Elemental Soft Haste cap

[Gendori]

Haste is a curve. The trend line is generally a power 2 formula. There are essentially two areas where haste = mastery, and outside of which it's either mastery > haste or haste > mastery, depending on whether you're between the two areas or not. I'd post the plot graph, but LibreOffice doesn't handle as many columns as Excel does so I can't produce the graph at home. It'll have to wait until I'm in the office on Monday.

It's also worth noting that these cross over areas can change, depending on the rest of your gear. This means that what works for one person won't work for another because they have more/less gear, different talents, etc.

Your reforge graph indicates that you're going from one of those areas where haste > mastery to the haste = mastery crossover. At some point above your reforge plot you may see it going up again, but the problem here is that you're sacrificing one stat for another, whereas my plots are all looking at the status quo + additional stats.

As I've said previously, your results are valid for you, but will not be valid for others.

Ok, I feel like were starting to get somewhere here, and Im pretty much in agreement with what you said, except the last part, theres a problem with that.

First off heres my profile for frame of reference

http://us.battle.net/wow/en/characte...ndori/advanced

When I'm deciding how to reforge My character I can't Look at a Set of Plots and say Oh If I add 100 haste I get 300 dps but If I add 100 Mastery I get 280 DPS.

When I decide to add 100 Haste, Im sacrificing something, I can sacrifice 50 INT or 100 Crit or 100 Mastery. Ive reforged all of my crit off to get to the 7060 Number Im at now, so now that Ive reforged all my crit off, where do I go next? Heres where My reforge Plots come in

Here's the latest reforge plots for my character, The Curve is still quite similar to the one I did a Month ago when I had drastically different gear

http://img42.imageshack.us/img42/1903/plots.jpg

At 7060 Haste and 13246 Mastery Im doing 155964 DPS.

Heres what I get for 1000-5000 less and More haste instead of Mastery

2060 Haste 18246 Mastery-152329
3060 haste 17246 Mastery-153732
4060 Haste 16246 Mastery-154271
5060 Haste 15246 Mastery-155078
6060 Haste 14246 Mastery-155505
7060 Haste 13246 Mastery-155964
8060 Haste 12246 Mastery-156073
9060 Haste 11246 Mastery-156098
10060 Haste 10246 Mastery-156668
11060 Haste 9246 Mastery-156907
12060 Haste 8246 Mastery-157150

As you can see I am seeing steady DPS gains reforging into Haste all the way until I get to 7060 (the number has in fact moved up a little bit from before where a Month ago this was around 6050). The Difference in reforging from 2060 Haste 18246 Mastery to 7060 Haste 13246 Mastery is around a 2.33% PATCHWERK DPS gain. If I go the Full Monty and go alllllll the way up to 12060 Haste and 8246 Mastery, Im only looking at a Paltry .76% PATCHWERK DPS gain. Clearly at some point here, Haste has Lost some Value.

Now Youre just gonna have to trust me when I tell you that I have run reforge plots for many many characters the last month, and I have yet to see someone with the EoTE/EB Talents NOT have this smoothing out of the reforge plot curve at around the same area.

Here's where it gets complicated, I could Reforge off all of that Mastery and add more Haste and realize that .76% DPS gain, But is this Relevant to the current tier of raiding? Not Really! Theres 12 really Unique bosses this tier, and Most have some sort of Cleave or AoE component. The Problem with this is that Haste is By far Our worst stat in every situation where you are adding a Mob. The More you add, the worst Haste is, Let me show you the weights Im getting with 2 Mobs

INT-6.09
Crit-3.91
Haste-2.0
Mastery-2.85

Here's 3 Mobs

INT-10.54
Crit 7.39
Haste 3.64
Mastery 5.22

4 Targets

INT-11.63
Crit-8.03
Haste-4.06
Mastery-6.02

And 5 Targets

INT-12.46
Crit-8.09
Haste-4.82
Mastery-6.67

So as you can see, in ever situation where we add mobs, haste becomes by far our worst stat. So the Question for me when Im deciding my reforging is, Is it worth it to sacrifice .7% of my patchwerk DPS to realize a Decent gain elsewhere? For me, I like to play to what our strengths are. We are strong at AoE damage. to me .7% is a drop in a bucket and well worth sacrificing to be very strong at cleaving and AoE

So there ya go, I'm not advocating a simple cut and dry approach. If you want that Ill say go for Hit cap of course, then reforge to around 6-7k haste (use simcraft plots to see where your setup begins to flatten out), then if you care about your patchwerk DPS, stack more haste, Otherwise youll get better results on Cleave and AoE fights with Mastery.

