Frost DPS | Cataclysm 4.0.6, This Is My Life - Page 28

xbit · 04/01/11, 5:31 PM

Just an update on the question I asked yesterday about spell crit. I did some more testing today where I had 2.75% melee crit from agility, and 0.00% spell crit, this time on a boss dummy. Everything I used (strikes obviously, but also blood plague, DND, death coil, and howling blast) was able to crit. I was not autoattacking, so KM wasn't procced at any time. The only thing that never crit was frost fever, and I left it running for 1000 ticks. Even things that did crit did so far less often than would be expected for a 2.75% crit chance. the OP quotes crit suppression of spells of 3% which obviously can't be true here or none of them would have crit, unless boss dummies don't implement crit suppression (my strikes shouldn't have crit either with a 4.8% suppression). So provided the dummies don't implement any kind of crit suppression (or different suppression applies to different spells...) FF seems to be on spell crit and everything else seems to be on melee crit? I will test letting both diseases run and note any crit chance difference between them.

xbit · 04/01/11, 5:58 PM

Originally Posted by matthewseidl

Looking at KM procs in simc. Its my understanding KM can proc from both weapons at 5 PPM, so for dw frost, I'd expect ~20% of melee hits to generate a KM. But in simc, I see 288 MH + 288 OH attacks (average) with only 48 procs. Even if its only MH which procs, 48 is still fairly low.

For 2h frost, I expected 30% of melee hits to generate KM procs, and the number support that.

288 MH auto attacks != 288 MH non-misses. If you look at the number that didn't miss, the math will work out to be correct. That said, the sim currently only procs KM off of main hand auto attacks, which I believe is incorrect since it was trivial to equip a weapon in my offhand on live, but not in my main hand, and still see procs. I will get that fixed and it should be available whenever 406-19 is released. Thanks for the feedback

Edit: Fixed in svn as of r8322 for those who compile their own.
Edit2: I'm an idiot. Punches can still proc KM. The original implementation was correct.

Steelo · 04/02/11, 9:15 PM

Hey I just speced frost for the first this in the xpac and read up all consider's post. I just have one question that wasn't addressed in the main post. How many mobs need to be present for you to switch from the single target priority and use the AoE priority? I was thinking 3+ mobs or does HB out damage OB on even 2 mobs? Srry if this was already posted somewhere in the responses, I skimmed through them and couldn't find an answer.

Thanks alot

knowname · 04/03/11, 12:17 AM

Originally Posted by xbit

288 MH auto attacks != 288 MH non-misses. If you look at the number that didn't miss, the math will work out to be correct. That said, the sim currently only procs KM off of main hand auto attacks, which I believe is incorrect since it was trivial to equip a weapon in my offhand on live, but not in my main hand, and still see procs. I will get that fixed and it should be available whenever 406-19 is released. Thanks for the feedback

Edit: Fixed in svn as of r8322 for those who compile their own.
Edit2: I'm an idiot. Punches can still proc KM. The original implementation was correct.

KM procs at 5 PPM from main hand(or right fist if thats how you roll) modified by several factors such as haste and hit. Been that way for a long time.

effexor · 04/04/11, 7:55 AM

What is the amount of haste you guys are simming or trying out and having great numbers with 2h? 900 rating or more?

Zenwhoa · 04/05/11, 3:48 AM

Originally Posted by effexor

What is the amount of haste you guys are simming or trying out and having great numbers with 2h? 900 rating or more?

900 is extremely low. From all the tests I've run, haste is the best stat to stack regardless of 1h/2h FP/UP. Aim for at least 1100-1300; you may notice higher numbers from DW with lower haste values. Also, you may get more damage as 2H pre ~1300 haste if you use Frost Presence (and dump RP / switch to UP for lust). It really does depend on a lot of factors though including gear, playstyle, and the nature of the raid encounter.

