http://www.wowhead.com/?spell=37239
I'm not proposing stopping what you are doing to heal for sustained periods of time, but when the T5 set bonus is an instant LHW, do you really think that overloading your DPS and having to turn off autoattack to avoid pulling aggro is a better solution than swapping your AP bracers out for something with some +heal on them? WoW is a game of inches and no amount of improvement goes unnoticed.

That's the 2-piece T5 bonus. And I would respond by saying that's another example of confused itemization on Blizzard's part. If the bonus were instead further threat reduction on melee / shock damage, it would be a much better bonus, because it would make you better at your primary raid role, which is what raid gear is supposed to do.

T5 enhancement bonuses are by far the best set bonuses we get, period. Even if you want to complain and say the 2pc doesn't increase your dps, just throwing those LHWs on yourself increases your raid performance, as you require less healing (and melee invariably takes damage).

T6 are examples of absolutely terrible set bonuses, as they do not scale at all, unlike every other set offered to every other subclass, which provide % scaling bonuses.

Just an FYI, I'm running a big test now after cleaning the code some, using equivalent item budget steps for AP, Hit, and Crit. (100AP, 3.16% hit, and 2.25% crit). I've also made the reporting a lot more robust so that if we get odd results, everyone can see exactly what isn't acting as suspected. I'm pretty sure it will need to run over the course of the weekend, so on Monday when I get back to the office I should have the actual results.

Each test is being run done with a 10 hour combat period.

UR is a 10 second buff, that gets refreshed every 2-3 swings with raid buffs. There's plenty of time to start casting a LHW and swap in 2 +heal weapons and still refresh the buff. Of course, blizzard is nerfing this by removing the offhand healing weapons, but there are still good +heal shields and offhands.

I don't think I'll ever hear someone say "Stop healing! We'd rather you stand there and do nothing while you wait for threat to build."

As far as the set revamp, I haven't seen any info on this. Is this just for t5 and t6, or is t4 getting changed as well?

PS. Disquette, you deserve a medal.

Agreed that the simulation is awesome. But I'd really like to see how relatively valuable raw haste rating is for shamans versus the item budget.

Also it would be great if its value could be compared in 2.0 with downranked windfury against 2.1 wf5/wf5.

Beyond haste rating, there's also a theory that weapons significantly slower than 1.5 (with flurry) waste time in weapon cooldown and windfury cooldown since every 1H weapon is unable to windfury for at least one swing, and thus the best WF weapon would be exactly 1.5s fully hasted. It sounds like these simulations disprove the theory and say that slower is better, period. My assumption is that since WF is a chance on hit and not guaranteed, the weapon cooldown (vs WF cooldown) effect is minimized. Any thoughts on that?

From linked thread:
New T5 stats
Chest - 36 str, 28 agi, 40 sta, 24 int, 1xred, 1xyellow, 1xblue bonus: 4 crit rating; 15 hit rating
Legs - 36 str, 28 agi, 52 sta, 27 int, 1xred, bonus: 2 agi; 17 hit rating
Gloves - 32 str, 21 agi, 30 sta, 20 int, 21 hit rating
Shoulders - 27 str, 33 sta, 18 int, 1xblue, 1xyellow, bonus: 3 crit rating; 19 crit rating 5mp5
Helm - 36 str, 28 agi, 40 sta, 19 int, 1xmeta, 1xyellow, bonus: 4agi; 17 hit rating
Changes
To chest: 6 str, 9 agi, 12 sta, 5 int, -1 hit rating, -25 damage/healing, blue socket changed to yellow
To legs: 4 str, 2 agi, 32 sta, 6 int, 17 hit rating, -20 damage/healing, socket bonus from 6 sta to 2 agi
To gloves: 3 str, 21 agi, 3 sta, 17 hit rating, -1 int, -23 crit rating, -26 damage/healing, -6mp5
To shoulders: 5 str, 18 sta, -1 int, 2 crit rating, -1mp5, red socket to blue, bonus 3 str to 3 crit
To helm: 2 str, 3 agi, 16 sta, 7 int, bonus from 4 hit rating to 4 agi, 17 hit rating, -39 damage/healing
Net change: 20 str, 35 agi, 81 sta(!), 16 int, 50 hit rating, -21 crit rating, -110 damage/healing

New T4 stats
Chest - 29 str, 25 sta, 19 int, 2 x red, 1 x yellow, bonus: 4str; 21 crit, 6mp5
Legs - 31 str, 31 agi, 46 sta, 21 int, 8mp5
Gloves - 24 str, 18 agi, 22 sta, 22 int, 16 hit, 4mp5
Shoulders - 23 str, 15 agi, 27 sta, 15 int, 1 x red, 1 x yellow, bonus: 3 hit; 11 hit
Helm - 31 str, 23 agi, 34 sta, 23 int, 1 x meta, 1 x red, bonus: 4 hit

I think there are one or two errors in there but it should still give you an idea. Personally, I'm going to be trying to pick up the Cyclone shoulders for sure now, they are fairly good as-is with the sockets, and the addition of 14 hit is all the better. We're not killing bosses that drop T5 yet so I think they would last a while. Socketed, 70 AP, 15 agi, 4 crit rating, 14 hit rating... throw on an Aldor/Scryer inscription for another 30AP, 10 crit rating... nice.

