Mage Stats weight (25 replies to this topic)

Hi, someone can help me ? i need to find the weight stats fire frost arcane pré-fear / T14 HM dunno how can i found it
thx

itsokaylolz, on 15 November 2012 - 08:04 PM, said:

Straight from the last version of Simulation Craft:

logs/mage_arcane_pre_raid_scale_stats.log:    Weights :  Int=3.48(0.07)  SP=2.88(0.07)  Hit=2.00(0.07)  Crit=1.16(0.07)  Haste=1.57(0.07)  Mastery=1.53(0.07)
logs/mage_arcane_t14_heroic_scale_stats.log:    Weights :  Int=4.30(0.11)  SP=3.48(0.11)  Hit=3.43(0.11)  Crit=1.58(0.11)  Haste=2.18(0.11)  Mastery=2.12(0.11)
logs/mage_arcane_t14_normal_scale_stats.log:    Weights :  Int=4.07(0.09)  SP=3.41(0.09)  Hit=2.68(0.09)  Crit=1.55(0.09)  Haste=2.12(0.09)  Mastery=1.93(0.09)
logs/mage_fire_pre_raid_scale_stats.log:    Weights :  Int=3.66(0.09)  SP=2.74(0.10)  Hit=2.67(0.10)  Crit=1.98(0.09)  Haste=1.64(0.09)  Mastery=1.26(0.09)
logs/mage_fire_t14_heroic_scale_stats.log:    Weights :  Int=4.83(0.14)  SP=3.60(0.14)  Hit=4.16(0.14)  Crit=3.22(0.14)  Haste=2.45(0.14)  Mastery=1.99(0.14)
logs/mage_fire_t14_normal_scale_stats.log:    Weights :  Int=4.35(0.11)  SP=3.09(0.11)  Hit=3.14(0.11)  Crit=2.60(0.11)  Haste=2.27(0.11)  Mastery=1.73(0.11)
logs/mage_frost_pre_raid_scale_stats.log:    Weights :  Int=3.64(0.07)  SP=2.95(0.07)  Hit=3.63(0.07)  Crit=1.25(0.07)  Haste=1.38(0.07)  Mastery=1.04(0.07)
logs/mage_frost_t14_heroic_scale_stats.log:    Weights :  Int=4.45(0.09)  SP=3.54(0.08)  Hit=6.03(0.09)  Crit=1.99(0.08)  Haste=2.04(0.08)  Mastery=1.80(0.08)
logs/mage_frost_t14_normal_scale_stats.log:    Weights :  Int=4.35(0.07)  SP=3.50(0.07)  Hit=4.95(0.07)  Crit=1.71(0.07)  Haste=1.71(0.07)  Mastery=1.53(0.07)

i don't get this part mmmmmm mayby it's ysut because i'm to blond atm lol^^

Damien, on 17 November 2012 - 09:26 AM, said:

MichaelTheys, on 17 November 2012 - 06:13 PM, said:

Well it isn't easy on the eyes, it is literally straight copypasta'd from simcraft to forums.
What don't you get about it?  I'm just going to clean it up and make it more reader-friendly without changing any of the dataz.

Arcane:
Pre-Raid:
Int: 3.48(0.07)  SP: 2.88(0.07)  Hit: 2.00(0.07)  Crit: 1.16(0.07)  Haste: 1.57(0.07)  Mastery: 1.53(0.07)

Normal Raid:
Int: 4.07(0.09)  SP: 3.41(0.09)  Hit: 2.68(0.09)  Crit: 1.55(0.09)  Haste: 2.12(0.09)  Mastery: 1.93(0.09)

Heroic Raid:
Int: 4.30(0.11)  SP: 3.48(0.11)  Hit: 3.43(0.11)  Crit: 1.58(0.11)  Haste: 2.18(0.11)  Mastery: 2.12(0.11)


Fire:
Pre-Raid:
Int: 3.66(0.09)  SP: 2.74(0.10)  Hit: 2.67(0.10)  Crit: 1.98(0.09)  Haste: 1.64(0.09)  Mastery: 1.26(0.09)

Normal Raid:
Int: 4.35(0.11)  SP: 3.09(0.11)  Hit: 3.14(0.11)  Crit: 2.60(0.11)  Haste: 2.27(0.11)  Mastery: 1.73(0.11)

