Has anybody recently tested Arcane Missile JoW proc rate?
Testing done here provides very strong evidence that JoW is a PPM system for both weapon attacks and spells. The Fireball tests were particularly revealing.
Due to information in that thread SimulationCraft was changed a while back to use 15PPM for both attacks and spells. However, channeled spells that trigger "spell hits" such as Mind Flay and Arcane Missiles were not tested. I use the "tick time" for each missile to determine the chance of proc'ing JoW given a PPM system.
This has had the effect of making the mana returns from JoW very much on par with BoW and Mana Spring Totem.
Can anyone point to some rigorous AM testing? How exactly does Rawr handle it? (or is it specific to each module?)
I don't think any such data exists for AM. Rawr currently uses straight proc chance, not PPM.
For better comparisons I think it would help if we come up with a test character setup. I've tried looking at the simcraft profiles you're using, but those gear links don't seem to work for me. If we can come up with a list of gear, buffs and fight parameters it'll be easier to compare both side by side to determine where things differ (if they differ at all).
The most interesting thing from this discussion is the Clearcasting / Arcane Potency bug. This is currently not modeled and I don't think we ever came to a conclusion how exactly it works other than it is bugged. Do we know if it was fixed in 3.1 and in what way if it was?
For 4T8 I will only say that unless the proc chance is above 20% it will not have any significant effect on optimal cycle selection.
This was recently added to the patch notes for 3.1:
* Arcane Potency: Fixed a bug where sometimes the benefit was consumed before altering the critical strike chance of the mage. In addition, will no longer benefit the critical strike chance of Molten Armor’s damage.
That might mean that it works correctly with Arcane Missiles now.
Originally Posted by Crowl
If you have to control a robot dinosaur that fires lazers and there's a time when you shouldn't be shooting those lazers then the encounter is clearly flawed beyond hope of fixing.
I don't think any such data exists for AM. Rawr currently uses straight proc chance, not PPM.
For better comparisons I think it would help if we come up with a test character setup. I've tried looking at the simcraft profiles you're using, but those gear links don't seem to work for me. If we can come up with a list of gear, buffs and fight parameters it'll be easier to compare both side by side to determine where things differ (if they differ at all).
Our JoW difference could certainly explain much of my gap. Perhaps we can get a conclusive answer on AM/Jow soon.
I only have access to a PC for a small slice of the day (three children, one computer)..... most of my work is done on my laptop. But I'll start making time for this in the evenings so I can run Rawr regularly. Let's use whatever profile you think best. Rawr.Mage is still the benchmark as far as I'm concerned.
So I downloaded Rawr's latest beta release, 2.2.0b6 released on April 02, uploaded my character from the armory and enabled the necessary buffs/debuffs and turned on 3.1 mod. I am getting similar dps to what I was getting with the older version, 2.2.0b5, however all of a sudden now it is suggesting AB3Abar3MBAM (AB-AB-AB-Abar on MB proc do AM after ABx3) over AB3AM for a typical 5 minute fight. I am not aware of any 3.1 changes that directly or indirectly buffed Abar over AM so I couldn't understand why the switch back to Abar from AM.
So I downloaded Rawr's latest beta release, 2.2.0b6 released on April 02, uploaded my character from the armory and enabled the necessary buffs/debuffs and turned on 3.1 mod. I am getting similar dps to what I was getting with the older version, 2.2.0b5, however all of a sudden now it is suggesting AB3Abar3MBAM (AB-AB-AB-Abar on MB proc do AM after ABx3) over AB3AM for a typical 5 minute fight. I am not aware of any 3.1 changes that directly or indirectly buffed Abar over AM so I couldn't understand why the switch back to Abar from AM.
Abarr has technically always been better than AM. Rawr usually skipped over Abarr in favor of AM for higher mana efficiency which it then translated into a higher uptime for ABSpam3MBAM. I honestly never found it to be worth following this pattern, and checking in on the top dps arcane mages I found most of them opted to use ABarr in favor of AM.
