Dots and you: The Affliction Warlock Thread - Page 57

Galanna · 04/02/09, 6:40 AM

Originally Posted by madindehead

Sorry to bring this up again, but where is this coming from?

Wowhead: OM NOM NOM still has the same talent as before, ie the 12% increased shadow damage below 35%.

Yes, but WorldofWarcraft.com -> Info -> Classes -> Warlock -> Talent Calculator still has the same talent as even before, which is "increases the critical strike chances of your Shadow spells by 5% when your target is at or below 35% health.", which is incorrect.

VenomByte · 04/02/09, 10:22 AM

Originally Posted by shauno

The necklace issue was calculating the +5 SP bonus as +8 SP .. that's what I get for late night coding >_< baah, a few number of bugs in the gem section of code... should be all fixed now, just running another recalculate.. lets see what we get this time.

chardev.org v6 ~ a World of Warcraft character planner

Now THAT looks a lot better huh

Thanks *alot* for your input

I'm running it now for 368 hit rating and soon will add in 2pc/4pc bonuses T7

368: chardev.org v6 ~ a World of Warcraft character planner

I'm currently tweaking it to have a wider item choice and be faster with more options.. see how it goes.

I've written an optimisation program myself. My results vary slightly to yours.

On the 289 hit, the difference is negligable, but my build is chardev.org v6 ~ a World of Warcraft character planner
The only differences are these: Item Comparison Tool - World of Warcraft - I have arrived at Matriach's Spawn and Gloves of the Fallen Wizard instead of Surplus Limb and Gloves of Token Respect. I have more spirit at the expense of haste.

Using the weights given earlier (are you still using these or have they changed?), this appears to give my spec a 0.46 higher weighting, equivalent of ~1 haste (virtually nothing)

On the 368 hit rating spec, the results differ more significantly. My result is here: chardev.org v6 ~ a World of Warcraft character planner
The differences are as follows: Item Comparison Tool - World of Warcraft

Using the same weighting, and assuming I've not botched my maths, this gives my set a higher score by 11.98 (worth approximately 18 spell power). I think something may not quite be right with your 368 hit model, as that's a reasonably significant difference.

I have not yet attempted to model T7 bonuses, trinkets, or profession bonuses which may affect hit, such as blacksmithing and jewelcrafting, but otherwise my algorithm should be covering the search space completely, including socketing for +hit with +8/+16 gems and utilising socket bonuses where appropriate.

EDIT: For anyone who's interested, the stat difference between my 289 & 368 hit profiles is as follows:
-1 sp
-57 haste
-0 crit
-22 spirit
-19 int
which is I believe ~50 dps, give or take. That suggests to me that gearing for 368 hit and putting your 3 talent points elsewhere would be more optimal

supplicium · 04/02/09, 1:28 PM

If this is supposed to be the ideal set up going into ulduar (which I believe is the point now) then you need to compensate for the fact that a chaotic skylfare is far superior do to dot crits are actual crits now rather then a separate effect and there by 3% inc crit damage is definitively better (rather than debatably better) IE you need 3 blue gems.

Also next patch professions are probably JC and tailoring for min/max. One gives you significantly more static damage and one gives you a high procrate (45seconds 80% proc rate) of a large amount of damage.

VenomByte · 04/02/09, 3:15 PM

The entire gearset will likely be off to some degree as far as patch 3.1 is concerned, since the weighting given to crit is significantly lower than it is likely to be after the 3.1 changes are live. Once we have a reliable set of new weightings, the ideal gear can be recalculated.

shauno · 04/02/09, 8:10 PM

Hey VenomByte,

Good to have a comparison to work with.
I have found a few bugs with my score adding and fixed most of it up and have come to this 368 hit rating set:

chardev.org v6 ~ a World of Warcraft character planner (glad can save CharDev again!)

Only swapped the chest out for Heigan's Putrid Vestments

Your score: 1999.75
My score: 1999.82

Stat differences with using my set:

+10 Intellect
-66 Spirit
+51 Haste

Although - TBH, with Fel Armor, that's a tossup between 22 SP or 51 haste, really. But the stat weights essentially control that.

But yes, as far as 3.1 is concerned, once there are some definitive stat weights, it will be easy to re-run.

Skellum · 04/02/09, 8:39 PM

Question, should I be using Ember or Chaotic skyflare as affliction right now?

gia · 04/02/09, 10:56 PM

I've been thinking about getting the range talents (grim/destructive reach) for a while and after trying out ulduar on the PTR I've really made up my mind, I find that I often lose dps time because of moving around and/or the boss moving or being kited out of range. I did a few attempts at Vezax the other night and I often found myself having to stand on the shadow crash buff and being out of range of the boss. Maybe it's just me being new at warlock or the fact that we were still learning the fights though. Feel free to tell me that I just need more practice.

Anyway, I've played around with simcraft a bit and dropping 2 points from nightfall and 1 point from demonic power is roughly a 70 dps decrease to gain those 6 yards, is that my best option or are there better alternatives?

I've also thought of getting speed enchant or meta gem for movement fights, thoughts?

Kurotowa · 04/03/09, 4:32 AM

Buffs to things like run speed and spell range are damned near impossible to plug into a DPS calculation, because they don't directly impact the numbers. What they do impact is things like "how quickly can you not stand in the fire" and "does the range give you an additional and superior location to stand that is not in the fire" and so on. (Or, for several of the Ulduar bosses, stand IN the fire.) The value of these is somewhere between zero and a whole lot, depending on the fight. On your classic Patch DPS test, where you do nothing but stand in one spot and cast, they're worth zero. On the more chaotic fights they can be worth a great deal, especially if you're still learning it, because they're basically giving you a cushion to your response time and positioning.

