Topic Details

Frost 2h Gemming

25 replies to this topic Started by Vuoq, Dec 02 2012 03:41 AM
- - - - -
  • Please log in to reply

Posted 02 December 2012 - 03:41 AM

#1
Vuoq
  • Vuoq
  • Peasant
  • NO DEFAULT AVATAR
  • Posts: 19
  • Reputation: neutral (0)
  • Joined: 16-November 12
Heya, Well I have been reading and from talking to some friends, that when we have good enough gear, say atleast 4 piece tier, by which point we should have decent to good baste str from our gear, we can and should start gemming for 320 strenght for better dps? Also that although str stat weight is higher, we get double haste per gem than we would str , so the overall stat weight gain is like 1 str vs 2 haste, and it comes out on top ? (hope that makes sense) has anyone tried gemming 320 haste instead of str gems? does it come out on top and if so around how much str should we start trying to gem for haste. Cheers

Posted 02 December 2012 - 11:18 AM

#2
Talabas
  • Talabas
  • Vagrant
  • NO DEFAULT AVATAR
  • Posts: 9
  • Reputation: neutral (3)
  • Joined: 27-November 12
  • LocationZurich, Switzerland
Once 2 points of haste are worth more than 1 point of strength then, yes, it makes sense to gem haste in preference to strength. Exactly when you start doing this depends on your individual stat weights so you'd need to check them using a decent simulation tool such as Simulationcraft. I'm currently averaging item level 484 on my DK and strength still appears to be the way to go in my personal case (and it appears to be a while until I'm going to need to take the haste gemming strategy). In particular, for boss fights with significant movement, the value of haste drops considerably compared to strength in my case.

Posted 03 December 2012 - 01:41 AM

#3
Vuoq
  • Vuoq
  • Peasant
  • NO DEFAULT AVATAR
  • Posts: 19
  • Reputation: neutral (0)
  • Joined: 16-November 12
I had a look with simcraft, and it came up with this for my current stat weights: Str: 2.73 Haste: 1.84 So since haste gems give double the str gems, 1.84 x 2 = 3.68 , making haste gemming better for my gear already? I have read that some people suggested that gemming for hastte then rerunning the sim will just show the opposite that strenght will become better. Is there a site or something I can change my gems then check stat weights , or is there a way to do this in Simulationcraft . I am pretty new to the software. Also as you mentioned, with bosses that have alot of movement surely the value of haste would drop and str would reign supreme as a all around dmg increase. Edit: I found a website, Ask.MrRobot. It optimized my current gear for 320 haste gems, while still using the 320 str JC gems x2. Odly enough this site is arguing that I cann stay 0.23% under the hit cap and get away with it by having more secondary stats. Anyone agree with this ?

Edited by Vuoq, 03 December 2012 - 02:08 AM.

Posted 03 December 2012 - 11:25 AM

#4
Talabas
  • Talabas
  • Vagrant
  • NO DEFAULT AVATAR
  • Posts: 9
  • Reputation: neutral (3)
  • Joined: 27-November 12
  • LocationZurich, Switzerland
Wow, I'm not saying the results are wrong but I'm surprised that haste is coming out so high when you're running Simulcraft. As you mention that you're quite new to the software, can I just check that you've downloaded the latest version from http://www.simulationcraft.org and that you've set at least 10000 iterations from Simulcraft's options (the default of 1000 is not sufficient if you want any reasonable level of accuracy)? Also ensure that you're selecting all the stats you want to analyse from the "Scaling" tab under Options. You can also change the style of boss fight within the Options under the Globals tab. By default it's a straight Patchwerk-style tank n spank fight. Moving onto AskMrRobot, I've only found that to be accurate if you take the stat weighting from Simulcraft and manually enter it into the tool. The default stat weights that AskMrRobot provides have always proven to be way off for me personally. By default, AskMrRobot will go under the hit and expertise caps to maximise DPS where it makes sense. There's differences of opinions about this feature and you'll see a lot of discussion and explanation within the AskMrRobot forums about this topic. Personally, I'm with AskMrRobot on this and being slightly under cap isn't an issue if other secondary stats can be increased to provide higher DPS, assuming the stat weights that the tool is using are accurate (i.e. use Simulcraft and input its results into AskMrRobot). Many others would disagree with me on this topic though.

