Sweet Informational Thread - Page 42

Max Rebo · 07/02/08, 1:02 PM

Originally Posted by Wizeowel

Yes, indeed there is more going on: you always have a 1% chance to miss.

Missing 1 out of 10 frostbolts on a level 72 mob and coming to EJ to post that Blizzard fixed the ghost hit isn't really a shining example of exhaustive research.

Thanks for not reading my entire message and actually trying to understand what I wrote.
I know there is always a 1% chance to miss.
I did read other sites, but I still had doubts, so I came here.
I never made a definitive claim that Blizzard fixed ghost hit. I don't know where you read otherwise.
Since when is exhaustive research required before asking a question here? This is part of my research.
I hate being argumentative and that's not what I came here for, but really... you didn't help much.

------------------------------

EDIT: The first few replies to my question(s) have been the most helpful.
I suppose that I'm taking some of the WWS reports too literally. From one of the replies on the previous page, it seems resist % includes phases where bosses are immune, thus skewing the final result.

Since I haven't been using WWS for too long, let me ask this:

In the WWS parse that Zinaida linked, the FB miss rate is 5.8%, but the actual number of misses displayed is zero. Any idea if this is a glitch or some shortcoming of the program?

I'm just trying to understand all the numbers...

Zinaida · 07/02/08, 1:35 PM

The "miss" column refers to melee. The "resist" column refers to spells. Higher Nethermancer Zerevor will also sometimes recieve an immunity buff which causes spell hits to register in the "other miss" column, and these are included in the overall miss rate calculation.

Wizeowel · 07/02/08, 1:59 PM

Originally Posted by Max Rebo

it seems resist % includes phases where bosses are immune, thus skewing the final result.

If you check your actual combat log you'll see that your spell was just a regular miss, not any kind of immunity. The result is not skewed, you missed 1 out of 15 casts, that's 6.7%

Max Rebo · 07/02/08, 2:19 PM

@ Zinaida / Wizeowel:

Thanks for clarifying. As I mentioned, I'm still getting used to sifting through the WWS reports.
I've never really delved too deep into any sort of combat analysis until recently. I think I understand things a bit better now.

Ozmethod · 07/02/08, 4:22 PM

Currently using the s1 necklace of Sub, so obviously I have an upgrade to do. Did some quick messing around with Rawr and another mage DPS calc, and apparantly the SSO necklace is currently the best in the game. Surely this cant be true? Rawr doesnt have the new Gaurdian necklaces, but they arent that huge of a step above the Vindicator. So, is the SSO necklace really the "best" in the game?

Also, if it really is the best option pre-sunwell, is it better to be scryer or aldor?

Gofa · 07/02/08, 7:01 PM

The new pvp neck is really good and it should be better than the sso neck. Aldor is much better than scryer for the sso necklace.

solbergb · 07/02/08, 8:33 PM

Guardian Pedant of Subjigation:
damage 28 haste 24 resil 18 35 sta 18 intellect, yellow socket, +3 sta socket bonus

It's better than Domination, as they've put the "haste" itemization only partly
into spell damage, the rest into more stam and int. Dominations a bit better if
for some reason you're haste capped. It's a level 154 item which uses its itemization
well for a mage.

Shattered Sun Pedant of Acumen is a level 115 item with this itemization.
damage 37 19 sta 18 int and a proc that varies based on your faction

Without the proc, Subjigation is clearly better - just socket a +12 spell damage gem in
and you're better in every stat provided by Acumen. You don't need RAWR for that.

The thing is that the proc is pretty significant but also gear/spec dependent in how
much it adds to the itemization. RAWR has almost certainly done the math better
than I can, so what it says is probably right for you but I'd check again once
the Guardian stuff makes it into RAWR. It's unlikely the scryer version is better,
the Aldor version might be.

In my case I went with the Guardian because I earn Honor points a lot faster than
I earn SSO and faction rep. I had an easier time with the decision though, as
my character is Scryer.

