We have agreed on no such thing, you are misinterpreting my position on single target. I do recommend using as many HBs as you are able, and gaming unholy runes if you have focus free, as it maths and sims out to be a worthy dps increase. I stated as such in my very first post if you want to go back and read.

You are contradicting yourself, then.

Mhobius · 12/25/11, 9:47 PM

Some things I'd like to share:

- Glyph of Pestilence no longer causes Plague Strike to hit 20% harder, it was fixed some patches ago

- Simulationcraft doesn't take Devotion Aura in consideration, beside it's only ~150 atk power with Bladed Armor, it's still a dps increase

- I really don't understand the point of any spec with no Improved Blood Tap, like 5/31/5 or 4/31/6. They don't bring pretty much nothing better to the table, beside easier gameplay. They DON'T grant more dps than the Improved Blood Tap build, which seems the only spec viable for "Masterfrost", it's still the best even for the normal priority. You don't need longer disease duration, even for AoE fights, because Frost Fever is always up and you don't spread Blood Plague with Pestilence as Frost. I brought this argument some pages ago and I still got no real answer why you shouldn't use this spec. Some people said that it's almost impossible to spend Blood Tap every 30 sec and use it properly, but it doesn't make sense, because it's off the GCD and even with +4 sec delay (and probably more), you're still doing more dps than your whatever spec. I'm playing this build since the day I switched to Unholy Presence haste stacking and it's nothing hard to pull off. I really don't understand why most end game DKs avoid Improved Blood Tap, it feels to me that they don't know how better the spec is, rather than a personal gameplay preference. 7/31/3 pulls ~120 more dps than the other variants and the simulator clearly shows its superiority. Also, an additional rune every 30 sec instead of 1 min provides more burst, which can be used during AoE fights or certain situations.

CortDK · 12/25/11, 11:42 PM

Originally Posted by Lamperouqe

HB x2 versus OB while low on resources includes space to utilize rime charges, when high enough on resources that overwriting rime starts to become a problem, the "masterfrost" priority will be obliterating as well

You're still giving it 100% utilization which is wrong. There are numerous cases where in two GCDs OBL would be a priority over using 2 rime procs in a low resource state where with masterfrost it would continue to use HB. You have to remember masterfrost uses less than half OBLs than a normal priority and spends a majority of it's time using HB.

And you are still ignoring the value of additional RP in your dps calculations.

But what I don't understand is even using your maths it shows HB to do more damage. It will only be higher still when the numbers are corrected. Yet you still think it would be a dps decrease. Could you explain that?

Because it's pretty clear to me that your numbers, although I think they are showing the difference to be lower, support masterfrost and don't refute it.

Lamperouqe · 12/26/11, 1:23 AM

Originally Posted by CortDK

You're still giving it 100% utilization which is wrong. There are numerous cases where in two GCDs OBL would be a priority over using 2 rime procs in a low resource state where with masterfrost it would continue to use HB. You have to remember masterfrost uses less than half OBLs than a normal priority and spends a majority of it's time using HB.

It's not very common to have 4 runes available in 2 GCD's, having a KM proc to make you prioritize the second OB directly after the first OB, or having 2 runes of the same rune type up after the first OB all the while under low resource regeneration. If you are able to utilize at least one rime proc (even if it procs two) before having to OB again, you have already made obliterating an increase.

As uncommon as it is to be forced to double OB under low resources, the chance to proc rime twice in a row overwriting it is 0.45 * 0.45 = 0.2025

Originally Posted by CortDK

And you are still ignoring the value of additional RP in your dps calculations.

Sure, if not capping, you get an extra FS about every 24th HB on average.

Originally Posted by CortDK

But what I don't understand is even using your maths it shows HB to do more damage. It will only be higher still when the numbers are corrected. Yet you still think it would be a dps decrease. Could you explain that?

Because it's pretty clear to me that your numbers, although I think they are showing the difference to be lower, support masterfrost and don't refute it.

Are you saying that KM doesn't affect the calculation? And that HB not being a melee attack doesn't devalue it?

To make things clear, those numbers were calculated without factoring KM in. You refuse to read and comprehend, don't you? I gave you a lot of reasons why KM heavily favours the standard priority, and why HB not being a melee attack puts it at a disadvantage. You need to look at the numbers with those things in mind, not with naked eyes.