Try out these action lists instead of the defaults:

1H FP 4.1 priority
actions+=flask,type=titanic_strength
actions+=/food,type=beer_basted_crocolisk
actions+=/presence,choose=frost,if=!buff.bloodlust.react&runic_power<=20
actions+=/presence,choose=unholy,if=buff.bloodlust.react
actions+=/army_of_the_dead
actions+=/snapshot_stats
actions+=/blood_fury,time>=10
actions+=/golemblood_potion,if=!in_combat|buff.bloodlust.react|target.time_to_die<=60
actions+=/auto_attack
actions+=/pillar_of_frost
actions+=/raise_dead,if=buff.heart_of_rage.react&buff.rune_of_the_fallen_crusader.react
actions+=/raise_dead,time>=15
actions+=/outbreak,if=dot.frost_fever.remains<=2|dot.blood_plague.remains<=2
actions+=/howling_blast,if=dot.frost_fever.remains<=2
actions+=/plague_strike,if=dot.blood_plague.remains<=2
actions+=/obliterate,if=frost=2&unholy=2
actions+=/obliterate,if=death=2
actions+=/obliterate,if=buff.killing_machine.react
actions+=/blood_tap,if=buff.killing_machine.react
actions+=/empower_rune_weapon,if=target.time_to_die<=120&buff.killing_machine.react
actions+=/blood_strike,if=blood=2
actions+=/frost_strike,if=runic_power>=90&!buff.bloodlust.react
actions+=/frost_strike,if=runic_power>=95
actions+=/howling_blast,if=buff.rime.react
actions+=/howling_blast,if=(death+unholy)=0&!buff.bloodlust.react
actions+=/obliterate
actions+=/empower_rune_weapon,if=target.time_to_die<=45
actions+=/frost_strike
actions+=/howling_blast
actions+=/blood_tap
actions+=/blood_strike,if=death=0
actions+=/empower_rune_weapon
actions+=/horn_of_winter

2H UP 4.1 priority
actions+=flask,type=titanic_strength
actions+=/food,type=beer_basted_crocolisk
actions+=/presence,choose=unholy
actions+=/army_of_the_dead
actions+=/snapshot_stats
actions+=/blood_fury,time>=10
actions+=/golemblood_potion,if=!in_combat|buff.bloodlust.react|target.time_to_die<=60
actions+=/auto_attack
actions+=/pillar_of_frost
actions+=/raise_dead,if=buff.heart_of_rage.react&buff.rune_of_the_fallen_crusader.react
actions+=/raise_dead,time>=15
actions+=/outbreak,if=dot.frost_fever.remains<=2|dot.blood_plague.remains<=2
actions+=/howling_blast,if=dot.frost_fever.remains<=2
actions+=/plague_strike,if=dot.blood_plague.remains<=2
actions+=/obliterate,if=frost=2&unholy=2
actions+=/obliterate,if=death=2
actions+=/obliterate,if=buff.killing_machine.react
actions+=/blood_tap,if=buff.killing_machine.react
actions+=/empower_rune_weapon,if=target.time_to_die<=120&buff.killing_machine.react
actions+=/blood_strike,if=blood=2
actions+=/frost_strike,if=runic_power>=90&!buff.bloodlust.react
actions+=/frost_strike,if=runic_power>=95
actions+=/howling_blast,if=buff.rime.react
actions+=/howling_blast,if=(death+unholy)=0&!buff.bloodlust.react
actions+=/obliterate
actions+=/empower_rune_weapon,if=target.time_to_die<=45
actions+=/frost_strike
actions+=/howling_blast
actions+=/blood_tap
actions+=/blood_strike,if=death=0
actions+=/empower_rune_weapon
actions+=/horn_of_winter

Make a chardev profile. Try using the frost presence priority if you are running at an item level below ~359, even if you are running as a 2H'er. These work fine for simulations on 4.06 too, although they may improve slightly with slightly different RP on FS numbers depending on patch and whether or not you get the RP talents.

Edit: I kept blood strike in the priority so the list could still be used for live. It shouldn't actually ever cast when running 4.1 simulations.

Kaffe · 04/05/11, 2:47 PM

Originally Posted by effexor

What is the amount of haste you guys are simming or trying out and having great numbers with 2h? 900 rating or more?

I believe ultimately your goal as 2H Frost is to reach around 2000 Haste Rating to enhance rune regeneration. 900 is quite low even for ~346 iLevel gear. You should ideally be able to break 1000 in all blues with enchants and reforging. Remember to gem pure strength unless the socket bonus can cover for the lost strength when using hybrid gems.

knowname · 04/05/11, 4:59 PM

Originally Posted by Zenwhoa

900 is extremely low. From all the tests I've run, haste is the best stat to stack regardless of 1h/2h FP/UP. Aim for at least 1100-1300; you may notice higher numbers from DW with lower haste values. Also, you may get more damage as 2H pre ~1300 haste if you use Frost Presence (and dump RP / switch to UP for lust). It really does depend on a lot of factors though including gear, playstyle, and the nature of the raid encounter.