So in short my advice is that if you feel like they're helping, by all means take them. They're training wheels, and should be used to train with. After you know the fight and have tightened your play you can drop them again.

VenomByte · 04/03/09, 4:40 AM

Originally Posted by shauno

Hey VenomByte,

Good to have a comparison to work with.
I have found a few bugs with my score adding and fixed most of it up and have come to this 368 hit rating set:

chardev.org v6 ~ a World of Warcraft character planner (glad can save CharDev again!)

Only swapped the chest out for Heigan's Putrid Vestments

Your score: 1999.75
My score: 1999.82

Stat differences with using my set:

+10 Intellect
-66 Spirit
+51 Haste

Although - TBH, with Fel Armor, that's a tossup between 22 SP or 51 haste, really. But the stat weights essentially control that.

But yes, as far as 3.1 is concerned, once there are some definitive stat weights, it will be easy to re-run.

I've fixed a slight bug I had in my system and re-run it for 289 and 368 hit.

For 289, I get the same set as I did before. For 368, I now get exactly the same gear set as you do. Given that the two independently written apps have now converged on the same setup, I am 99% sure that what we've come up with is indeed optimal for the set of weights given.

It's worth noting though, that as little as a +/-0.1 on any of the weights can change the resulting gear set slightly. When they're recalculated for 3.1, it would be nice to have an extra decimal place on them, if this is possible within the margins of error used to arrive at the numbers.

shauno · 04/03/09, 5:00 AM

Originally Posted by VenomByte

I've fixed a slight bug I had in my system and re-run it for 289 and 368 hit.

For 289, I get the same set as I did before. For 368, I now get exactly the same gear set as you do. Given that the two independently written apps have now converged on the same setup, I am 99% sure that what we've come up with is indeed optimal for the set of weights given.

It's worth noting though, that as little as a +/-0.1 on any of the weights can change the resulting gear set slightly. When they're recalculated for 3.1, it would be nice to have an extra decimal place on them, if this is possible within the margins of error used to arrive at the numbers.

Good to hear. I will link my findings here:

No 4PC T7 Bonus:

289 hit: chardev.org v6 ~ a World of Warcraft character planner
368 hit: chardev.org v6 ~ a World of Warcraft character planner

With 4PC T7 Bonus:

289 hit: chardev.org v6 ~ a World of Warcraft character planner
368 hit: chardev.org v6 ~ a World of Warcraft character planner

Cheers.

VenomByte · 04/03/09, 7:32 AM

I have fixed another bug, which prevented two rings with zero hit rating being used. This changes both my specs (both are non-T7)

New 289 spec is here: chardev.org v6 ~ a World of Warcraft character planner

New 368 spec is here: chardev.org v6 ~ a World of Warcraft character planner

The scores just keep going up!

Hit rating gems are now optimal, which is an interesting result.

(Note: I just realised I overwrote my old chardev links with these, so if following links earlier in the thread proves confusing, this is why)

Troffel · 04/03/09, 10:21 AM

shauno, VenomByte: Can you instruct your software to calculate equipments for alliance, which have Heroic Presence (Racial ability of Draenei)?

The targeting hit ratings are 262 and 341.

I am very curious what solution you get for 262 hit rating with the 4pc T7 bonus.

NinjaSquirrel · 04/03/09, 10:39 AM

Like Troffel, I am also curious on what your tool generates to account for Heroic Presence. As alliance, I have an advantage that our shadow priest is also draenei.

My thoughts were to change out the Signet of the Malevolent for the Band of Channeled Magic, change the yellow socket gem in the Valorous Leggings from +19 spell power to +9 sp and +8 hit, and changing the +18 spirit boot enchant to Icewalker.

This change would equal out to 260 hit (3 shy of the cap)
Overall difference:
+7 int
-7 stamina
-29 hit
+9 spell power
+16 haste
+14 spirit
+12 crit

I am wondering if being 0.1% shy of the hit cap is worth changing to this set up or if there was a better combination to drop to the 263 hit rating?

Troffel · 04/03/09, 10:57 AM

Originally Posted by NinjaSquirrel

...
I am wondering if being 0.1% shy of the hit cap is worth changing to this set up or if there was a better combination to drop to the 263 hit rating?

Hopefully the calculation of their tools take the minimum of the target hit rating and the hit rating from gear in the dps-calculations.
The tools may give a hit rating from gear different from the target hit rating, but with a maximum for the dps.
My expectations is that the gear, which delivers a maximum of dps, will have a little less hit rating than the target hit rating.

I have to admit that I hate, if a haunt misses.

VenomByte · 04/03/09, 12:21 PM

Originally Posted by Troffel

shauno, VenomByte: Can you instruct your software to calculate equipments for alliance, which have Heroic Presence (Racial ability of Draenei)?

The targeting hit ratings are 262 and 341.

I am very curious what solution you get for 262 hit rating with the 4pc T7 bonus.

Isn't the 10% target 263 when you round up? Similarly I believe the 13% comes to 341.016, but I think that can safely be left as 341.

My result for 263 is as follows: chardev.org v6 ~ a World of Warcraft character planner

For 341 I get: chardev.org v6 ~ a World of Warcraft character planner

For T7 sets you'll have to wait for Shauno. My App isn't set up to do them yet.

EDIT: If you are willing to go just under the 10% cap, you get: chardev.org v6 ~ a World of Warcraft character planner at 261 hit. This is 1.3 hit points under the cap, but 1.8 weighting points higher in dps (a little under 3 spell power's worth). In theory this would give you marginally better dps.