Here there be discussion on various aspects of the EP system for Enhancement shaman.

Item 1 on the agenda - long ago our original EP values, which we now call the T5 EP, were based on a somewhat arbitrary set of gear that was supposed to represent what stats a shaman in T5 content might have on average. We've since totally lost what those stats were, and I have no idea if those stats were based on a full T5 geared out shaman, or a shaman just entering T5. We're also missing things like a value of Expertise for the T4/T5 weights.

So, should we re-do each set of the EP values? What should be the criteria for each segment - should they be determined using the best-in-slot of the previous tier, or the best-in-slot gear from the end of the zone? (ie, sunwell EP - based on best gear available in sunwell, or based on the best gear that you would arrive in sunwell?)

Here's our current EP values from the wiki - Shaman: Enhancement

Stat	T4 EP	T5 EP	T6 EP
Attack Power	1	1	1
Strength	2 (2.2)	2 (2.2)	2 (2.2)
Agility	1.74 (1.91)	1.8 (2)	1.54 (1.69)
Crit Rating	1.97	2	1.74
Hit Rating	1.34	1.4	1.69
Armor Penetration	0.22	0.28	0.35
Expertise Rating	3	3	3.18

I feel a starting T5/T6/SWP EP set would be of more value, primarily due to the fact that once you're kitted out in best in slot for a certain tier, there are no upgrades for you from that tier. So why would EP matter, as it's a system used to make gear choices? Additionally, very few people are lucky enough drop-wise to get best in slot for every item slot, so I'm not sure best of tier level X EP values would be accurate for the majority of raiding shaman.

Granted, there'll be some overlap. E.g., beginning T6 EP would be equivalent or near equivalent to end T5 EP, and so on for the other tier level changes. And if you're not going to craft your own EP values, a set of EP that's forward-looking seems the more logical set of values to use.

The obvious caveat being that you shouldn't be a lazy sumbitch and use the damn sim yourself. Kids these days, and other such cliche grumblings of old people.

The stats in itself do tell little, i think they need accompanied with settings used for generating it. Buffs, Boss armors and things like that i guess would be same for all. Thing that changes by specs would be AP, Crit, Hit and so on. Hardest part actually would be decision on weapons. I guess bare minimum within reason could be S2 weapons, altough you can get S3 also without going to kara even. Or should exempt arena weapons along with BS ones?

As of what to choose i guess full T4, T5, T6* gear as in before SSC/TK, after SSC/TK (but before MH/BT) and before Sunwell.

Right the stats that were used to generate the EP values definitely need to be recorded, thats why I'm prompting this to ask how we should go about doing it.

I think the biggest monkey wrench to deal with in this problem is badge/ZA gear. Since it's accessible to anyone even before entering T4/T5 raiding respectively, where do we include it in the gear sets used to derive general EP values? Especially now that badges can net you BT item level gear.

Edit: Perhaps a solution would be best in slot available at a given item level? If so, then we could do a ilvl 125 set, ilvl 138 set, and ilvl 151 set for entering T5, entering T6, and entering SWP respectively.

I wouldn't call it a monkey wrench perhaps, but the badge/pvp gear does cause a problem when we're trying to stratify things into neat little segments of gear. That's one of the reasons that a couple months ago I was asking if maybe we were missing some hidden relationship behind the interaction of our stats that would lead to the "grand unified EP theory."

While a GUT for EP values would be intensely awesome, I think an iLvl cutoff for EP/gear sets would probably be more feasible

I think this is the most reasonable way to approach it at this point. It's certainly the simplest way to filter the gear when trying to assemble "best" gear setups at each raid tier.

On the topic of how to define the tiers: the general challenge is that there are two competing objectives: first, figuring out what's best in slot, and second, figuring out how to get there. To answer the first question, one needs to score stuff based on best possible gear; to answer the second, you want to score based on current gear so you know what your largest available upgrades are. Depending on the exact gear in question, these weightings may be pretty similar... or not.

Now, my experience in Sunwell is that for most classes, the upgrades are pretty clear. There are exceptions, to be sure, but for the most part it's pretty obvious what the best item is going to be; thus, since the answer to the first question is easy, I might weight to make the second is easy as possible. The problem being, of course, that everyone has different current gear. But I think it's safe to say that at the current time, the most common instance of a "T6 shaman" is one that has finished BT and farmed it a bit, but has very few if any pieces of Sunwell gear. Thus, I would probably base T6 rankings off end-BT gear with maybe a piece or two of early Sunwell loot (in particular, with 3 T6 tokens dropping off each of the first 3 bosses, it generally doesn't take too long for people to get one or two of the Sunwell T6 pieces).

