I was afraid it'd be a problem that cost $ to fix.

It did seem strange that video lag would enter the equation as it seemed like the cast buffer being sent to server would compensate, but I was worried that my computer performance was indeed the culprit.

Thank you very much for the reply.

Givent the recent findings about Resistance Mechanics in WotLK it picked my interest again to revisit the level-base partial resistance behavior. I haven't seen many data collections done in WotLK regarding this. The tests before the expansion concluded that the level-based resistances are average 2% per level difference, but with the changes to the resistance mechanics in WotLK it's worth reexamining the issue.

WWS of boss parses are largely hard to interpret since the resisted amount reported is not modified by damage amplifying effects. With this in mind I collected a reasonable sample of data on boss target dummy with no debuffs and known crit bonus multiplier so it was easy to parse the combat log and extract the resists.

We know that partial resists appear in 10% increments. The recent findings have concluded that partial resists follow a triangular distribution around average resist

P(x) = 0.5 - 2.5*|x - AR|

where x is the possible resist and AR the average resist.

This formula however cannot be used for AR < 0.1 since some of the distribution moves over into negative values for x. To correct for this while preserving the average observe the following:

E(X) = -0.1 * P(-0.1) + 0 * P(0) + 0.1 * P(0.1) + 0.2 * P(0.2) = 0.1 * (P(0.1) - P(-0.1)) + 0.2 * P(0.2)

We see that the average will be preserved if we use the following distribution:

P'(0) = P(0) + 2 * P(-0.1)
P'(0.1) = P(0.1) - P(-0.1)
P'(x) = P(x) for x >= 0.2

It is not obvious that Blizzard must have used this method, but it seems the most natural and as will be seen later is supported by data.

If we simplify this for 0 <= AR <= 0.1 we get the following distribution:

P(0) = 1 - 7.5 * AR
P(0.1) = 5 * AR
P(0.2) = 2.5 * AR

The outcomes of the experiment can be interpreted as samples from multinomial distribution. Each possible resist itself follows a binomial distribution, but they are not independent. We know that fusing several cells of a multinomial distribution also results in a binomial distribution. We will use this under assumption that the resists follow the above distribution to construct confidence intervals on AR.

The data collected is as follows:

total 10964 non-misses

6096 with 0% resisted
3361 with 10% resisted
1507 with 20% resisted

Fusing the 10% and 20% cells we have 4868 events from 10964 samples. Since this is a binomial distribution we can compute confidence intervals for the proportion. Using 99% confidence interval we get 0.4318 <= 7.5 * AR <= 0.4563 or 0.05757 <= AR <= 0.06084. (note that a more accurate way might be to compute simultaneous confidence intervals on the multinomial distribution)

This seems in line with our previous assumption of 2% per level difference so I think it is safe to assume that is still the case. The resist table for 6% average resist given the above distribution should be

55% with 0% resisted,
30% with 10% resisted,
15% with 20% resisted.

Compare this with

55.60% with 0% resisted,
30.65% with 10% resisted,
13.74% with 20% resisted

in our sample. So the assumption on the shape of probability distribution seems to be quite reasonable.

In Resistance Mechanics in WotLK, a comment says the resisted portion is not affected by damage multipliers(I am not sure it is buff/debuff on the caster or target), I think it could modify the 6% when there are curse, aura or AP.

No, if you check the combat log you'll see that the displayed resisted amound is unaffected by any damage multipliers. Note this is just something you have to take into account when parsing the logs, the end result how much damage you lose is the same no matter how it is displayed in the log since resistance is just another multiplier (you really lost the listed value in log multiplied by the damage multipliers, including crit multiplier).

The specific post is this: Resistance Mechanics in WotLK, he tried to explain why there is a 6 % resistance (28000 damage, 1800 resisted) in combat log. In his formula the damage multiplier C includes all buff and debuff on the target.