Originally Posted by Mhobius

I really don't understand the point of any spec with no Improved Blood Tap, like 5/31/5 or 4/31/6. They don't bring pretty much nothing better to the table, beside easier gameplay. They DON'T grant more dps than the Improved Blood Tap build, which seems the only spec viable for "Masterfrost", it's still the best even for the normal priority.

True, IBT does simulate higher dps than epidemic for the standard priority as well. It's devalued the longer you keep it unused and off CD, as you mentioned, but it's not enough to make it a loss.

There is a bug on BT I've noticed happening for a long time as frost, though. If you cast it immediately after the first death rune finishes refreshing in attempt to use it on the second death rune, the first death rune, which is already active, consumes it. I've learned to play around this by letting the first death rune sit active for like a second before using BT to activate the second death rune succesfully.

CortDK · 12/26/11, 4:33 AM

Originally Posted by Lamperouqe

To make things clear, those numbers were calculated without factoring KM in. You refuse to read and comprehend, don't you? I gave you a lot of reasons why KM heavily favours the standard priority, and why HB not being a melee attack puts it at a disadvantage. You need to look at the numbers with those things in mind, not with naked eyes.

Nice personal insults.

As I said several posts back, all that matters is if the benefits of the DPS gain outweigh the loss of the disadvantages. Stacking mastery significantly reduces the number of KM procs, but doing so nets a DPS gain (albeit a relatively small one of a few hundred dps) because the gain outweighs the con, in that particular example.

Given you have put no maths at all or any evidence at all to support your opinion that the losses would override the gains, I'd suggest it is you that might have naked eyes. The mistake you are making is looking at pros and cons without providing any maths or sims or even in game logs, and assuming the cons support your position.

Regardless, given you are now relying purely on personal opinion instead of evidence and your debating has degraded I'm not spending any more energy to correct you. You can't even fix your basic OBL calculation after a page of arguing.

minytrooper · 12/26/11, 5:26 AM

Originally Posted by Mhobius

Some things I'd like to share:

- Glyph of Pestilence no longer causes Plague Strike to hit 20% harder, it was fixed some patches ago

- Simulationcraft doesn't take Devotion Aura in consideration, beside it's only ~150 atk power with Bladed Armor, it's still a dps increase

- I really don't understand the point of any spec with no Improved Blood Tap, like 5/31/5 or 4/31/6. They don't bring pretty much nothing better to the table, beside easier gameplay. They DON'T grant more dps than the Improved Blood Tap build, which seems the only spec viable for "Masterfrost", it's still the best even for the normal priority. You don't need longer disease duration, even for AoE fights, because Frost Fever is always up and you don't spread Blood Plague with Pestilence as Frost. I brought this argument some pages ago and I still got no real answer why you shouldn't use this spec. Some people said that it's almost impossible to spend Blood Tap every 30 sec and use it properly, but it doesn't make sense, because it's off the GCD and even with +4 sec delay (and probably more), you're still doing more dps than your whatever spec. I'm playing this build since the day I switched to Unholy Presence haste stacking and it's nothing hard to pull off. I really don't understand why most end game DKs avoid Improved Blood Tap, it feels to me that they don't know how better the spec is, rather than a personal gameplay preference. 7/31/3 pulls ~120 more dps than the other variants and the simulator clearly shows its superiority. Also, an additional rune every 30 sec instead of 1 min provides more burst, which can be used during AoE fights or certain situations.

1) you say Glyph of Pest, no longer causes PS to hit 20% harder, that could be your "~120" dps difference easily if the sim you using still has the bugged glyph programmed into it, being as you PS a lot more often with 7/31/3 build.
2) as you stated easier gameplay, having to PS at 30s, 90s, 150s, etc. and Outbreaking at 0s, 60s, 120s, etc. is just simple.
3) for some bosses you may leave the boss for a while to kill an add or what not, allowing the dot to get 2 extra ticks off

On the flip side.