Try out these action lists instead of the defaults:

1H FP 4.1 priority
actions+=flask,type=titanic_strength
actions+=/food,type=beer_basted_crocolisk
actions+=/presence,choose=frost,if=!buff.bloodlust.react&runic_power<=20
actions+=/presence,choose=unholy,if=buff.bloodlust.react
actions+=/army_of_the_dead
actions+=/snapshot_stats
actions+=/blood_fury,time>=10
actions+=/golemblood_potion,if=!in_combat|buff.bloodlust.react|target.time_to_die<=60
actions+=/auto_attack
actions+=/pillar_of_frost
actions+=/raise_dead,if=buff.heart_of_rage.react&buff.rune_of_the_fallen_crusader.react
actions+=/raise_dead,time>=15
actions+=/outbreak,if=dot.frost_fever.remains<=2|dot.blood_plague.remains<=2
actions+=/howling_blast,if=dot.frost_fever.remains<=2
actions+=/plague_strike,if=dot.blood_plague.remains<=2
actions+=/obliterate,if=frost=2&unholy=2
actions+=/obliterate,if=death=2
actions+=/obliterate,if=buff.killing_machine.react
actions+=/blood_tap,if=buff.killing_machine.react
actions+=/empower_rune_weapon,if=target.time_to_die<=120&buff.killing_machine.react
actions+=/blood_strike,if=blood=2
actions+=/frost_strike,if=runic_power>=90&!buff.bloodlust.react
actions+=/frost_strike,if=runic_power>=95
actions+=/howling_blast,if=buff.rime.react
actions+=/howling_blast,if=(death+unholy)=0&!buff.bloodlust.react
actions+=/obliterate
actions+=/empower_rune_weapon,if=target.time_to_die<=45
actions+=/frost_strike
actions+=/howling_blast
actions+=/blood_tap
actions+=/blood_strike,if=death=0
actions+=/empower_rune_weapon
actions+=/horn_of_winter

2H UP 4.1 priority
actions+=flask,type=titanic_strength
actions+=/food,type=beer_basted_crocolisk
actions+=/presence,choose=unholy
actions+=/army_of_the_dead
actions+=/snapshot_stats
actions+=/blood_fury,time>=10
actions+=/golemblood_potion,if=!in_combat|buff.bloodlust.react|target.time_to_die<=60
actions+=/auto_attack
actions+=/pillar_of_frost
actions+=/raise_dead,if=buff.heart_of_rage.react&buff.rune_of_the_fallen_crusader.react
actions+=/raise_dead,time>=15
actions+=/outbreak,if=dot.frost_fever.remains<=2|dot.blood_plague.remains<=2
actions+=/howling_blast,if=dot.frost_fever.remains<=2
actions+=/plague_strike,if=dot.blood_plague.remains<=2
actions+=/obliterate,if=frost=2&unholy=2
actions+=/obliterate,if=death=2
actions+=/obliterate,if=buff.killing_machine.react
actions+=/blood_tap,if=buff.killing_machine.react
actions+=/empower_rune_weapon,if=target.time_to_die<=120&buff.killing_machine.react
actions+=/blood_strike,if=blood=2
actions+=/frost_strike,if=runic_power>=90&!buff.bloodlust.react
actions+=/frost_strike,if=runic_power>=95
actions+=/howling_blast,if=buff.rime.react
actions+=/howling_blast,if=(death+unholy)=0&!buff.bloodlust.react
actions+=/obliterate
actions+=/empower_rune_weapon,if=target.time_to_die<=45
actions+=/frost_strike
actions+=/howling_blast
actions+=/blood_tap
actions+=/blood_strike,if=death=0
actions+=/empower_rune_weapon
actions+=/horn_of_winter

Make a chardev profile. Try using the frost presence priority if you are running at an item level below ~359, even if you are running as a 2H'er. These work fine for simulations on 4.06 too, although they may improve slightly with slightly different RP on FS numbers depending on patch and whether or not you get the RP talents.

Edit: I kept blood strike in the priority so the list could still be used for live. It shouldn't actually ever cast when running 4.1 simulations.