Caveat, while Yes I have showed simcraft to favor crit over Mastery on 2-3-4 and 5 targets, I dont feel the tradeoff of Going Haste to 6-7k > Crit> Mastery is worth it, it would take a lot more work to figure it all out though

[binkenstein]

Smoothing != Breakpoint. What you're seeing is that the reduction of X of one stat is equal to the gain of X of the other. When you see a non-flat point on that graph it means that the dps gain from adding X of one stat is worth more than the reduction of X of the other, or vice versa. You're looking at a comparison between DPS changes, which cannot be used to infer breakpoints, which are massive changes in dps gains for one stat over a short period. As I said, I will grab some examples when I'm at work on Monday.

Additionally, I would not recommend using a PW setup for testing, as it's an unrealistic setup. I tend to use Light Movement for all of my personal work.

Yes, Mastery vastly outweighs Haste when looking at AoE/Cleave setups. This is why I'm favouring Mastery slightly more than haste.

[Booniehat]

Binkenstein and Gendori

Thanks for the insightful discussion. Approaching high values of haste myself and am interested to see what secondary stat I should look at. Bink - you say mastery. Gendori - your numbers show Crit as a LOT more. What gives? Is it haste to 7K ish, then rest to crit?

[Gendori]

Binkenstein and Gendori

Thanks for the insightful discussion. Approaching high values of haste myself and am interested to see what secondary stat I should look at. Bink - you say mastery. Gendori - your numbers show Crit as a LOT more. What gives? Is it haste to 7K ish, then rest to crit?

Thats for AoE and Cleave situations, Single Target I have crit decently behind haste and Mastery for me, Im not sure if its worth the Tradeoff TBH. If I just reforge into lots of Mastery it will be good in every situation.

Im actually Puzzled With the AoE results slightly. I would think Mastery would be as good or better than crit on 4 or 5 targets, Its probably because Ive dumped so much crit, that I dont have the Spellpower buff up 100%, I can run more Plots(!) to figure out whats going on =)

[Deleted]

What if you're talenting PE? From what I understand, the more haste you have, the more damage your FE/EE will do...or am I incorrect?

Can anyone confirm or deny this please? Also I'm hearing a lot of talk about EotE/EB setup for most of this debate; is it because at certain stat weights it's an increase to use that build rather than PE? Is it based on if you have the legendary meta gem and 4 set? Or is it literally personal choice?

I assumed PE was best for near enough every fight in ToT due to the heavy movement of a lot of fights; am I wrong?

[Booniehat]

Can anyone confirm or deny this please? Also I'm hearing a lot of talk about EotE/EB setup for most of this debate; is it because at certain stat weights it's an increase to use that build rather than PE? Is it based on if you have the legendary meta gem and 4 set? Or is it literally personal choice?

I assumed PE was best for near enough every fight in ToT due to the heavy movement of a lot of fights; am I wrong?

I can only attest to practical experience - yes, PE/AS is great in TOT simple cos of the billion "get out of crap" mechanics at any point. Eleblast is fantastic if you can sit and tunnel the boss. In T14, that was pretty ok. T15? I can't think of many fights thats ok.

On the other hand though, for VERY cleave heavy fights, I've started to go with a Echo + Primal Ele build. Someone tell me if I'm doing it wrong. But it makes the most logical sense - Your cleave is multiplied by the sorta-mastery of Echo (as opposed to the 'meh' AS), and your Primal Fire Elemental has an AOE effect. Makes no sense to have Eleblast.