I think Kavan's data is the first set of data casted on a boss, all data collected in Resistance Mechanics in WotLK is casted on players.

Hey guys i was hoping to get some help here. I have been reading forums on DPS rotations with FFb spec, and can not figure out what i am doing wrong. I have tested on dummies for hours on th ptr and live servers. I do the normal rotation of 1 scorch, LB, FFB till hotstreak procs. I am only putting out about 3.4k dps with all my CDs. Once in awhile depending on the fight i may go to 4k dps but rarly (with heroisims and CDs). Im for some reason can not keep up with the DPS in my guild and am willing to take any advice that i could be doing wrong.

heres my armory, any advice please. The World of Warcraft Armory

Thanks.

Thanks for the help im also at work Ya last night i was in Uldar and i have mookin in grp and spriest. I was at 34.51 crit and over 2k spell damage. I wish i had a wws report for you. But even in Naxx i was behind in DPS. I work on getting more hit gear. I mean i wasnt even pushing the top 10 in DPS last night. im actually not going to be raidinf much longer sadly because my dps is to low, compared to everyone else. When im pushing 4k dps everyone else is pushing 8k. When i get a wws ill have to link it.

I find it really hard to believe that you pull 3.4k on target dummies but yet only push 4k in full raid situation.

Things that may cause your target dummy DPS to look higher: LB exploding on the targets on the side, not enough DPS time and testing done.

Based on your 34.51% crit and 2k sp, was that including all the raid buffs?

Sorry i didnt say i pulled it on a dummy. I did that damage in raid. Then when Hero and all my CDs i did around 4k dps.
Back in Naxx raids the other mages where doing like 5k dps with there CDs. I must be doing something wrong. I did see a mage do 4k dps on a dummy, do you guys think i would pull more dps off as Arcane? I mean with my hit rating and all.

WWS or shut up.

Sorry that comes off harsh. But we can't accurate analyze what your doing from your 'description' of what your doing.

Log your performance. Ask people to analyze the logs. Anything else will only catch obvious and very superficial problems.

I'm going to assume that you are using some kind of mod to track living bomb, scorch, and hot streak procs. If not, you really need to get something to do that for you. Rescorching too early / Lowering your living bomb uptime / Missing hot streaks are all steps towards wrecking your dps.

Are you queuing your spells? That means while your FFB is casting you're spamming the key for your next spell. If you are spending ANY amount of time NOT casting a spell it will have a severe impact on your dps.

Here's how I typically cast:
I am now spamming my pyroblast button while the third FFB is casting. Note that I don't move or cancel my third FFB because that would mean the time I spent attempting to cast the third FFB would be wasted - that's a DPS loss. A better way to think about it might be that I am always thinking one spell ahead - the current spell I am casting is a given - I'm more worried about what spell I should cast next rather than think about what spell I should currently be casting.

Also, to avoid queuing a non-HS pyroblast, AFTER my third FFB leaves my hand, I actually start moving back and forth here until I see my pyroblast cast, then I stand still and start mashing my FFB key again - not necessary but I have never ever had to cancel a casted pyroblast, so either it's working or I'm just playing good . I actually have gotten so used to it that I can plan minor positional adjustments around hotstreak procs, but I digress.

Also, I've heard other mages erroneously state that it is beneficial to save hotstreak procs til after a GCD to avoid ignite munching - this is false. You should cast a hot streak pyroblast whenever you get the proc - just queue it after the spell you are currently casting. The only time you should risk losing a hotstreak proc is to refresh a living bomb or to rescorch if it is about to drop off.

Aside from that, its a matter of getting your reaction time down so that you can catch hot streak procs very quickly - I often refresh a living bomb right as it goes off - if it crits, there's a chance that it could generate a hot streak proc - it is difficult but not impossible to catch the hot streak in the global cooldown caused by the living bomb refresh.

I can't see your gear from your armory link. I've searched the armory and there is no LVL 80 Mage by the name of Sockett on US or EU servers. Of course, this might just be an armory bug.