Heroic Raid:
Int: 4.83(0.14)  SP: 3.60(0.14)  Hit: 4.16(0.14)  Crit: 3.22(0.14)  Haste: 2.45(0.14)  Mastery: 1.99(0.14)


Frost:
Pre-Raid:
Int: 3.64(0.07)  SP: 2.95(0.07)  Hit: 3.63(0.07)  Crit: 1.25(0.07)  Haste: 1.38(0.07)  Mastery: 1.04(0.07)

Normal Raid:
Int: 4.35(0.07)  SP: 3.50(0.07)  Hit: 4.95(0.07)  Crit: 1.71(0.07)  Haste: 1.71(0.07)  Mastery: 1.53(0.07)

Heroic Raid:
Int: 4.45(0.09)  SP: 3.54(0.08)  Hit: 6.03(0.09)  Crit: 1.99(0.08)  Haste: 2.04(0.08)  Mastery: 1.80(0.08)


mmm, formatting...

~Sage

Sagetastic, on 18 November 2012 - 02:14 AM, said:

So this means that Haste is more important than Mastery for Arcane mages, especially if you have normal level raid gear. This is interesting because most players probably have reforged using the 'old' formula:

int > hit (+expertise) 15% > mastery > haste > crit

Even Icy Veins uses the old formula in their Arcane mage guide.

Siniwelho, on 18 November 2012 - 11:59 PM, said:

Haste is only better by 0.04(preraid) / 0.19(normal) / 0.06(heroic)

Normal might have the largest discrepancy, but these are tough calls imo.
I forget what, but didn't something about arcane and soft haste capping change this xpack?

In the end, it is really just getting as much as you can out of the gear you get when you can't get what you want!
Unless you're one of those lucky guys that has all BiS, then I hate you. <3

~Sage

Sagetastic, on 19 November 2012 - 05:51 AM, said:

'by only .04, .19, .06' when you are comparing the number 2 is 2%, 9.5% and 3% better. That is pretty significant.

I thought the big haste change was mana regen changing with haste... I could be wrong though, as I don't really play arcane very much.

Also a note, those 2,9.5,3% differences are all all within the error range given by Simcraft (in brackets after the numbers), so as long as your gear has haste/mastery + another stat, you are looking good for the most part (ideally haste + mastery or hit)

midgettoes, on 19 November 2012 - 09:13 AM, said:

I guess arcane is just going to be redheaded stepchild for a while.

<.<;

~Sage

Go fire, stack crit, lol.

These were simulations made with the default profiles of Simulation Craft. The stat values in the profile can influence the calculations of the stat weights. For example, imagine the weight of haste in a profile where there is so much haste that you are hitting the global cooldown on your main spells.

So, I used the T14 heroic profile and made Haste Rating and Mastery Rating vary from 2000 to 10000 (while Critical Strike Rating remained at 2000). The highest DPS was obtained between 6500 Haste/5500 Mastery and 7500 Haste/4500 Mastery, with the absolute best result with 6750 Haste Rating and 5250 Mastery Rating.

That said, I need to make some more tests and see what the average mana level is when using the stat priority from the default T14 heroic profile. Also, this profile is using Rune of Power as a Tier 6 talent, and things may be different with Invocation.

One problem with Arcane currently is that no one is raiding with that spec, so you can't really inspect top players to see how the spec should be played in progression.

Damien, on 20 November 2012 - 02:41 PM, said:

Yes, that is a part of the problem. Nobody wants to develop Arcane when Fire is so much better.

But then again the simulations give accurate data only for encounters without movement. Personally I lost my faith to SimulationCraft when I ran some Arcane simulations with different Fight Style parameters and got results like:

- HelterSkelter: 591 secs fight, 6 Rune of Power casts
- Heavy movement: 569 secs fight, 6 Rune of Power casts

It seems that SimulationCraft has modelled Rune of Power as a buff and doesn't care about the mechanics involved. It's needless to say that this way the results won't meet the reality.

Siniwelho, on 20 November 2012 - 03:31 PM, said:

The best way to test is likely an "all things being equal" environment because even in the same scenario such as a certain boss you may be targeted by some random feature more so than the next time. You may have to move out of your Rune of Power just after casting, at times, so the better way would be without the mechanics, at least to see where things stand. So from there, fight mechanics are an understood expectation and no matter what the stat weights should be constant. The only thing that may change from one fight to another would be things like talents and glyphs, based on the individual encounter as well as raid make-up. For me, I always think I am lacking but I just call it part of old age.