Posted 03 December 2012 - 11:41 AM

#5
Talabas
  • Talabas
  • Vagrant
  • NO DEFAULT AVATAR
  • Posts: 9
  • Reputation: neutral (3)
  • Joined: 27-November 12
  • LocationZurich, Switzerland
One other thing, you can use http://chardev.org to change your gemming strategy and import that into Simulationcraft if you want to see what effect it has on your DPS. (You can also use that site to change gear, enchants, etc. to create all kind of "What If" scenarios). Just make sure you use a decent number of interations within Simulationcraft, at least 10000 but preferably higher if you're fine waiting longer for it to do its work.

Posted 03 December 2012 - 11:44 AM

#6
Vlad
  • Vlad
  • Administrator
  • Posts: 1,404
  • Reputation: exalted (294)
  • Joined: 09-May 12
  • LocationRennes, France
Talabas really said everything I would have wanted to say. Just to quickly get the "under the hit cap" thing out of the way - being under the hit cap is only a minor DPS "loss", which of course gets bigger and bigger the more under the hit cap you are. It's not something that will instantly break your DPS if you're 0.01% under the cap. So, Ask Mr Robot (and anyone else who runs the numbers) will find that sometimes the DPS loss from being under the hit cap by a tiny amount is made up for by the DPS gain from putting that Hit Rating into other secondary stats. It usually happens when your options are "either 0.05% under the cap, or 0.10% over the cap", in which case you end up wasting a lot of stats by getting capped. I'd trust Mr Robot on this one. Regarding the values of Strength and Haste - the same rule applies as was mentioned. If Strength is not at least twice as good as Haste for you, then you should go with Haste gems. I'm still trying to figure out some things about this, and I'll be talking to the people at Ask Mr Robot to make sure we're all in good understanding :)

Posted 04 December 2012 - 12:27 AM

#7
Vuoq
  • Vuoq
  • Peasant
  • NO DEFAULT AVATAR
  • Posts: 19
  • Reputation: neutral (0)
  • Joined: 16-November 12
Thanks for replies, I did some more sims with 10k itterations and haste still came on top. So I now gemmed for haste and with askrobot using my simmed stat weights, I am 0.7% under hit cap . Which doesnt seem bad I have seen a few DKs using both lei shen + dark vortext (both normal) , instead of relic of xuen + one of these. Is there something logical behind this? According to EJ relix of xuen is meant to be BIS ... even above the heroic trinkets for this tier. Although when I sim, my dps is something like this, from highest to lowest: Vortex+ Leishen - Highest simmed dps Relic+ Vortex Relic+ Lei Shen - Lowest simmed dps Does Simcraft taken into account trinket procs? if so is this really more accurate over the EJ BiS guide.? EDIT: Im thinking that since haste is better for me now anyway, I should be using Vortex + leishen(for max haste) Since the haste proc is insane.. but the 105 second ICD is abit worrying that it will be quite low up time =/ Would just like a solid opinion on anyone who has faced this too.. ty EDIT 2: But then again the 1k str passive vs 1k str haste passve (relic vs lei shen), str should win there as haste is only better when its nearly double in value..right? or is haste being a overall stat weight automatically make it first pick.

Edited by Vuoq, 04 December 2012 - 01:11 AM.

Posted 04 December 2012 - 10:39 AM

#8
Vlad
  • Vlad
  • Administrator
  • Posts: 1,404
  • Reputation: exalted (294)
  • Joined: 09-May 12
  • LocationRennes, France
I had a talk with the guys at Ask Mr Robot about this, and after their most recent sims, they actually found that Strength is more than twice as good as Haste in basically every gear level (even at BiS level, which goes against what EJ claims). I'd note that they use Simulationcraft for their sims, but also a bunch of home-made programs. With this in mind, they've altered the weights on the optimizer, which changes everything regarding gemming, basically. It's now Strength in red, str/haste in yellow, and str/hit in blue.