Lets "pro up" a little gentleman. If your in FP regardless if you are running 2h or dw'ing, haste will never out scale mastery. Your miss interpreting scaling deltas. Once you reach the GCD cap, mastery will scale better than haste. The GCD cap is different for every build. In 4.1 you will simply need a calculator to find out what yours is. Your rune CD will be displayed in the character window with haste already applied. There are going to be several ways to build a frost death knight in 4.1 that can be successful, but I assure you if your going to use FP mastery will still be your life blood. If you want to talk numbers send me a pm.

Zenwhoa · 04/06/11, 5:38 AM

Originally Posted by knowname

Lets "pro up" a little gentleman. If your in FP regardless if you are running 2h or dw'ing, haste will never out scale mastery. Your miss interpreting scaling deltas. Once you reach the GCD cap, mastery will scale better than haste. The GCD cap is different for every build. In 4.1 you will simply need a calculator to find out what yours is. Your rune CD will be displayed in the character window with haste already applied. There are going to be several ways to build a frost death knight in 4.1 that can be successful, but I assure you if your going to use FP mastery will still be your life blood. If you want to talk numbers send me a pm.

I haven't run any simulations on the newest build, but I literally ran hundreds of simulations developing those action lists and trying to figure out the best stats and specs. I've posted a healthy amount of that data several posts up but go ahead and run them yourself - unless you have an insane amount of haste (definitely over 1500), there is almost no chance you are GCD capped in frost presence, especially in 4.1 where we get even more wiggle room with the removal of BS from our rotation. And, haste was very clearly the highest scale factor for both 1H and 2H frost presence specs. I believe it still wins out in 4.1 for most scenarios, I've noticed similar in-game results in raids and on target dummies that supported the simulator (again, some of which I have previously posted).

Crit and mastery are extremely close, but crit generally wins out. Mastery gets a slight bump in 4.1, but it still shouldn't overtake crit for most players. However, I have also posted a lot of speculation on the fact that on every fight but Chimaeron it is possible to have multiple mobs DoT'ed at the same time, which helps push mastery higher and potentially total frost damage above that 45% mark. When 45% of your damage is frost, mastery undeniably becomes better than crit (check a few posts back for my math on this). This seriously depends on playstyle, gear, and what that players specific role might be on an encounter.

I think a source of confusion on this came from a combination of scale factors being extremely close for frost presence specs and a few random previous bugs in simulations. I'm not really sure on that one, but I definitely am coming up with haste being supreme. I've posted a wealth of statistics in my previous replies, so feel free to try and contradict me. Like I said, I welcome it - I'm searching for the real answers just as hard as everyone else here. Zimeron told me that he is implementing my new action list, so there is at least one thing I appear to be correct on. All you are throwing out there is random opinion with absolutely nothing backing it up. Come on, at LEAST post a log of you doing 982351829 DPS while stacking mastery

knowname · 04/06/11, 2:11 PM

Originally Posted by Zenwhoa

I think a source of confusion on this came from a combination of scale factors being extremely close for frost presence specs and a few random previous bugs in simulations. I'm not really sure on that one, but I definitely am coming up with haste being supreme. I've posted a wealth of statistics in my previous replies, so feel free to try and contradict me. Like I said, I welcome it - I'm searching for the real answers just as hard as everyone else here. Zimeron told me that he is implementing my new action list, so there is at least one thing I appear to be correct on. All you are throwing out there is random opinion with absolutely nothing backing it up. Come on, at LEAST post a log of you doing 982351829 DPS while stacking mastery

World of Logs - Real Time Raid Analysis

Nothing special, but you asked for a log. I am not trying to make light of your hard work. I am trying to get to the bottom of things just as much as you are.

xbit · 04/06/11, 2:18 PM

Originally Posted by Zenwhoa

I've posted a wealth of statistics in my previous replies, so feel free to try and contradict me. Like I said, I welcome it - I'm searching for the real answers just as hard as everyone else here. Zimeron told me that he is implementing my new action list, so there is at least one thing I appear to be correct on. All you are throwing out there is random opinion with absolutely nothing backing it up. Come on, at LEAST post a log of you doing 982351829 DPS while stacking mastery