Pre-Sunwell, Aldor is better hands down if you're exalted. Their shoulder enchants
are obviously superior and their necklace proc is better. Scryer mainly has a nice
+hit trinket at revered that is nice when you are transitioning from blues to
purples, but it isn't really relevant when you're transitioning into the top tier
content. Usual caveats for needing a particular recipie or some such apply.

Pintofbrew · 07/03/08, 8:16 AM

Originally Posted by Max Rebo

At the risk of being redundant, as I'm sure this topic has been covered ad nauseum, I have a question/concern about the hit cap for frost with 3/3 EP. I realize that the first page of this topic mentions the extra 1% per point in EP for frostbolt, thus making the effective hit cap for FB 126, as is mentioned on many other sites, but my recent observations are telling me different.

One example I can think of is a recent Shattered Halls run. During the first half of the run, my spell hit was about 135 or so. For the first boss (Nethekurse), my WWS report showed 6.7% FB resist. Am I reading this properly, or is something wrong with my gear? During the 2nd and 3rd boss fights, I swapped a few pieces of gear to reach ~166 spell hit, and WWS shows zero FB resists. It's quite possible that the first boss has some sort of magic resist, so this example may be poor. I'm not sure...

Example WWS: WWS Loading...

Another example that comes to mind is a recent full Kara run in which there was another frost mage with more or less equivalent gear to me. I actually had a bit more crit and our frost bonus damage was about the same. My spell hit was ~135, his was 164. I was consistently down on damage by 2% when we compared numbers. Later, when I swapped gear to get my hit to 164, our damage was about equal -- meaning my damage increased by a noticeable margin. There wasn't much AoE here... just frostbolt spam. I do not have a WWS report for this one, unfortunately.

I don't know if there is a definite answer, but just from my observations, it seems Blizzard has "fixed" the ghost hit for frostbolt. Am I missing something here? It's fine if I'm wrong, but my observations don't jive with what is currently accepted.

I know that these examples are far from being a controlled environment, but it's the best I have.

I'll indulge your query because the actual testing of the EP bug is lost somewhere deep in an old thread we don't use any more and a pig to find.

Back then, it was detected on WoW forums that some chap who had raided frost (which was really bad at the time) for months had logs of thousands of Bolts, recording between 0.9 and 1.3% resist at any hit value beyond 130ish.

There were the following possible scenaria:

1) Emp Fbolt was giving 5%
2) Binary effect was interfering with "innate boss resilience" and cancelling the partial resist along side 5% of the hit
3) EP was giving it's old 2% per point benefit to Fbolt or
4) WC was supplying 5%

We all speculated and did minor-size sampling, which was all moot, because nothing lvl 73 stood still and ate bolts long enough, we didn't want to raid Frost, and even if we did we didn't want to risk sucking balls with 4% miss rate. Not to mention Dr. Boom was too low level. Until some God-sent man, copied a lvl 65 mage to PTR, chugged some mp5 elixirs, and came up with a dozen Parsings showing:

1) WC gave 0% hit.
2) Emp Bolt gave 0% hit.
3) 3 points in EP gave 6% hit.
4) All of the Above gave 6% hit.

Each test cycle was in the region of 600-700 frostbolts and the discrepancy from assumed value was to the tune of 0.5%. This was the last chapter in determining the Frostbolt bug.

This is irrefutable evidence (1) EP is giving 6% instead of 3% (2) it is 6% and not 5% (3) No other talent has anything to do with it.

There is no reason to wear more than 125hit as frost, assuming no dranei in group, no ToW in group and 3/3 EP. Period. This is one of the most well-established data-researching efforts we've seen, outside of raid-data on EJ.

Edit: As a side-note, in future, if you do intend to turn up with artefacts about anything, please note:

(not specifically aimed at you, Max)
1) Sample sizes of less than 500 samples are ridiculously prone to dilution/spikes/flukes. Trying to determine a 1% miss, when 1% represents "two frostbolts" is idiotic statistics. So is assuming "1% = 1 in 100", it shows statistical ignorance.
2) 5-man data is irrelevant; mobs that are not 73 do not have 17% innate resist rate. It can also be argued that 73 mobs that aren't bosses may also provide wrong results, because they don't include innate boss resilience, which may or may not affect the tests.
3) Samples in a non-controlled environment are extremely bad value; if you do a run where a mob is Banished or switches on CloS, or Spell Reflect, or Reflect Shield, depending on which damage meter you use you'll record a possible resist when it isn't due to your miss rate. The same is true of being Mind Controlled and attacking your own team-mates , who may also resist.