Posted 04 December 2012 - 06:24 PM

#9
Vuoq
  • Vuoq
  • Peasant
  • NO DEFAULT AVATAR
  • Posts: 19
  • Reputation: neutral (0)
  • Joined: 16-November 12
Ah really Posted Image Strange that both sites have such different opinions on the matter...

Here is what I got from my sim results:

Str gemmed dps: 97222
Str gemmed 0.7% under hit cap dps: 97337
Haste gemmed 0.7% under hit cap dps: 97744

All with the same gear, all with Relic+Lei Shen.

Also my gear is fully optimized according to Askrobot, so I guess they did not implement their new stat weights yet.

Did they say why they found strenght being twice as good as haste?

I mean according to my sims and the rule of 2xhaste-1xstr comparision, haste comes out on top from simcraft atleast.
I cant imagine their simcraft results being different unless they changed some options.

Edited by Vuoq, 04 December 2012 - 06:29 PM.

Posted 04 December 2012 - 06:31 PM

#10
Vlad
  • Vlad
  • Administrator
  • Posts: 1,404
  • Reputation: exalted (294)
  • Joined: 09-May 12
  • LocationRennes, France

Ah really Posted Image Strange that both sites have such different opinions on the matter...

Here is what I got from my sim results:

Str gemmed dps: 97222
Str gemmed 0.7% under hit cap dps: 97337
Haste gemmed 0.7% under hit cap dps: 97744

All with the same gear, all with Relic+Lei Shen.

Also my gear is fully optimized according to Askrobot, so I guess they did not implement their new stat weights yet.

Did they say why they found strenght being twice as good as haste?

I mean according to my sims and the rule of 2xhaste-1xstr comparision, haste comes out on top from simcraft atleast.
I cant imagine their simcraft results being different unless they changed some options.


I'll see if I can get someone from Ask Mr Robot to post a reply to this and explain better. In any case, I'll try to get more clarification.

Posted 04 December 2012 - 06:50 PM

#11
Vuoq
  • Vuoq
  • Peasant
  • NO DEFAULT AVATAR
  • Posts: 19
  • Reputation: neutral (0)
  • Joined: 16-November 12
Thank you, I dont doubt their expertise or anything, im just trying to understand why str is being seen twice as strong all of a sudden, especially when the stat weights would suggest otherwise ^^. I personally have always felt getting as much strenght as possible would make sense, since we have alot of str % based stuff, like pillar of frost. and in general hitting as hard as possible sounds the best way for dps. But I have looked into this whole haste gemming thing recently and following the process , made to think haste is better. Although a big factor for me still is that having all that haste largely assumes on constant attacking on the boss, any movement, which most raid bosses have some movement at least, that haste value starts to drop rapidly compared to having tons of str and just hitting as hard as possible before moving again etc...

Posted 04 December 2012 - 07:26 PM

#12
Vlad
  • Vlad
  • Administrator
  • Posts: 1,404
  • Reputation: exalted (294)
  • Joined: 09-May 12
  • LocationRennes, France

Thank you,

I dont doubt their expertise or anything, im just trying to understand why str is being seen twice as strong all of a sudden, especially when the stat weights would suggest otherwise ^^.

I personally have always felt getting as much strenght as possible would make sense, since we have alot of str % based stuff, like pillar of frost. and in general hitting as hard as possible sounds the best way for dps. But I have looked into this whole haste gemming thing recently and following the process , made to think haste is better.

Although a big factor for me still is that having all that haste largely assumes on constant attacking on the boss, any movement, which most raid bosses have some movement at least, that haste value starts to drop rapidly compared to having tons of str and just hitting as hard as possible before moving again etc...


I'm certainly not an expert at the maths behind these things, but what I can say is that things like "hitting harder (STR) should clearly be a bigger DPS increase" are "optical illusions", I would say. Things get calculated and simulated and it's not just that "haste should result in more DPS", it's that it does. In this particular case, I believe that initially, it was thought/simmed etc. that Strength was not at least twice as good as Haste, which made Haste the better option for increasing your DPS. I couldn't say what has changed, though, to reverse the findings.