1) single target 7/31/3 might pull ahead (if the sim isn't using the bugged glyph)
2) certain other fights 7/31/3 might pull ahead as well, fights where adds die fast (or should), or burst damage is needed, ie black horn/spine/hagara
3) "masterfrost" (lead spec atm) uses this build

Other points to note, as far as most end game DKs avoiding Imp BT, they have probably played 5/31/5 for quite a while, and may not feel comfortable switching to this other build with the new patch and trying to push progression. To switch to a different build for as little as ~120 dps may not be worth it to them, as they will more than likely screw it up, somehow and forget to refresh BP for the extra 8 seconds or something similar, resulting in dps loss. With progression its best to stay in a "comfort zone" so to speak with rotation and disease times. "Masterfrost" on the other hand was potentially a big enough dps gain to try out, but the priority on it is so strict and unforgiving that some DKs are just going back to the original "builds"

All in all, it's just a matter of preference, the dps difference is so small it doesn't really matter.

Mhobius · 12/26/11, 9:13 AM

Originally Posted by minytrooper

1) you say Glyph of Pest, no longer causes PS to hit 20% harder, that could be your "~120" dps difference easily if the sim you using still has the bugged glyph programmed into it, being as you PS a lot more often with 7/31/3 build.

The dps difference isn't influenced by the glyph, because this bug isn't taken in consideration by the simulator, simply because it got fixed long time ago. I did test this, so it's not a random assumption.

gharnef · 12/26/11, 4:49 PM

Originally Posted by Mhobius

The dps difference isn't influenced by the glyph, because this bug isn't taken in consideration by the simulator, simply because it got fixed long time ago. I did test this, so it's not a random assumption.

I can confirm this also, build 15050 does not have that bug in it.

As for 7/31/3, it does sim as a DPS gain, but delaying BT for more than a second or two each CD completely nullifies the dps gain. This has been established for quite some time.

The reason for this is that you are trading an unholy rune once every minute (PT+OtB -> Pt+Pt+OtB) for an additional death rune every minute (CD from 60s -> 30s) so at it's best it is the difference between an unholy rune and a death rune (which is sometimes 0), but if the death rune isn't activated immediately it lowers the value of the tradeoff with every additional second of delay.

Netukka · 12/26/11, 11:02 PM

Im not going to get involved in the argument, but something started bothering me...

Originally Posted by Lamperouqe

The following math is assuming "masterfrost" pulling the same amount of KM OB's out, overvaluing HB. As you already admitted, it doesn't happen in reality.

OB MH hit 19933 + 0.12 * 38581 = 24562
OB OH hit 12595 + 0.12 * 24518 = 15537
Rime and 2p t13 (25447 + 0.18 * 51459) * 0.72 = 24990
OB without KM 24562 + 15537 + 24990 = 65089
1.72 globals used on average, 2 runes consumed at once which helps to keep runes regenerating, 37.2 RP generated on average.

In the rime calculations. You give rime a 72% chance to proc on Obl. Where does this number come from? I thought the 2pc set bonus simply gave you 27% chance to proc 2x HB and 18% chance for normal 1 instead of 45% for one. Is the math wrong or did I miss something? Or did you simply go 27+27+18=72 and call it a day?

Also the way the probabilities are calculated and the damages due to them is wrong. Just take a look at the OB hits for example. The calculations assume a 12% crit chance, but say you had a 100% chance. The calculation would look like 19933 + 1 * 38581 (where does that number come from anyway?) = 58514. That would mean every MH Obl would obviously crit, but it would also deal 3x hit dmg instead of 2x, which is quite clearly false.

Zimeron · 12/26/11, 11:10 PM

Originally Posted by Mhobius

Some things I'd like to share:

- Glyph of Pestilence no longer causes Plague Strike to hit 20% harder, it was fixed some patches ago

- Simulationcraft doesn't take Devotion Aura in consideration, beside it's only ~150 atk power with Bladed Armor, it's still a dps increase

We don't have Glyph of Pestilence affecting Plague Strike and Devotion Aura is properly increasing AP when the DK has Bladed Armor. If you have examples of GOP affecting PS or Devo not increasing AP, please post the profiles on our issue tracker.

Lamperouqe · 12/27/11, 12:16 AM

Originally Posted by gharnef

As for 7/31/3, it does sim as a DPS gain, but delaying BT for more than a second or two each CD completely nullifies the dps gain. This has been established for quite some time.

The sim is already delaying it to use it at the right moment, but that's true for untalented BT as well. The average BT intervals were something like ~34 sec with IBT and ~64 sec without.

It's a gain of 1 PS worth of damage over the course of a minute, including a GCD cost, though. The longer you're unable to find room for BT when it comes off CD, the smaller of an increase it is. Using BT on a rune that's 50% refreshed would halve its value, so using it hastily isn't wise either. Making simple mistakes in the use of BT can lead to it being a decrease, which makes 5/31/5 the safer option. I'll try how IBT plays out in practice today.