The problem with Rawr's algorithm is that it uses an average for the mana efficiency of the different cycles, meaning if you get unlucky with your ABSpam3MBAM cycle and end up having to cast 8 ABs, because the average mana consumption of the cycle is far less than the mana consumption of ABSpamming at 3 stacks. Maybe Kavan can explain it better, and I apologize ahead of time if I got that wrong. I'm just going off my own observations and testing in game.
Maybe my Xperl is showing it wrong, but on Live the ooFSR timer resets with each missle fired. Haven't tested mana regen numbers, though.
The OOFSR is related to clearcasting AM, but it's hardly even noticable now with the amount of haste in most arcane sets. You only got 1 tick unless you started getting back to back CC procs.
Well surely one of the advantages of using ABx3AM is that you have an extra chance to proc Missile Barrage, the highest Arcane dps option. Always using AM allows you to pick up some extra Missile Barrage procs that using ABar would miss.
Well surely one of the advantages of using ABx3AM is that you have an extra chance to proc Missile Barrage, the highest Arcane dps option. Always using AM allows you to pick up some extra Missile Barrage procs that using ABar would miss.
How so? Missile Barrage procs on AB and Abar not AM. Using ABx3 and then Abar gives you 4 chances to proc Barrage. The advantage with ABx3AM is less overwrites. If your using Abar your gonna use it even if Barrage procs on your 3rd AB because you cant stop to wait and see if it procs.
This means you have the ABar and 3 more ABs before you use Barrage which has a high chance of overwriting and wasting a Barrage proc. Always finishing with AM, besides the more efficient aspect, means you have a far smaller chance of overwriting and wasting a Barrage.
Abarr has technically always been better than AM. Rawr usually skipped over Abarr in favor of AM for higher mana efficiency which it then translated into a higher uptime for ABSpam3MBAM. I honestly never found it to be worth following this pattern, and checking in on the top dps arcane mages I found most of them opted to use ABarr in favor of AM.
The problem with Rawr's algorithm is that it uses an average for the mana efficiency of the different cycles, meaning if you get unlucky with your ABSpam3MBAM cycle and end up having to cast 8 ABs, because the average mana consumption of the cycle is far less than the mana consumption of ABSpamming at 3 stacks. Maybe Kavan can explain it better, and I apologize ahead of time if I got that wrong. I'm just going off my own observations and testing in game.
The way I understand it is after the Abarr coefficent change and the Glyph of Arcane Missiles change, it became a better option to use only ABx3AM all the time with the exception of cooldowns rolling in which case you would use ABSpam3MBAM. ABx3[Abarr or Mbam] is still a higher DPS rotation but the mana cost is much higher, and the reason ABx3AM is used is for the mana efficiency which later allows you to use that ABSpam3MBAM during cooldowns. If you were to use ABx3[Abarr or Mbam] then you would have to continue that cycle through your cooldowns and would end up at roughly the same DPS, but with more mana problems.
Originally Posted by Dochas
How so? Missile Barrage procs on AB and Abar not AM. Using ABx3 and then Abar gives you 4 chances to proc Barrage. The advantage with ABx3AM is less overwrites. If your using Abar your gonna use it even if Barrage procs on your 3rd AB because you cant stop to wait and see if it procs.
This means you have the ABar and 3 more ABs before you use Barrage which has a high chance of overwriting and wasting a Barrage proc. Always finishing with AM, besides the more efficient aspect, means you have a far smaller chance of overwriting and wasting a Barrage.
The poster above you is most likely referring to the T8 4pc set bonus which gives you a chance to retain your Brain Freeze, Hot Streak, or Missile Barrage. The set bonus though is not seen as a very good DPS boost for Arcane Mages.
The way I understand it is after the Abarr coefficent change and the Glyph of Arcane Missiles change, it became a better option to use only ABx3AM all the time with the exception of cooldowns rolling in which case you would use ABSpam3MBAM. ABx3[Abarr or Mbam] is still a higher DPS rotation but the mana cost is much higher, and the reason ABx3AM is used is for the mana efficiency which later allows you to use that ABSpam3MBAM during cooldowns. If you were to use ABx3[Abarr or Mbam] then you would have to continue that cycle through your cooldowns and would end up at roughly the same DPS, but with more mana problems.