I'd ask what exactly changes the values of Haste and Strength if you have to move. Unless by move you mean that the boss is out of your attack range altogether (in which case I'd imagine Haste is actually better since it passively increases your rune regeneration, while Strength does nothing).

Posted 04 December 2012 - 08:19 PM

#13
Vuoq
  • Vuoq
  • Peasant
  • NO DEFAULT AVATAR
  • Posts: 19
  • Reputation: neutral (0)
  • Joined: 16-November 12
Hmm I see, maybe I have been thinking of the value of haste in the wrong way. I know it increases our rune regen abit more,but that seems to scale quite slowly. I have been thinking its more to do with alot of extra white dmg we may gain? which is why I have been thinking of haste being inferior for movement when we are not in range of the boss to attack. So the question really comes down to why strenght is being seen as twice as good as haste now to truly understand whats going on here. PS: Having another look at the EJ Frost DPs 5.04 thread, they dont make it entirely clear as to which gem they are suggesting for each slot, they simply outline the best gem in each of the colour categories.

Edited by Vuoq, 04 December 2012 - 08:21 PM.

Posted 04 December 2012 - 10:55 PM

#14
Vlad
  • Vlad
  • Administrator
  • Posts: 1,404
  • Reputation: exalted (294)
  • Joined: 09-May 12
  • LocationRennes, France

Hmm I see, maybe I have been thinking of the value of haste in the wrong way. I know it increases our rune regen abit more,but that seems to scale quite slowly. I have been thinking its more to do with alot of extra white dmg we may gain? which is why I have been thinking of haste being inferior for movement when we are not in range of the boss to attack.

So the question really comes down to why strenght is being seen as twice as good as haste now to truly understand whats going on here.

PS: Having another look at the EJ Frost DPs 5.04 thread,
they dont make it entirely clear as to which gem they are suggesting for each slot, they simply outline
the best gem in each of the colour categories.


So I've had another chat with some from Ask Mr. Robot about this, and the change in stats comes from the newer version of Simulationcraft than what had previously been used. Where the change comes from specifically is hard to tell, and one would need to see what was changed between Simcraft versions.

In any case, they are of the opinion that it's worth gemming for Strength now. As to why EJ lists what they list in "BiS" gear, I don't know. Either it's out of date, or it's based on a different way of simming the stats.