---------- Post added 2013-04-27 at 11:36 AM ----------

Also, if it helps - I simmed my char in 3 seperate situations:
1. Single target
2. 3-target
3. 5-target (Cos thats the max)

Results for the various stat configurations.

HASTE > MASTERY > CRIT [I.e. "standard" Elemental stat prio]
Single Target: 122.6 K
3 Target: 195.7 K
5 Target: 218.9

HASTE > CRIT > MASTERY
Single Target: 122.3 K
3 Target: 197.9.5K
5 Target: 220.7K

CRIT > HASTE > MASTERY [This is what Simcraft is telling me I should be using]
Single Target: 121.6 K
3 Target: 205.4K << +10K from standard
5 Target: 230.3K << +12K from standard

CRIT > MASTERY > HASTE
Single Target: 121.2 K
3 Target: 202.5 << +7K from standard
5 Target: 228.1K << +10K from standard

MASTERY > HASTE > CRIT [Doing this just for completion's sake, its weird]
Single Target: 122.0 K
3 Target: 192.9
5 Target: 218.7K

MASTERY > CRIT > HASTE
Single Target: 122.1
3 Target: 192.9
5 Target:218.4

CONCLUSIONS
In my gear, Secondary stats play very little impact in way of Single target. All combinations were virtually equal.
For Cleaving(3) and full AOE (5), CRIT > HASTE > MASTERY shows the highest simmed DPS. Whether or not its true in reality, I have not tested.
If the above 2 are true, I'd safely reforge CRIT > HASTE > MASTERY for almost no loss in Single Target DPS, but substantial gains in AOE.





Method:
Used my toon: Oikodomeo on Thaurissan.
Used 1000 iterations (good enough for DPS, not scale factors)
Used 1 target, 3 target, 5 target patchwerk setting
Talents were constant: Primal Ele, Ancestral Swiftness.*
Glyphs were constant: Unleashed Lightning and Chain Lightning Glyph*
Used Ask Mr Robot to modify stat weights as above, to replicate the gemming/reforging process without actually spending gold.
Please let me know if my method is invalid. Just trying to add to discussion.

*Echo of the Elements will change results - I'll sim that another time. I used PE/AS at most TOT encounters are best suited to that.
** Don't care about the Glyph of Chain lightning as I'm not tryin to get the actual numbers, but rank the DPS among the various combinations of stats.

[Gendori]

So I figured out why my stat weights were wrong for 4 and 5 targets...I forgot I had unglyphed chain lightning tge other day when I ran some LFR....Ill redo my 4and 5 target sims when I get home from work today....ill also set to light movement as blink suggested

[Intrepid1]

I really cannot understand why there are some shamans sacrificing so much intellect for secondary stats. Even in your own Simc weights int is stronger. Your not only hurting yourself single target dps wise but completely gimping your fire elemental by stacking mastery gems. Me and you Gendori have the same ilvl yet I have 4100 more spellpower. I really have to question what you are doing when your own character is optimized around CL spam only and not around single target dps.

[Gistwiki]

I really cannot understand why there are some shamans sacrificing so much intellect for secondary stats. Even in your own Simc weights int is stronger. Your not only hurting yourself single target dps wise but completely gimping your fire elemental by stacking mastery gems. Me and you Gendori have the same ilvl yet I have 4100 more spellpower. I really have to question what you are doing when your own character is optimized around CL spam only and not around single target dps.

Every elemental has weights that show 1 int > 1 haste or mastery. However, many have weights that show 2 haste or mastery > 1 int. Because secondary stat gems have 2 of their stat for every 1 stat point on a primary gem, in instances where 2 haste or mastery > 1 int it'll be more single target dps to use secondary stat gems.

The argument here is that even though haste may appear to be generally slightly better than mastery for single target (varies by the player) that Mastery is so much better than haste that the small single target damage delta is more than made up for in the AoE gain. It's a point of view I respect.