Aldrir · 07/07/08, 5:03 AM

Originally Posted by Vontre

Specific Builds:
2/48/11: Fire with Icy Veins. Best possible dps.

0/40/21: If you really need two Ice Blocks. This build is actually competitive for dps as well because it allows for two Icy Veins during Bloodlust.

10/48/3: Standard fire but no Icy Veins, Clearcasting instead. Used for better mana efficiency, but generally a poor choice compared to the first two fire options.

10/0/51: (or other frost variant): Deep Frost with winter's chill and Water Elemental. Very strong for guilds working on Kharazan or specific encounters that benefit from mage kiting skills. Scales poorly.

40/18/3: Arcane Blast with Fireball/Scorch.

40/0/21: Arcane Blast with Frostbolt, strongest 'arcane build' due to Icy Veins and Cold Snap.

48/0/13: Heavy Arcane with Icy Veins. This build benefits from itemization synergies because all spells used are Arcane spells, effectively lowering the effective hit cap by 7%. Unfortunately Arcane Missiles is poor dps, and while theorycraft has hinted at possible use for this spec, for the moment it is considered universally inferior to Arcane/Frost.

33/28: Arcane/Fire 'huge crit' build. This is a poor build that should never be used in raids.

could you please post links to these builds?

Pintofbrew · 07/07/08, 7:55 AM

It is assumed these builds are rather obvious. I'll save Vontre the hassle, however,

3/47/11 Talent Calculator - World of Warcraft . One point in Arcane Focus helps with RoS and Council more than 1/2 in Blazing Speed.

0/40/21 Talent Calculator - World of Warcraft

In both these builds, imp. Frostbolt is for Al'ar and Illidan P2, in the latter case imp. Blizz is for Illidan Shadow Demons, and in both cases imp. Nova is for Shadow Fiends on Illidan as well.

10/48/3 Talent Calculator - World of Warcraft

10/0/51 Talent Calculator - World of Warcraft

40/18/3 Talent Calculator - World of Warcraft I'd avoid this spec.

40/0/21 Talent Calculator - World of Warcraft Cold-Snapping an extra IV is practically what keeps this spec superior to:

46/0/11+4 Talent Calculator - World of Warcraft

33/28/0 Talent Calculator - World of Warcraft Some believe this can compete in raid environments. They are few, but have come up with competitive WWS logs. This spec also suffers at low item levels from lacking a lot of hit.

cdutka · 07/07/08, 6:42 PM

[quote=Pintofbrew]
10/0/51 Talent Calculator - World of Warcraft
[quote]

I've always wonder why you would use this spec, it seems to me that permafrost & Imp. Blizzard are not really needed in raiding out side of Hyjal. I also find Cone of Cold/Arcane Blast more effective AoE then Blizzard. The spec I use is quite similar, however I have moved the two needed filler points to Cone of Cold, while adding Magic Absorption 5/5 and 2/3 Arcane Meditation. My guild is only upto RoS (50% through phase 3), anyways just my two cents.

[b]17/0/44[b] Talent Calculator - World of Warcraft

Roywyn · 07/07/08, 8:18 PM

Originally Posted by cdutka

Originally Posted by Pintofbrew

10/0/51 Talent Calculator - World of Warcraft

I've always wonder why you would use this spec, it seems to me that permafrost & Imp. Blizzard are not really needed in raiding out side of Hyjal. I also find Cone of Cold/Arcane Blast more effective AoE then Blizzard. The spec I use is quite similar, however I have moved the two needed filler points to Cone of Cold, while adding Magic Absorption 5/5 and 2/3 Arcane Meditation. My guild is only upto RoS (50% through phase 3), anyways just my two cents.