Slant, I might add haste a bit later. My priorities right now are:

1) get the big test done to either confirm or reject the initial "hit really isn't that good" result
2) work with the Perl version that Friedrich wrote
3) Haste procs (or constant haste)

I finally put a time stamp on the output of the steppings of the simulation. It turns out that it runs at about 300:1 speed. 10 hours (600 minutes) of in-game combat takes it 2 minutes to run. Needless to say, this is not very quick. I'm hoping that the Perl version will run more quickly, though I have no reason to think that it will or won't.

The other issue I have with haste, or procs in general, is that I'm not sure what the actual chance on hit or PPM of such procs are. Is there a good repository for this data? In particular, I can model a general scenario of:

Weapon W has a proc giving X haste has a Y PPM or Z chance on hit, for an expected up time of A%

However, If I start having multiple haste procs (mongoose, blackout truncheon, abacus, drakefist), I start losing sight of how I'm going to do this.

Maybe you need to clarify here what exactly you think your raid role is. Reading your posts in the past 2 pages I'd say you seem to think that you are a rogue with a group buff. Point you in the direction of the current DPS target and that's all you need eh?

Those of us who have played shaman for quite a bit longer understand that our role is fundamentally deeper than that. An enhancement shaman that refuses to stop attacking and heal when needed, or to even just once in awhile throw a spot heal, is as worthless as you claim we are.

Why worry about specific procs at all? The user can calculate the effective constant haste. Check my math here, back of a napkin scribbles incoming.

Lets say you have a 2.7s weapon that grants 19% haste for 10 seconds at 1PPM.

Each hit would have a 4.5% chance to proc, the proc would be up 10s per minute, and autoattacking alone the constant haste equivalent would be 19(10/60)=3.17%. If you assume 100% flurry uptime, you get 1.30PPM for 19(13/60)=4.12% constant haste. If you assume 100% flurry uptime and 6 stormstrike instant attacks per minute you're at 1.57PPM for the equivalent of 19(15.7/60)=4.97% constant haste. And finally, 100% uptime of 30% haste from flurry, 4.97% constant haste from the item proc, and 6 SS instants per minute, 1.62PPM for 19(16.2/60)=5.13% constant haste from the proc.

5.13% constant haste is 53.92 haste rating, which the user would plug into the simulator and simulate away.

Rather than making an unsupported assertion and telling me to "learn2play," would you care to elaborate on which current raid encounters you feel enhancement shaman spot heals on players other than themselves are either desirable or necessary? In particular, the 25-man raid game seems quite focused on DPS races where every last bit of damage counts. I can point to Magtheridon phase 1, Hydross, and Prince Malchezzar (phases 1 and 2) as examples of fights I have played where it's almost certainly much better for me to continue to DPS my target rather than attempt to heal another player, because it's critical that as much damage is done to the boss as fast as possible. If I heal on those fights it will be as you said to top myself off now and then so I can save a little healer mana. I could imagine a scenario where I know a damage spike is coming on a tank, so I could get a HW ready. But why should many stat points on my raid gear be dedicated to fulfilling a function that, if you replaced me with an elemental shaman, he'd do better?

PvP and 5mans are a different ballgame, granted - but my point is specifically about raid armor sets which are supposed to help you with your primary raid role.

See, there you go again. Why do you assume that our "primary" role is DPS? As moogle alluded to earlier, we are a support class filling in where needed. If some healing is needed, then heal. Top of a priest so that he can focus on the tank instead of pausing to heal himself, heal yourself for the same reasons, whatever. Support your group.

I didn't expect to see this kind of discussion here. Do we really need to derail with the old hybrid versatility versus raiding focus flamefest? This stuff was old back when nagafen first killed my ranger in EQ.

Here's a quick summary, and then we can move on. Raids are about overall efficiency not individual flexibility. Enhancement shaman healing is not an effective use of a raid slot.

Sorry, but you can't afford to do that on Hydross, at least not in his current incarnation. If you're dealing with 7 waves of adds, that fight requires around 9k sustained raid dps to get him and all the adds down before the enrage. You can't bring classes to the raid that aren't either top-notch DPS, tanks, or healers. Hell, you can barely even bring casters who aren't in Tier 5+ (read: tailoring gear)! If you're not doing all the dps you can possibly do on Hydross, your guild is definitely better off replacing you with a rogue. As it stands, an enhancement shaman does have a place in the raid for his dps and the buffs he brings. Not for his heals.