[Gendori]

I really cannot understand why there are some shamans sacrificing so much intellect for secondary stats. Even in your own Simc weights int is stronger. Your not only hurting yourself single target dps wise but completely gimping your fire elemental by stacking mastery gems. Me and you Gendori have the same ilvl yet I have 4100 more spellpower. I really have to question what you are doing when your own character is optimized around CL spam only and not around single target dps.

My stat weights are showing that 2 Mastery >=1 Point of Int In Single Target situations

In any cleave and AoE situation Im showing 2 Mastery >>>> 1 INT

Im not really losing anything by regemming 2 Mastery for one INT In single target, but am realizing a decent gain in AoE/cleave situations, therefore since Im not losing anything, I redid all my gems.

Note that Its really only when you get up to 520ish in item level that the stat weights start to look like this, if youre in T14 gear or LFR Gear your weights are probably still 1 INT>2 Haste,Mastery

[Deleted]

Would just want to add that secondary stats plays "very little impact" on single target dps can still mean alot. The difference between 300 food and 275 food is very very slim. People tend to use 300 in high progession anyways, to get that little extra edge.

I've had many sub 1% wipes on heroics and yeah, those are really annoying.
There is also alot of different factors to check, your trinkets mainly.

Nice found nontheless. Don't see it being that useful on any encounter beside Durumu HC/Council HC but will be nice for getting those pesky ranks on bosses like Horridon and tortos. ^^

[Gendori]

Would just want to add that secondary stats plays "very little impact" on single target dps can still mean alot. The difference between 300 food and 275 food is very very slim. People tend to use 300 in high progession anyways, to get that little extra edge.

I've had many sub 1% wipes on heroics and yeah, those are really annoying.
There is also alot of different factors to check, your trinkets mainly.

Nice found nontheless. Don't see it being that useful on any encounter beside Durumu HC/Council HC but will be nice for getting those pesky ranks on bosses like Horridon and tortos. ^^

I get what youre saying, but if you really care that much, you could simply reforge every fight. 99.9% of people arent THAT OCD, so we come up with solutions that work for as much as possible. Switching talents and glyphs for every fight is about as far as Im willing to go unless were really really stuck on something then I might reforge for whatever were stuck on

[Michelangelo]

Not sure if this was resolved yet.

Gendori, your work is specific to your character, but you're using language that equates your Elemental having a lesser return from Haste at 6050 rating compared to Mastery as ALL Elementals have a lesser return from Haste at exactly 6050. What Bink is pointing out is that the value for which Haste roughly = Mastery is different for each Elemental. We very too much with our current gear that there the "magic numbers" for DPS perfection always change, compared to a Shadow Priest that aims for specific breakpoints regardless of gear.

[Gendori]

Not sure if this was resolved yet.

Gendori, your work is specific to your character, but you're using language that equates your Elemental having a lesser return from Haste at 6050 rating compared to Mastery as ALL Elementals have a lesser return from Haste at exactly 6050. What Bink is pointing out is that the value for which Haste roughly = Mastery is different for each Elemental. We very too much with our current gear that there the "magic numbers" for DPS perfection always change, compared to a Shadow Priest that aims for specific breakpoints regardless of gear.

Did you read the whole thread? Ive acknoweleged this alrwady. In fact the most recent plot I did told me it was around 7000 haste where reforging more forsaking mastery had diminishing returns.

The point is that there is going to be a certain point for everyone where reforging more haste forsaking mastery is not going to be any meaningful increase in dps. It is at that point that you must decide if adding more is going to be worth it, knowing that haste is by far your worst stat when youre adding targets to an encounter. It all depends on what youte trying to do, what gear level youre at, and what your guild is working on.

The only way to know for sure is to sim out your plots yourself and find that point

[binkenstein]

Ok, graph time.

First up, Shadow Priests.

This is the 8085 breakpoint. The red line is the gain between datapoints (which are 10 rating apart), while the blue line is the stat weight as SimCraft would calculate it (+1000 of a stat). You can see the big plateau there on the blue line, which corresponds to a massive spike on the red. This is a "classic" breakpoint, although you should note how the haste weight before this jump is roughly the same as after it.