Originally Posted by CortDK

As I said several posts back, all that matters is if the benefits of the DPS gain outweigh the loss of the disadvantages. Stacking mastery significantly reduces the number of KM procs, but doing so nets a DPS gain (albeit a relatively small one of a few hundred dps) because the gain outweighs the con, in that particular example.

I wasn't even talking about mastery reducing the amount of KM procs, I was talking about the utilization of KM procs. The standard priority is able to use KM on OB more often. The difference of KM OB and KM FS in damage is not insignificant.

Sigh, I just don't see the minimal theoretical dps gain worth it with the pros and cons we've been debating for a few pages now. I've yet to see it perform better in practice. If in possession of a souldrinker, it's most definitely not worth it at least.

Originally Posted by Netukka

In the rime calculations. You give rime a 72% chance to proc on Obl. Where does this number come from? I thought the 2pc set bonus simply gave you 27% chance to proc 2x HB and 18% chance for normal 1 instead of 45% for one. Is the math wrong or did I miss something? Or did you simply go 27+27+18=72 and call it a day?

Rime still has a 45% chance to proc. What I calculated was how many HB's on average a single OB would produce through rime.

Originally Posted by Netukka

Also the way the probabilities are calculated and the damages due to them is wrong. Just take a look at the OB hits for example. The calculations assume a 12% crit chance, but say you had a 100% chance. The calculation would look like 19933 + 1 * 38581 (where does that number come from anyway?) = 58514. That would mean every MH Obl would obviously crit, but it would also deal 3x hit dmg instead of 2x, which is quite clearly false.

I calculated the crit damage increase without subtracting the hit damage from it, which was my bad. I knew I made some sort of a mistake there, but the same is true for the HB damage in the formula.

CortDK · 12/27/11, 12:27 AM

Originally Posted by minytrooper

1) you say Glyph of Pest, no longer causes PS to hit 20% harder, that could be your "~120" dps difference easily if the sim you using still has the bugged glyph programmed into it

Simc didn't model that particular bug.

gharnef · 12/27/11, 12:27 AM

Originally Posted by Netukka

Im not going to get involved in the argument, but something started bothering me...

In the rime calculations. You give rime a 72% chance to proc on Obl. Where does this number come from? I thought the 2pc set bonus simply gave you 27% chance to proc 2x HB and 18% chance for normal 1 instead of 45% for one. Is the math wrong or did I miss something? Or did you simply go 27+27+18=72 and call it a day?

Also the way the probabilities are calculated and the damages due to them is wrong. Just take a look at the OB hits for example. The calculations assume a 12% crit chance, but say you had a 100% chance. The calculation would look like 19933 + 1 * 38581 (where does that number come from anyway?) = 58514. That would mean every MH Obl would obviously crit, but it would also deal 3x hit dmg instead of 2x, which is quite clearly false.

Yes.

As for the mathematical error it should be changed to:

HIT_DAMAGE + CRIT_CHANCE *(CRIT_DAMAGE - HIT_DAMAGE) = AVERAGE_DAMAGE

You could also express it as:

HIT_DAMAGE + CRIT_CHANCE*(CRIT_MULTIPLIER-1)*(HIT_DAMAGE) = AVERAGE_DAMAGE

To fix the numbers according to my expressions it would be:

OB MH hit 19933 + 0.12 * 18648 = 22171
OB OH hit 12595 + 0.12 * 11923 = 14026
Rime and 2p t13 (25447 + 0.18 * 51459) * 0.72 = 21693
OB without KM 22171 + 14026 + 21693 = 57890

To de-math the explanation: the initial calculation was having crits behave as a crit PLUS a hit - leading to heavy over-evaluation of crits.

On another completely separate note: I was looking at the simC BiS list and it has No'kaled as the BiS mainhand for frost, but with strength - obviously an oversight, HOWEVER, just to check I tested it with agility on it - as it would in game, and it simmed as an increase - I tried again with two No'kaleds and it came as an even greater increase. This is on Build 15050. It obviously wouldn't be valid for Masterfrost as the loss of strength would more greatly impact howling blast than obliterate, but it looks like No'kaled might be the actual BiS for Frost this tier.