17/0/44 Talent Calculator - World of Warcraft

Permafrost and Imp. Blizzard are very good for learning Illidan.
I'd definately recommend taking them - especially since he's about the sole reason to spec frost in BT/HS.

Imp. CoC is a terrible talent. First because is CoC is terrible (spam AE for DPS, or talented Blizzard for CCing waves).
Second because CoC has a very low AoE damage cap that is not increased by the talent.
Yes, the talent essentially does nothing at all on mass AoE.
Just like Arcane Power doesn't do much/anything either.

For Illidan, I took shatter/frostbite as well, which had me 56 in frost, with 5 points to spend on arcane.

Magic Absorption is generally seen as a bad talent as well since your chance to resist non-binary spells is fixed and you resist very little. And you don't really have any use for that mana in a 1X/0/4X spec.

If you go past 10 points in arcane, make sure to pick up Magic Attunement (i.e. "Imp Amp") and then choose a filler of choice (usually 3/3 Arc. Impact).

When it comes to choices/flavour, I've never had Incineration.
I've been using 2/45/14 as I prefer the utility of Imp. Blizzard on Illidan/Z'A/Hyjal over 4% crit on a spell that does 3% of my damage.
That's not a spec recommendation (I fully expect and understand people disagreeing), just throwing in a flavour.

Goedel · 07/07/08, 9:00 PM

My preferred frost spec for raids became something like this 18/0/43. A few points are discretionary, including the 2 in Improved Blizzard. In the Arcane tree, I think it can be nice to have one mage in a raid with Magic Attunement, and the frostie works well for that.

However, I had also seen discussion recently, though perhaps not passing the level of inconclusive anecdotal observation, that Magic Absorption does much more than had been supposed. A friend and I--also not with such rigorous testing--have seemed to notice more opportunities for it to work as well, with a host of hidden opportunities such as chances to "resist" bosses' picking random targets, events beyond "Boss casts Nasty Bolt at you for 5k damage" that don't typically show up otherwise in a combat log. This 18/0/43 spec is mana efficient enough already that the gains from Magic Absorption are likely to be superfluous, but I would like to explore the mysterious details behind this further, as it could be a very reasonable choice in a 50/0/11 or 40/0/21 build.

Neither I nor the other mages in guild have used Permafrost or Improved Blizzard on Illidan, just Frost Nova, Cone of Cold, and a finishing Ice Lance or Arcane Explosion if necessary. That assumes either the mages move to the infested person or vice versa, and I guess others might do it from a distance, but moving certainly works just fine.

Pintofbrew · 07/07/08, 9:15 PM

Everything Roywyn said rings true. Permafrost and i-Blizz have nothing at all to do with MH. In fact, in MH the only thing you give a damn about is AoE DPS, and for that both Blizz and CoC are utter bullshit.

Shatter/CoC are only for Parasites on Illidan, Perma/iBlizz are only for Shadow Demons on Illidan.

Even 10 points in Arcane are more or less redundant for Frost. Some, like Goedel, spec 18 in Arc for meditation, though unless you're never grouped with an SP I'd say it's utterly redundant.

If you must spec Arc Med at least spec only 2/3, because with 3/3 you have to skimp on Ice Floes and in certain cases that does impact your DPS. In long boss-fights with correct CD management (this is doubly true of long T5 content) you'll Cold Snap twice, meaning one more WE and one more IV. A non-trivial sum of real DPS.

Not to mention, reduced CoC cooldown is excellent for Illidan P5 when the bloody parasites are double-rate.

Edit: Having said that, 2/2 Ice Floes and only 4/5 WC isn't a bad idea at all.

Goedel · 07/07/08, 9:23 PM

Aye, as Pintofbrew suggests, even setting aside the murky Magic Absorption potential, 18 in Arcane can be more than enough mana. The flip side is that none of the missing points in Frost are really significant for raid boss DPS, so if the mana might be useful and the utility/PVP bits might not, you might as well. My original move to that spec came from Zul'Aman timed runs, when the benefit to efficiency (read: speed) was tangible and some of the buffs from a 25-person raid were missing. I kept it for raids when, indeed, I could not count on a shadow priest. (I later switched to fire and then to my currently beloved 40/0/21, which amuses me when 1012 m/5 while casting (in proximity to Jaina) is not more than enough.)