Use Lightweave Embroidery on your cloak. And are those your PVE boots? Armory is showing PvP boots right now. If you don't have a pair, get the [Sandals of Crimson Fury] from wyrmrest accord rep.

And also get to the hit cap.

So after 3 nights in Ulduar I'm wondering what other mages and even other casters are seeing as far as relative damage dealing with the 3.1 changes. Pre 3.1 I was nealy always top 3 in boss fights, and our top 5 were generally a mix of a rogue, 2 hunters, myself and a lock, with other our raiding mage and an enhance sometimes popping up and a mix of melee and casters from there down. So far just looking at recount (WWS is choking on logs for Ulduar which cross midnight for me) pretty much every fight with the exception of an amazingly lucky RNG run I'm seeing pretty much every fight sort out in this order: melee, hunters and then casters. I'm guessing this is impart due to the 5% crit debuff nerf and was wondering if anyone else is seeing this sort of stratification by dps type.

I've noticed quite a bit of disparity depending on the fight. On Destructor, I'm consistently #1 or #2 (neck-to-neck with a death knight) by a very large margin from the rest of the group as an arcane mage. Controlling mana is a bit of an issue, but after a few attempts I got the hang of it and am able to burst on almost every heart phase (hitting ~6000 dps, +-200). In almost every other boss fight I've done so far, the hunters rock the charts, followed closely by the DK and a ret paladin.

Very strong results as FFB. EJ switched to wow meter online since wws is unable to work with ulduar for what its worth.

FYI WowArmory now displays dualspecs (as well as glyphs).

I've had really good results with FM-FFB as well, to the point where I am threat capped at times in a bunch of Ulduar fights (after MI wears off). Invis works half the time, the other half getting interrupted by AoE or random add aggro. The threat issue does make me miss Arcane's -40%, but that's still much less of an issue than Arcane's reliance on full Evos to achieve only sub-FFB dps.

That said, if they ever do end up fixing the FM-swap trick, switching to Fire/ttw (for similar dps) shouldn't be bad at all. With Ulduar BiS gear, or what we know so far, Rawr shows Fire/ttw might not even have to Evo at all, for up to a 9min 25man fight (not accounting for AoE spam).

My Armory and talents : The World of Warcraft Armory

For rotation , currently I am using :

-living bomb
-scorch

-fireblast
-fireball
-fireball
-fireblast

moves like this, refreshing scorch and pyro if hotstreak procs.

I am wondering, if I should not use fireblast at all. It makes me loose some time to cast fireball.

Also, curious if hot streak can proc next to another, so I can throw 2x pyro one after another ? or the 2nd one replaces 1st hotstreak ?

Thanks in advance.

I haven't taken a look at your armory yet, but a couple things immediately jump out:

Fireblast should only be used in two cases:
1) While moving, when you can't cast Living Bomb or Hot Streak Pyro.
2) If the target will die before your fireball casts

Any other time, and it is a DPS Loss.

Hot Streak cannot be procced by Pyroblast, but due to spell travel time, it is possible (and not horribly uncommon) to get 2 hot streak pyros in a row. Check in the FFB thread for lots and lots (and lots) of information on when this happens.

I really think the evocation issue has been blown out of proportion. Maybe it's different for the fights I haven't done or hard modes, but I've done up through Mimiron and only a couple of times had problems with evocation. I'd still like them to change the spell but I don't think arcane is suddenly a dead spec because of it.

At any rate thus far I've still been pretty competitive in my guild enjoying my -40% threat, great burst/aoe damage

And 100% pushback reduction on Mimiron is really helpful as well.

But that's just my take.

Well, both Fire and Frostfire specs have 100% pushback reduction too (with concentration aura).
Still, as far as I understand, Arcane scales a bit worse than Fire/Frostfire, this is why it is considered a 'worse' spec for 3.1