Posted 05 December 2012 - 03:49 AM

#15
Vuoq
  • Vuoq
  • Peasant
  • NO DEFAULT AVATAR
  • Posts: 19
  • Reputation: neutral (0)
  • Joined: 16-November 12
Right, sorry to bother you some more, but I just want to get to the bottom of this as best I can , since im getting mixed feelings from my findings and what EJ/Askrobot are saying and im not sure why.. But after your last post I downloaded the new 510-4 simcraft, which installed as 510-3 for some reason, I have windows 7 64 bit if that makes a difference. Before I post some results+stats I got, is 'scale over' option meant to be left at default? or should I change this to raid DPS. These are the results I got with the new simcraft, and my exact same gear at ilvl 490. I also did manual 5min DPS test on the boss dummy in ebon hold. Haste Gemmed: DPS on Boss Dummy 5min test, no food buff , no prepot, no flask, no army. DPS relic+vortex : 68-69k DPS relic+leishen: 69k-70k Haste Gemmed Simcraft DPS: 10k iterations DPS: relic+ vortex : 91174 DPS: relic+ leishen: 91022 compared to old Simmed DPS of : 97744 Stat weights: str 2.71 ap 1.00 exp 1.84 hit 1.66 cri 1.15 has 1.52 mas 0.79 Character stats Rune regen : 8.87 melee haste: 50.24% AP: 34,894 Strenght Gemmed: DPS on Boss Dummy 5min test, no food buff , no prepot, no flask, no army. DPS relic+vortex : 70-71k DPS relic+leishen: 69-70k Strenght Gemmed Simcraft DPS: 10k iterations DPS: relic+ vortex : 91174 compared to old Simmed DPS of 97531 DPS: relic+ leishen: 91022 compared to old Simmed DPS of 97222 Stat weights: str 2.71 ap 0.95 exp 1.76 hit 1.65 cri 1.09 has 1.53 mas 0.80 Character stats Rune regen : 9.09s melee haste: 46.73% AP: 36,070 Before I discuss what I think abit, would just like to say im still not sure if Simmcraft takes into account trinket procs... I mean from my sims with the previous simcraft(505-5), vorteX+leishen togethere where always higer dps simmed, I think that is mainly because it looks at the passive stats. If so then vortex will ofcourse always sim higher as it has done because its 1k strenght vs 1k haste, which strenght would win easy. But then you need to look at vortex having a sucky ICD compared to Leishen Would like a confirmation on this if possible please. So, by looking at what I founnd from simcrafting, strenght for example is still 2.71 stat weight, vs (1.52x2) 3.04 haste stat weight. Therefore haste gemming should still be superior no? However the default stat weights for my character on Askrobot are; strenght: 3.27 and haste 1..55 (x2 =3.1) , So on those basis str is better. But if new simcraft is showing quite different stat weights and they said they have used a new version of simcrafting to get these new stat weights, it doesnt seem right, especially as many have suggested that askrobot is best used when you enter your own stat weights via simcraft result. Has anyone here tested there simcraft stat weights in comparison to askrobot etc? do you find a similar difference here? I included character stats just to compare how rune regen,ap,melee haste compares between the two gemming methods. Rune regen speeed doesnt seem a whole lot different to me, where as melee haste seems quite more noticable. Although AP is almost 2k higher in str gem, since AP has a low stat weight, I dont think that matters so much. Apoligies again for continuing to come back to this, but I have seen a number of post regarding this gemming issue and I think this thread is starting to get some good information regarding the matter. Just to note, my gear is ilvl 490 with only 1 tier piece (496 gloves). Please let me know what you think when you have a chance and thank you for your help thus far. Cheers

Edited by Vuoq, 05 December 2012 - 04:12 AM.

Posted 05 December 2012 - 10:34 AM

#16
Vlad
  • Vlad
  • Administrator
  • Posts: 1,404
  • Reputation: exalted (294)
  • Joined: 09-May 12
  • LocationRennes, France

Right, sorry to bother you some more, but I just want to get to the bottom of this as best I can , since im getting mixed feelings from my findings and what EJ/Askrobot are saying and im not sure why..

But after your last post I downloaded the new 510-4 simcraft, which installed as 510-3 for some reason, I have windows 7 64 bit if that makes a difference.
Before I post some results+stats I got, is 'scale over' option meant to be left at default? or should I change this to raid DPS.

These are the results I got with the new simcraft, and my exact same gear at ilvl 490.
I also did manual 5min DPS test on the boss dummy in ebon hold.

Haste Gemmed:
DPS on Boss Dummy 5min test, no food buff , no prepot, no flask, no army.
DPS relic+vortex : 68-69k
DPS relic+leishen: 69k-70k

Haste Gemmed Simcraft DPS: 10k iterations
DPS: relic+ vortex : 91174
DPS: relic+ leishen: 91022 compared to old Simmed DPS of : 97744

Stat weights:
str 2.71
ap 1.00
exp 1.84
hit 1.66
cri 1.15
has 1.52
mas 0.79

Character stats
Rune regen : 8.87
melee haste: 50.24%
AP: 34,894

Strenght Gemmed:
DPS on Boss Dummy 5min test, no food buff , no prepot, no flask, no army.
DPS relic+vortex : 70-71k
DPS relic+leishen: 69-70k

Strenght Gemmed Simcraft DPS: 10k iterations
DPS: relic+ vortex : 91174 compared to old Simmed DPS of 97531
DPS: relic+ leishen: 91022 compared to old Simmed DPS of 97222