While I do agree with you on haste, since we're using the same sim to base our theory on, in practice it's going to be hard to know which is right based on logs, since the difference between stacking one stat vs another is only about a trade off of 1000 or so of one stat for another stat, which is only a few hundred dps between stacking whatever you think the worst stat is vs stacking the best (between mastery, haste, and crit of course), and dps can easily change by a few hundred just from RNG and raid comp.

gamma_ray · 04/06/11, 7:26 PM

Originally Posted by xbit

While I do agree with you on haste, since we're using the same sim to base our theory on, in practice it's going to be hard to know which is right based on logs, since the difference between stacking one stat vs another is only about a trade off of 1000 or so of one stat for another stat, which is only a few hundred dps between stacking whatever you think the worst stat is vs stacking the best (between mastery, haste, and crit of course), and dps can easily change by a few hundred just from RNG and raid comp.

For simulation purposes of figuring out stat weights for a general consensus, we must take off all raid buffs that we ourselves do not provide. Different stats will scale differently with buffs depending on your raid comp, that much is obvious. When running sims to post here for information's sake, we have to take out as many variables as possible and zero in on the stats themselves.

If you are creating stat weights for your specific character then yes, you should put in buffs that you receive from other members of your raid. Having a rogue with Honor among thieves for instance will give you that 5% crit specific to your raid. That has no bearing here when coming up with theorycrafting DK's.

I just feel like this needs to be said in order to make sure we are taking out as many variables as possible. Theorycrafting is theory only, we cannot try and take into account specific fight mechanics, doing so will only create more questions when we still have yet to solve basic ones in front of us.

xbit · 04/06/11, 7:43 PM

Originally Posted by gamma_ray

I just feel like this needs to be said in order to make sure we are taking out as many variables as possible. Theorycrafting is theory only, we cannot try and take into account specific fight mechanics, doing so will only create more questions when we still have yet to solve basic ones in front of us.

It was my impression that when specifics are involved, the consensus was to include all raid buffs/boss debuffs by default. Does it really matter what the weights are only self buffed when the vast majority of end game raiders (the target audience of this forum) are usually going to have the majority if not all of those buffs? My point about logs not being useful was that the dps variance of frost across different reforging is probably below the noise level of actual raid dps variation. This is the same reason that doing dps testing on target dummies is shunned. World of Logs is useful for comparing against simc report output (once you've verified buffs and debuffs match), but showing one WoL parse with mastery reforging doing more dps than another WoL by the same DK haste reforged doesn't prove anything.

Bobfred21 · 04/06/11, 7:45 PM

Originally Posted by gamma_ray

For simulation purposes of figuring out stat weights for a general consensus, we must take off all raid buffs that we ourselves do not provide. Different stats will scale differently with buffs depending on your raid comp, that much is obvious. When running sims to post here for information's sake, we have to take out as many variables as possible and zero in on the stats themselves.

If you are creating stat weights for your specific character then yes, you should put in buffs that you receive from other members of your raid. Having a rogue with Honor among thieves for instance will give you that 5% crit specific to your raid. That has no bearing here when coming up with theorycrafting DK's.

I just feel like this needs to be said in order to make sure we are taking out as many variables as possible. Theorycrafting is theory only, we cannot try and take into account specific fight mechanics, doing so will only create more questions when we still have yet to solve basic ones in front of us.

Any serious raiding DK should be trying to get as close to every raid buff as possible - these days it's not terribly hard.
That having been said, if I'm in a ten man guild that's missing mark of the wild and 5% crit looking at a set of stat weights generated using NO buffs will be less accurate than looking at stat weights where those two buffs were taken into account as well as all other buffs.

It's much more valuable to figure out what are stat weights are raid buffed, because when we fight a raid boss we're going to be doing our best to have as many of those available as possible and there's little use knowing which stat weights are more valuable on target dummies. Likely, it doesn't matter very much and the order will come out the same anyway - but sims should be trying to factor in as many variables as possible to a reasonable extent - and it's far more reasonable to assume we'll have every raid buff as opposed to having none of them.

Now, Xbit is still correct because of RNG and because specific fights have many things other than what buffs you have going on that could affect numbers.

EDIT: Beat to the punch.

Emoette · 04/07/11, 12:01 AM

Originally Posted by knowname

World of Logs - Real Time Raid Analysis

Nothing special, but you asked for a log. I am not trying to make light of your hard work. I am trying to get to the bottom of things just as much as you are.

Argoloth is too short of a fight for real analysis.