I don't think it's that cut and dry. Before ABarr was nerfed, the rotation was AB3[Abarr or Mbarr] and we didn't have mana problems. The change that was giving us mana problems early on was the assumption that we should use cooldowns immediately and then trying to use a high burn cycle through all of that, which simply wasn't workable without innervates. Since then we've moved towards a 2 CD cycle, the first being AP and any CD's that are shorter than AP (ie hyperspeed accelerators), and the 2nd being everything (IV, AP, mana gem, etc), and alternating between the two.
Now the key is, when Rawr tells (or was telling) you to use AB3AM, you should note that it is often telling you to use ABSpam3MBAM during non-AP portions of the fight, and this is where I disagree with the theory behind it. The benefit of using AB3AM in order to gain uptime of ABSpam3MBAM (during non-AP portions of the fight) is relatively small compared to just using AB3[ABarr or Mbarr] throughout that whole period, and the former is actually more risky to mana due to the RNG on MB procs. In theory, when you can use an average mana conumption on ABSpam3MBAM it works well, but in reality bad luck will ruin this plan.
AB3-AM gives 3 chances for AM procs. AB3-[MBarr]-ABarr gives 4 chances for MBarr proc, however after a proc you only have 2 chances to re-proc while restacking AB since you can't react on the 3rd one if it did proc (if AB2 didn't proc MBarr and AB3 did, you've already casted the ABarr by the time you noticed it).
Overall AB3-[MBarr]-ABarr is slightly more DPS than AB3-AM, but costs a bit more mana. When AP is not available, upgrading to the ABarr rotation is the best way to spend additional mana to do more damage, however doing ABspam[3MBarr] rotation during AP (or at least AP+some spellpower cooldown) is a better way to spend your additional mana. Basically you need to figure out how much total mana (including everything that replenishes it) you have left to use until the end of the fight and how much spare you'd have if you just stuck to AB3-AM rotation - and then if you realize there will be leftovers then it should be spent on AB3[3MBarr] during AP+stuff. If that's not possible, it should be spent on either AB3[MBarr] during just AP or on AB3-[MBarr]-ABarr (also preferably during AP). Check rawr for more details... But if someone's telling you that you should just use the ABarr rotation as your main rotation he's misleading you - you're not going to lose a lot at all but it's simply suboptimal.
Now the key is, when Rawr tells (or was telling) you to use AB3AM, you should note that it is often telling you to use ABSpam3MBAM during non-AP portions of the fight, and this is where I disagree with the theory behind it. The benefit of using AB3AM in order to gain uptime of ABSpam3MBAM (during non-AP portions of the fight) is relatively small compared to just using AB3[ABarr or Mbarr] throughout that whole period, and the former is actually more risky to mana due to the RNG on MB procs. In theory, when you can use an average mana conumption on ABSpam3MBAM it works well, but in reality bad luck will ruin this plan.
It's actually the other way around: the (dps) benefit of AB3Abarr3MBAM is relatively small compared to just using AB3AM. On average it's 1% more dps, but on average it's 22% more mana consumption. Having the occassional bad streak doesn't make this untrue, and moreover this can go either way, i.e. AB3AM can be higher dps than AB3Abarr3MBAM almost as often as it can can be less.
Getting a bad streak on ABSpam I think is rather unrelated to your standard rotation. Statistically you will go as far as AB6 25% of the time, so it might well feel that you are having bad streaks quite often. Rawr using averages doesn't make ABspam the wrong advice, you can still break the stack with ABarr at 5, 6, 7 ABs if you judge that your mana won't hold out until your next evocate. Just remember that statistically for every 1 bad streak there will be 2 "good streaks" and you don't want to cut off your nose to spite your face.
Maybe my Xperl is showing it wrong, but on Live the ooFSR timer resets with each missle fired. Haven't tested mana regen numbers, though.
The FSR starts when mana is spent for the most part. However, while you are maintaining a channeled spell (like, say, Blizzard or AM), you don't go OOFSR until after the channeled spell is finished. This includes clearcasted channeled spells. The only channeled spell that lets you go OOFSR while channeling is Evocation.