Stat weights:
str 2.71
ap 0.95
exp 1.76
hit 1.65
cri 1.09
has 1.53
mas 0.80

Character stats
Rune regen : 9.09s
melee haste: 46.73%
AP: 36,070

Before I discuss what I think abit, would just like to say im still not sure if Simmcraft takes into account trinket procs... I mean from my sims with the previous simcraft(505-5), vorteX+leishen togethere where always higer dps simmed, I think that is mainly because it looks at the passive stats. If so then vortex will ofcourse always sim higher as it has done because its 1k strenght vs 1k haste, which strenght would win easy.
But then you need to look at vortex having a sucky ICD compared to Leishen

Would like a confirmation on this if possible please.


So, by looking at what I founnd from simcrafting, strenght for example is still 2.71 stat weight, vs (1.52x2) 3.04 haste stat weight.
Therefore haste gemming should still be superior no?

However the default stat weights for my character on Askrobot are; strenght: 3.27 and haste 1..55 (x2 =3.1) , So on those basis str is better.
But if new simcraft is showing quite different stat weights and they said they have used a new version of simcrafting to get these new stat weights, it doesnt seem right, especially as many have suggested that askrobot is best used when you enter your own stat weights via simcraft result.

Has anyone here tested there simcraft stat weights in comparison to askrobot etc? do you find a similar difference here?

I included character stats just to compare how rune regen,ap,melee haste compares between the two gemming methods. Rune regen speeed doesnt seem a whole lot different to me, where as melee haste seems quite more noticable. Although AP is almost 2k higher in str gem, since AP has a low stat weight, I dont think that matters so much.

Apoligies again for continuing to come back to this, but I have seen a number of post regarding this gemming issue and I think this thread is starting to get some good information regarding the matter. Just to note, my gear is ilvl 490 with only 1 tier piece (496 gloves).

Please let me know what you think when you have a chance and thank you for your help thus far.

Cheers


We're working on something to get to the bottom of this, I'll let you know when I know more :)

Posted 08 December 2012 - 11:38 PM

#17
Vuoq
  • Vuoq
  • Peasant
  • NO DEFAULT AVATAR
  • Posts: 19
  • Reputation: neutral (0)
  • Joined: 16-November 12
Any luck with trying to understand whats going on here? =D

Posted 09 December 2012 - 03:24 AM

#18
cheguevarra
  • cheguevarra
  • Peasant
  • NO DEFAULT AVATAR
  • Posts: 14
  • Reputation: neutral (0)
  • Joined: 03-September 12
  • Locationusa,NE
it is very important. before i regem into all str instead of haste as mrrobot suggest now, id like to hear from you, vlad. yours always, Che.

Edited by cheguevarra, 09 December 2012 - 03:25 AM.

Posted 09 December 2012 - 01:42 PM

#19
Vlad
  • Vlad
  • Administrator
  • Posts: 1,404
  • Reputation: exalted (294)
  • Joined: 09-May 12
  • LocationRennes, France
Hey guys. We ran more simulations.

For reference, my own Frost Death Knight, in 492 gear, has 6,700 Haste Rating, 4,165 Critical Strike Rating, and 1,100 Mastery Rating. Simulationcraft gives the following weights Str=2.94, AP=1.12, Exp=2.08, Hit=1.90, Crit=1.14, Haste=1.22, Mastery=0.91. I'm currently gemmed for Strength, and not Haste. So this clearly shows Strength to be "way" ahead.

I increased my stats to add more Strength at the expense of Haste (so, taking the example further), and I got Str=2.91, AP=1.07, Exp=2.11, Hit=1.95, Crit=1.16, Haste=1.21, Mastery=0.88. Hmm. Haste and Crit are closer in value here, and Strength still wins by a lot.

Now, if instead, I gem for Haste, and not Strength, I get Str=2.93, AP=1.12, Expertise=2.07, Hit=1.91, Crit=1.17, Haste=1.21, Mastery=0.92. Hardly any change, and Strength still wins!