Fujisaw · 07/08/08, 2:14 AM

ok im a 50/0/11 arcane mage turning into a fire mage i was wondering if 7/43/11 because you dont have it as one of the fire specs to use. just by looking at it vs the fire specs you have this seems more optimal of a spec just wondering if i can get some back up on this or am i way off target.

Zmaj · 07/08/08, 3:09 AM

Originally Posted by Fujisaw

ok im a 50/0/11 arcane mage turning into a fire mage i was wondering if 7/43/11 because you dont have it as one of the fire specs to use. just by looking at it vs the fire specs you have this seems more optimal of a spec just wondering if i can get some back up on this or am i way off target.

If you are going to go fire, the 2 optimal specs like pintofbrew mentioned are 3/47/11 and 0/40/21.

Going more than 3 points into arcane as fire spec is a waste. Damage increasing talents should always take precedent over mana regeneration ones. The reason 0/40/21 allows you to go 2/3 Playing with Fire and Pyromaniac is the cold snap you pick up in the frost tree. It allows you to refresh your icy veins and is virtually the same dps as 3/47/11.

Fujisaw · 07/08/08, 3:39 AM

but i get everything a 3/47/11 gets but Pyroblast which shouldnt be used in raid, Blastwave AE is more effective anyways, Dragon's Breath AE is better in more than 1 way and 1 point in Incineration since i will only be casting 1 scorch every 7-8 Fireballs if i am in charge of that. With 7/43/11 i gain Clearcasting you can max incineration if u feel its worth the point but you still get 1 point in CC

Pintofbrew · 07/08/08, 7:10 AM

See this is what I don't get. There's a "sweet info" thread, with a guideline of what's considered "standard" for everyone in the mage comunity. You come up, juggle some talent points, make up an almost identical spec with quite a few utterly mind-boggling choices, ask "why doesn't anyone mention this spec I invented", somebody answers "it's sub-par" and you counter their statement with "no it's not".

Do you post wanting to hear an opinion, or do you post to argue why your choice of weird spec nobody else abides by is as good as if not better than what we all chose?

Blast Wave is incidental. Some like it some don't. Dragon's Breath is a lot more useful in AoE packs that fuck-up, as well as AoE packs when locks pull agro. It's safer to help them with a DB rather than a Nova, which will still kill anyone in melee range, which might be you.

You spec 60% AM interrupt resist. Congrats. You missed at least two points in Arc. Focus, thus you're not hitcapped for Spellsteal on Council and may *resist* the first one, dying, or may *resist* on stealing RoS, wiping your raid. Better still, you may *resist* countering CoH on council while BoP is up and cost your raid another 1:45 worth of encounter on the longest, most exhausting T6 content fight because you felt 60% uninterruptable AM was a superior choice.

I'm not even going to coment on Frost Warding, patently the worst early talent in the mage tree, marginally beaten by i-Wand.

Then the fantastic 3/5 i-Fbolt and none in either permafrost or i-Blizz. You have no AoE slow, you have no AoE stun, you can't help on Illidan Parasites or on AoE pulls in MH which botch and the paladin gets gibbed. You also have no nuke to cast on Illidan P2, but as you seem to be early BT I assume you'll respec for that.

Finally, the 4% clearcasting will do exactly: Fuck all. You're losing a significant amount of usefulness without gaining anything that'll change your game play for the better. Even with 10% clearcast, you'd still need an SP to perform. Now you still need one, just a tiny bit less. So when the encounter is over, you and your fellow 0% clearcasting mage who is in the same SP group, will both have used 1 pot, 0 gems and no Evo, just he'll be on 4k mana and you'll be on 5.1k. This is why we don't spec Clearcasting.