Affliction is next

You can see that this graph cuts out at +/- 4000. This is because I'm going off old T14 reports that I ran, and I didn't do a large enough range for this dataset. In any case, you can see how there are more plateaus here, but they are more chaotic in nature.

Lastly, Elemental

Again, an old dataset so the datapoints are 50 apart rather than 10. This was using a profile with 12056 haste rating, so it covers from 6056 to 18056, and you can see that there's a lot of variation from point to point, but there aren't any of the plateaus we've seen for the previous two cases.

It's for this reason that I state there are no haste breakpoints.

Equivalence points, where Haste = Mastery, one is higher than the other above, and vice versa below, are dependant on how much haste you have and how much gear you have. This comes back to the point I've been making for a while now where balancing stats & understanding how weights work, rather than blindly focusing on one particular stat, is the best way to go.

Lastly, optimising for cleave/AoE in a minor way is fine, but I would use multi-target sims to influence your gearing approach.

[Gendori]

Ok, graph time.


Again, an old dataset so the datapoints are 50 apart rather than 10. This was using a profile with 12056 haste rating, so it covers from 6056 to 18056, and you can see that there's a lot of variation from point to point, but there aren't any of the plateaus we've seen for the previous two cases.

It's for this reason that I state there are no haste breakpoints.

Equivalence points, where Haste = Mastery, one is higher than the other above, and vice versa below, are dependant on how much haste you have and how much gear you have. This comes back to the point I've been making for a while now where balancing stats & understanding how weights work, rather than blindly focusing on one particular stat, is the best way to go.

Lastly, optimising for cleave/AoE in a minor way is fine, but I would use multi-target sims to influence your gearing approach.

Thanks Blink,

And yes I have used Multi Target Sims to Balance my reforge around doing well in single target and Cleave/AoE

[Protoman]

Anyone have an idea what GC was talking about in a previous shaman comment where he said he was changing the cast time on LvB? I think it was in an attempt to make haste better, with stuff like blust/hero, Elem Mastery talent, and the legendary haste proc gem... as well as supposedly increasing Elem DPS.

My guess is they are increasing the cast time of LvB from 1.5 sec to 1.7-2.0 sec, which will make it so you don't "soft" cap haste quickly cause your LvB is at the 1sec gcd. This should let Elem stack haste again and not avoid having too many haste CD's like EM, Blust, and Bezerker racial plus legendary gem proc. Should be a good update but I haven't seen any patch notes that reflect this change other then the one comment by GC (on twitter I think).

[Deleted]

@Protoman
Wishful thinking :P As I recall he used "LB" cast time - so now glyph that was must-have is baked in for ele and doesn't reduce cast of Lightning Bolt by 5%. But I read that in first place as Lava Burst and was so happy they buff it ^^

[zenga]

Anyone have an idea what GC was talking about in a previous shaman comment where he said he was changing the cast time on LvB?

These are the 3 tweets I can find in the past weeks that mention ele:

https://twitter.com/Ghostcrawler/sta...92928453795841
https://twitter.com/Ghostcrawler/sta...80650617753600
https://twitter.com/Ghostcrawler/sta...58570598875136

[Protoman]

@Protoman
Wishful thinking :P As I recall he used "LB" cast time - so now glyph that was must-have is baked in for ele and doesn't reduce cast of Lightning Bolt by 5%. But I read that in first place as Lava Burst and was so happy they buff it ^^

These are the 3 tweets I can find in the past weeks that mention ele:

https://twitter.com/Ghostcrawler/sta...92928453795841
https://twitter.com/Ghostcrawler/sta...80650617753600
https://twitter.com/Ghostcrawler/sta...58570598875136

Damn....you guys are right, it's LB cast time and not LvB.....would have made alot of sense to see LvB cast time increased to 1.7sec and damage scaled accordingly to help with all those problems from haste procs/CD's and the 1sec gcd soft cap.....oh well. I like all the buffs Enh has gotten since its my main spec but I think Elem could use some buffs, not sure why GC keeps insisting they are fine.