I decided to look at someone with much better gear than me, and see if there were differences. I looked at the armory of Dmitrii, vodka's Frost Death Knight (http://us.battle.net...mitrii/advanced). He is gemmed for Strength (although for some reason he is sporting a single pure haste gem in one of his sockets). His weights are Str=2.97, AP=1.11, Expertise=2.10, Hit=1.93, Crit=1.39, Haste=1.79, Mastery=1.00. Wow! Here, Haste is showing up as being vastly superior to Strength.

We ran a few other generic simulations, in several slightly different gear set-ups (but all with stats similar to what you'd expect to have if you are prioritising Haste as your secondary stat, and are discarding Mastery).

Str=3.08 AP=1.10 Exp=2.45 Hit=2.21 Crit=1.24 Haste=1.58 Mastery=0.94 (this is with 9,000 Haste Rating, 4,000 Critical Strike Rating, and 2,000 Mastery Rating). > Haste wins (barely).

Str=2.84 AP=0.99 Exp=2.25 Hit=2.09 Crit=1.15 Haste=1.34 Mastery=0.86 (this is with 6,250 Haste Rating, 3,000 Critical Strike Rating, and 2,000 Mastery Rating). > Strength wins (barely)

Str=3.15 AP=1.11 Exp=2.53 Hit=2.24 Crit=1.19 Haste=1.69 Mastery=0.91 (this one is with a bit less usual stats, namely 9,000 Haste Rating, 2,000 Critical Strike Rating, and 4,000 Mastery Rating). > Haste wins (by a decent margin, but then you're unlikely to have this kind of stat distribution).

The list goes on, and there are variations like this all along - we tested this with regular amounts of Strength for the appropriate gear levels (15-19K Strength between pre-raid gear and Heroic gear).

Conclusion time? I don't know. Strength seems to be ahead, mostly. In some cases, Haste is ahead. In a few cases, they're almost equal. There isn't a pattern I can easily discern here, and the best thing I can say is to Sim your own character's weights and go with that.

Now, Ask Mr. Robot is using "fixed" weights for these stats. It's telling me to gear my character based on the same weights as Dmitrii. This, I believe, is a problem. I need to bring this up with the guys at Mr. Robot and see what they think.

So to bottom line all of it for you, I can tell you to import your character(s) in Simulationcraft and see what they say. Take that with a grain of salt; as I said above, they tell Dmitrii that he's "doing it wrong" and should be gemming for Haste instead of Strength, so either he's choosing to be wrong (it's possible), or he's not aware of Mr. Robot's suggestions (just as possible), or Mr. Robot is wrong with what it is suggesting (also possible).

I am sorry to keep saying "I will look more into it", but it doesn't seem like it's something with an easy fix or solution, mostly because I just don't know how much trust we can put in the results that Simulationcraft spits out (in this case, at least).

I'll keep you updated.

Posted 09 December 2012 - 03:47 PM

#20
Vuoq
  • Vuoq
  • Peasant
  • NO DEFAULT AVATAR
  • Posts: 19
  • Reputation: neutral (0)
  • Joined: 16-November 12
Thanks for replying, youve done alot of simming I see =) I have had some gear upgrades since my last post, at which time haste x2 outweight str , so I will sim again and enter my own weights into askrobot. Since I am currently str gemmed by going on askrobot default weights, instead of what simcraft was showing me as that would have lead to haste gemming in the 489 gear I had, now I am 495. However it does still seem to be a weird situation with how the simcraft is coming out with these stat weights. I think for a DK str might just still be the safe bet.. since we know its great,where as haste is just still in question for being greate/better. Not sure if its worth mentioning or looking at other classes, but ret paladins for example have often been seen similar to frost DK, for them at least haste wins by a land slide and the dps increase is really noticable too. Whereas for me at least gemming haste wasnt a hugely noticable difference in dps. Thank you for your on going advice and support. EDIT: Do you use patchwerk fight style for the simming? and are any other options besides iterations to 10k changed? Just incase that is causing my results to be any different...

Edited by Vuoq, 09 December 2012 - 03:58 PM.

Active Users

0 user(s) are browsing this forum 0 members, 0 guests, 0 anonymous users