Frenzi · 07/08/08, 8:52 AM

If you are going for a frost build in T5 or T6 content then I would recommend 13/0/48: Talent Calculator - World of Warcraft it has all the utility that a frost mage needs especially for AoE. Even so Imp Blizz is personal preference as I am not sure why you would ever need it for Illidan as permfrost and CoC is a far better combination than using imp Blizz. Parasites all die with 3 mages doing DB + BW easily enough anyway so I am not sure why frost is needed at all here?

To the point about clearcasting I think a lot of you are forgetting how much mana you can get back from an SP that has only just started in BT. I quit the game a few months ago with my T6 farming guild and certainly mana was never an issue but I have since joined a guild much lower down and only just started making moves in T6 and having Archi downed a couple of times and teron down in BT the SP's regen far less mana back. I am using a standard 2/48/11 build and I am going through pots, gems and sometimes even evocate as the spriests in the guild aren't at the gear level where mana becomes a non issue for mages.

Maybe 4% clearcast is just what he needs and works much better for him than it would for you at your level.

Norfair · 07/08/08, 8:58 AM

Originally Posted by Frenzi

If you are going for a frost build in T5 or T6 content then I would recommend 13/0/48: Talent Calculator - World of Warcraft it has all the utility that a frost mage needs especially for AoE. Even so Imp Blizz is personal preference as I am not sure why you would ever need it for Illidan as permfrost and CoC is a far better combination than using imp Blizz. Parasites all die with 3 mages doing DB + BW easily enough anyway so I am not sure why frost is needed at all here?

The main reason for mages to spec frost on that encounter and take improved Blizzard is for the demons that spawn during p4. It's dangerous to get close for CoC, since if you are the one being targeted by a demon there is very little time to dps it.

Frenzi · 07/08/08, 9:06 AM

Originally Posted by Norfair

The main reason mages to spec frost on that encounter and take improved Blizzard is for the demons that spawn during p4. It's dangerous to get close for CoC, since if you are the one being targeted by a demon there is very little time to dps it.

Ahh sorry I thought you were on about the parasites, my bad. I have never seen blizz be needed for the demons either though?

Pintofbrew · 07/08/08, 9:11 AM

Originally Posted by Norfair

The main reason for mages to spec frost on that encounter and take improved Blizzard is for the demons that spawn during p4. It's dangerous to get close for CoC, since if you are the one being targeted by a demon there is very little time to dps it.

If you're the one being targeted by a demon you're paralysed until it's dead. Not to mention, CoC's slow is way inferior to full-talented Blizz.

I suspect you meant "if you're being targetted by a parasite". Many guilds practice Nova-CoC. Using Blizzard is a monstrous mana-hog and does not freeze the wee bastards to start with, which is essentially the selling point of the Nova-CoC tech. With two mages dealing per parasite spawn, two GCDs is all you need to kill them and you never need to be within their melee range for any of the AoEs in question.

Norfair · 07/08/08, 9:12 AM

Originally Posted by Frenzi

Ahh sorry I thought you were on about the parasites, my bad. I have never seen blizz be needed for the demons either though?

I've put it a bit wrong: the main reason to spec frost is not for improved Blizzard. But the main reason to take improved Blizzard is for the demons in p4. It's surely not needed, but it does make it easier, so if you're respeccing solely for that encounter you might as well take it.

Originally Posted by Pintofbrew

If you're the one being targeted by a demon you're paralysed until it's dead. Not to mention, CoC's slow is way inferior to full-talented Blizz.

I suspect you meant "if you're being targetted by a parasite".

No, I really meant demon. The point was that the rest of the raid gets very little time to dps it if you are too close to Illidan prior to the demons spawning. Therefore, for slowing demons, use improved Blizzard since you can do this from a safe distance. For parasites, nova-CoC is indeed a better way.

Frenzi · 07/08/08, 9:22 AM

Yes when I used to play frost I played with the spec I linked, I was actually the guy Pintofbrew was referring to previously when he was talking about the elemental precision bug as I was at that point the guy playing frost at high end PvE, I switched after my first Illidan kill to 10/48/3 :P