Beeba, on 20 November 2012 - 05:03 PM, said:

Yes, that's why the standard simulation is a Patchwerk fight - no movement at all.

Quote

Here I have to disagree, because I see it like this: Arcane's mastery is Mana Adept. The more mana you have the bigger hits. If you have to move constantly it will be impossible to keep your mana up. This will lower the value of Mastery and increase the value ot the other stats. (Please don't ask who uses Arcane and Rune of Power in an encounter requiring a lot of movement.)

But when I posted about SimulationCraft I didn't actually think about the stat weights. I just thought that if you want make a simulation then why leave it half-way. Rune of Power is not a buff. It is a circle on the ground you have to stand on. If Amethyst Guardian spits an Amethyst Pool over your Rune of Power you will have to move AND create a new Rune of Power. Pretty much every time you have to move you will also have to cast a new Rune of Power because you can rarely re-use an existing Rune of Power. Six Rune of Powers in a 591 second encounter is doable only in a Patchwerk fight - not in HelterSkelter or Heavy movement.

i hope i am in the correct thread.
i'm having problems finding out what haste cap i should have.

http://eu.battle.net...úrâîgù/advanced


here's my char i'm talking about. i've used mr.robot but well, i still dont know which haste i should have.
cos like, it feels sometimes slow.

kuraigu, on 27 November 2012 - 01:57 PM, said:

Looking at http://www.totemspot.../entry.php?b=44 , you can see what haste gives you what break points. You can see that sitting at 3202 (your haste) your next significant tick in a raid environment comes ~4k haste. Since this is a huge amount of haste, I would have to say it is not worth reforging/gemming/etc to. Not until you have significantly better gear (ie more secondary stat allocation, including higher haste).

As for askmrrobot, it won't detect break points for you. It will take the weights you give it (or default) and optimise your gear so that the points are maximised. If you want to hit a certain haste breakpoint you can put a higher stat weighting on haste up to a certain level by playing around with the settings (try this out if you really want, but for your gear, as I said above, it isn't worth shooting for).

midgettoes, on 27 November 2012 - 03:10 PM, said:

i have reforged abit, please check again.

kuraigu, on 29 November 2012 - 02:29 PM, said:

Looking very quickly at your gear it all looks to be going in the right direction. Basically, follow askmrrobot unless you are incredibly close to a break point (see my previous post) in which case you could reforge slightly just to bump over that point. By incredibly close I would say <1 reforge away (so ~100 points).

Siniwelho, on 20 November 2012 - 03:31 PM, said:

The Helter Skeleter fight is very basic, it's just here to test the various raid event, but it makes you move for 5 seconds every 30 seconds. It basically means a 5 second downtime every 30 seconds unless you can cast spells while moving. I'm pretty sure the tool considers that the character still benefits from Rune of Power after moving.

Simulation results should always be taken with a pinch of salt.

Damien, on 29 November 2012 - 06:42 PM, said:

True, but in this case the needed amount is a truckload of salt.

I can understand that they assume that you can re-use the existing Rune of Power after movement. But I don't understand how they can increase the duration of Rune of Power from 60 seconds to about 90 seconds. If you cast the first Rune of Power just before the pull you will still need nine more casts to cover a 591 second encounter - even in a Patchwerk fight.

Imagine

Siniwelho, on 30 November 2012 - 11:57 AM, said:

Imagine now it one is using Invocation in a heavy movement fight. One has to set idle to complete the channeling of Invocation which is quite some time to do nothing. If you must move in that time, one must start channeling all over. During movement, regardless of spec or talents, there are not many choices of dps one can do. So whether you moved out of a circle or are simply moving with or without an Invocation buff, not a whole lot of dps is getting done. For fire, of course, the DOTs continue to burn. Much of a caster's best dps is done stationary, since movement can stop most casts or channels. I know I can place two Runes of Power near each other faster than I can cast one Invocation. I can move out of something and back while both are active. Currently, the Runes also last 20 seconds longer that the effect of Invocation as well as regenerate mana as Invocation reduces mana regen.