So, if you clearcast AM and then cast AB after that, at most you get full regen for the duration it takes to cast the AB. While you are channeling AM, you do not get OOFSR regen ever.
Is: "AB AB AB ([mbarr] or abarr)" the same thing as what rawr is telling me: AB3ABar3C?
Am i reading the tooltip wrong? The way i understand the rawr tooltip is if MBarr procs, cast MBAM on 3 stacks AB followed by an ABarr? The reason why im confused is that the general consensus is not to cast ABarr on zero stacks?
Is: "AB AB AB ([mbarr] or abarr)" the same thing as what rawr is telling me: AB3ABar3C?
Am i reading the tooltip wrong? The way i understand the rawr tooltip is if MBarr procs, cast MBAM on 3 stacks AB followed by an ABarr? The reason why im confused is that the general consensus is not to cast ABarr on zero stacks?
AB3ABar3C would be "AB AB AB [mbarr] abarr" in notation used by Manly. It is not clear to me why this rotation would be selected. Which Rawr version are you using? Any unusual fight parameters?
Was using 2.1.9. Just downloaded the new beta and the rotation changed to the normal one.
Another question I have is about the latency setting in rawr. From the rawr thread on this forum I gather that I'm not supposed to enter my in-game latency but rather the effective casting latency that is left over after trying to counter my in-game latency with Quartz?
Whehter or not the ABarr rotation gets picked or not really depends what your other options are for spending that extra mana, which depends on spellpower cooldown availability which greatly depend on gear and fight duration, as well as depending on whether or not with your specific gear AB spamming with just AP is a better or worse way to spend mana than the ABarr rotation.
But yes, AB3ABarr3c is AB AB AB [ABarr or MBarr], ever since the coefficient was nerfed it had been a bad idea to ever cast ABarr at 0 stacks unless you're moving.
Another question I have is about the latency setting in rawr. From the rawr thread on this forum I gather that I'm not supposed to enter my in-game latency but rather the effective casting latency that is left over after trying to counter my in-game latency with Quartz?
Yea. In most cases it's ok if you leave that at default. If you want to set things up as accurate as you can for your situation then what you want is look at a long stretch of chain casting, preferably a single spell, in combat log. Compute how long it should theoretically take you to cast those spells given your haste and how long it actually took you. Divide the difference by number of spells cast, that should be the value that matches the meaning of latency in Rawr.
Ok, I've skimmed through several of the mage threads today looking to see if there is a complete list of spells that trigger the effect for TtW.
I'm trying to set up one of my addons to monitor the target for any of the specific debuffs that trigger it, because I have a few tanks that I run/pug with and I want to harrass them if they're not keeping it up.
So far, this is all I've found:
Torment the Weak
On snarable targets, every snare or attack speed slow triggers Torment on that mob. Hamstring, Frostbolt, everything.
On unsnarable targets like raid bosses, only the attack speed slows Frost Fever, Thunder Clap, Judgements of the Just, Infected Wounds and the mage ability Slow triggers it.
In raids, Torment should be triggered by default unless you face immunities, untanked mobs or an extremely unusual and disadvantageous raid composition.
So, I'd like to be able to list specifically what we know causes it. Has someone posted this somewhere already? 'Cuz I didn't find it.
Edit: Ugh... There's no visible debuff that let's us know for sure that a Paladin has properly slowed the target... but you can check their talents (inspect) to see if they have Judgements of the Just.
In raids, Torment should be triggered by default unless you face immunities, untanked mobs or an extremely unusual and disadvantageous raid composition.
Technically, on aoe tanked mobs, the only "default" slow will be thunderclap and dk frost fever spreading via pestilence. Torment only works on single target spells so shouldn't be an issue except in wierd circumstances.
Torment the Weak
On snarable targets, every snare or attack speed slow triggers Torment on that mob. Hamstring, Frostbolt, everything.
On unsnarable targets like raid bosses, only the attack speed slows Frost Fever, Thunder Clap, Judgements of the Just, Infected Wounds and the mage ability Slow triggers it.
In raids, Torment should be triggered by default unless you face immunities, untanked mobs or an extremely unusual and disadvantageous raid composition.
04/02/09, 6:26 PM
Similar Threads