Edit: slant you're right, sorry. I'm done now.

Slant I am by no means advocating that an enhance shaman shows up to SSC and stands in the back and heals. Nor am I suggesting that anywhere near something like 10% of his total time should be spent healing. But to disregard the fact that our class can still throw a heal when needed to relieve pressure from someone else that is healing is no better than the guys in this thread that have brushed off the fact that close to 10% of our DPS is from elemental dmg.

I think I can use an analogy that is relevant here, considering I just got done running a test for you that directly ties in to this :-)

Attached is a screenshot of what happens when you go from a 1.9 speed weapon in each hand to a 2.0 speed weapon in each hand. When you have no other haste effects besides flurry, this crosses the magical 1.5 speed flurried barrier.

I'll let people come to their own more detailed conclusions, because I need to boogie over to a hot lady's house (yay!), but the fact that going to a slower weapon increases your overall windfury hit's at all, let alone a fairly significant 4.6%, means that we did indeed hit a breakpoint (column L). If I ran that same test again, with 2.1 speed weapons, we'd expect the # of windfuries to go down, which is intuitive (slower weapons have fewer chances to proc a windfury over a given amount of time).

Ok, so back to your question - why not just come up with an average haste rating? The more I thought about it, the more it's clear that this is what makes a simulation valuable. Let's say I have a 2.1 speed weapon, which gets flurried to 1.61 speed. Now let's say that the "average" haste increase is 6%, because of a 30% uptime on a 20% haste proc.

In the example above, I'll be swinging at 1.51 speed, which is *great* for windfuries! The reality of the situation, though, is that unhasted, I'll be swinging at 1.61, and hasted I'll be swinging at (2.1 / 1.3) / 1.2 = 1.35 speed. In this scenario, not only is the 1.61 flurried speed not as good as the 1.51 "averaged-haste" for Windfury procs, but furthermore, when the proc does happen, we start dropping another potential windfury since we dip below the 1.5 speed on the weapon.

For this reason, I'd rather model it accurately, though I'll fully admit that I don't know exactly how much difference it will make.

A quick update, after running the 2.1 speed tests, the resut was a 4.8% decrease in WF procs compared to the 2.0 speed test. Given that .1/2.0 = 5%, this is pretty reasonable to me. There's also a good reason it's not a full 5% decrease - the number of stormstrikes remains constant regardless of weaponspeeds, so that eats into the expected 5% decrease by a little bit. All in all, I'd say the 2.0 to 2.1 numbers are very sensible.

The *unreasonable* part is that the game is coded such that going from a 1.9 speed to a 2.0 speed doesn't also decrease WF procs by 5%.

So, going from a 1.9 to a 2.0 speed is a 10% WF swing increase versus a non-cooldown scenario (2.0 should be down about 5%, instead it's up 5%), and afterwards it trends back downwards as you'd expect.

Must...resist....urge to complain about state of enhancement shaman....

On a more serious note, it might be worth writing a module for the damage simulator that models PPM effects like crusader, mongoose, or the Dragonspine Trophy [I think it's a PPM effect]. I'll probably try to do that sometime next week.

You're absolutely right. Of course a true simulation would be superior, particularly because there are almost no items in the game that give constant haste. It would be an incredible amount of work to implement, and modeling average haste should be easy. But more power to you if you do it!

The sim is already proving its value showing the breakpoint. Those are some telling numbers.

On the topic of t5:

I've thought for a while, that this set bonus was the greatest thing ever, and that it really might revitalize my usefulness in a raiding situation. I already have a full mana bar at all times, because shocks are just too much threat - so burning it on heals sounds ideal (although heals are threat too =P).

After some testing, I can conclude that the 2-piece set bonus is crap.

It procs about as often as I feared it would, which is to say, almost never. When it does proc, you get a 15 second window to fire off that heal. No saving it for an opportune time, no saving it for after a big AE, or when someone is about to die. You pretty much use it on the first poor schmuck to take damage after the proc, and hope that person's dedicated healer doesn't already have a heal 1/2 way cast.

The proc is dumb. If they want to leave it being 1ppm, that's fine, whatever. The buff should at least last indefinitely (like natures swiftness), and IMO, it should stack up to 5 charges.

Did they change it? When I tested it on PTR I was amazed at how often it proc'd, probably once per mob that I was killing.

Also, what crazy set bonus off another class lets you "save" a proc? You use it when you get or too bad. That's how it works.

I'm sorry. I didn't realize that all mechanics were measured only by precedents. Relativism aside, the set bonus still doesn't measure up - to my hopes at least. A small heal window means that more often than not, my heal hits someone who (a) didn't desperately need it, or (b) already had a heal headed his direction.