The other problem you're going to run into is that the weighting of stats (at least for rogues, and I assume shaman are similar) tends to be fairly sensitive to exactly what stats you have at current. If you already have a lot of Armor Penetration gear, ArPen tends to be relatively more valuable; if you have a lot of Str/AP, that tends to make all the other stats (crit, hit, haste, ArPen) relatively more valuable. This is the reason for spreadsheets, of course - to figure it out specific to where in the parameter space you're currently located. Since shaman mechanics have so far defied spreadsheet-ification, you need to be at least a little bit careful about how you define your gear sets - rather than linking them specifically to a set of items, you might be better off using some base set of stats that has a quantity of each stat that's intermediate in the typical ranges of that stat for the relevant gear level. For instance, if a shaman at a given gear level has reasonable gearing options that put him anywhere between 0 and 500 ArPen... I might pick 250, even in no real set of gear actually gives that amount.

As touched on in the OP, I think it makes more sense to call this gear level "Sunwell Shaman" than T6. To me at least, it makes more sense that the average "T6 Shaman" weights reflect those of shaman currently progressing through BT/MH. In other words, the name of the ballpark weights should be representative of the content such a player is currently tackling.

If you're looking to pinpoint the most players at a certain gear level, your best option is to stop at the bottlenecks: by definition, you'll find the highest concentrations of people at bottleneck points, waiting to have the breakthrough.

These bottlenecks tend to be:
- 25-man raiding (either immediately before or after Gruul, after Kara/before Mag)
- pre-Vashj/Kael
- 4/5+5/9 T6 (not as difficult to break through as the first 2)
- farm T6 (big break in content)
- farm Sunwell (eventually)

I think using the "end set" you'd get from a particular raiding level would be best, so you know what you have to work towards.

However, I think certain assumptions need to be made (and I will most likely use these for Elemental as well).

Pre-T4: No badge gear, no raid gear, no heroic gear.
T4: Kara/Gruul/Mag, badge gear that costs less than 50 badges, heroics, primal nether craftables
T5: TK/SSC, badge gear that costs less than 75 badges, ZA, nether vortex craftables, plus T4
T6: MH/BT, all badge gear, heart of darkness craftables, plus T5
SW: Sunwell gear, sunmote craftables plus T6

BoE recipes for BoP items would be grouped into the instance they drop in.

I think this is a fairly good breakdown based on skill level and item accessibility. It's true that a T4 level player could obtain badge gear that costs more than 50 badges, but the number of badges that player earns per week would be alot lower than what a T5 or T6 player would earn, especially now that all raid bosses drop badges.

Setting aside the question of whether such a "grand unified theory" is even possible, I really don't think that such an undertaking would be useful, at least not if we choose to separate it by tier level. As several other posters have mentioned, not only do people get stuck at different points along the content spectrum, but there is no guarantee that the best items from any given tier of content will even drop; and if they do, guilds have varying numbers of raiding rogues, hunters, and dps warriors, along with different loot assignment policies that could really screw with who is able to get what drops in a given time frame.

A far more valuable endeavor would be to come up with a standard gear level independent from any Tier heirarchy (i.e. 1500 AP, 27% crit, 300 Armor Ignore, 2.5% haste, 15% hit including talents, dualing wielding S2 weapons), hold all but one variable constant, and see how increases and decreases in that variable affect DPS and EP values. Further tests could increase or decrease this base gear level and/or greatly stack a particular variable to see how it affects the overall weighting.

I mention this because I ran multiple simulations (3x at 10,000 hours on one day, and 15x at 5,000 hours the next) with my current gearset, and I came up with EP values that are significantly different from what the community would consider accepted standard values:

AP=1
Crit=2.21
Str=2.2
haste=2.19
Agi=1.95
Hit=1.89
ArmIg=0.33

Perhaps I just have far too much AP and Armor Ignore compared to what most shaman with my crit/hit/haste levels "should" have, but any arbitrary EP weight we come up with seems to be just that -- arbitrary, and useful only under a very specific set of cirumstances. The conclusion is that there is no substitute for just running the simulator yourself and getting personalized weights.

I've been getting pretty similar values to that as well Philondra. Although my agi numbers are much closer to crit cause of kings. But basically all my values are generally in the 1.85-2.15 range.

That doesn't make any sense, thats what the sim already does when you tell it to generate the EP values for you. And honestly you really don't have 'more AP/Haste than most shaman', I was running with 8% haste at one point, and there's definitely shaman out there with more unbuffed AP. If you take intermediate upgrades in T6 content you trade out 5-6% crit for haste. Based on your crit weight being so high (with a reasonabl crit % on your armory) I'd say you're running with something non-standard on your buffs/procs, or that you put in a bad input and didn't notice it.

Please reread my post. Nowhere did I say that I had more AP/haste than most shaman -- I may have phrased the sentence poorly, but I said that I might have higher levels of AP/Armor Ignore than would generally be expected from a Shaman with my levels of Haste/Hit/Crit -- in other words, my numbers may just be screwy because I have inadvertently stacked AP and armor penetration compared to similarly geared shaman. (My armory isn't showing [Dory's Embrace], which brings me up to just under 700 passive armor penetration.)

My point is that even giving general guidelines such as "t6 content has the following EP values for each stat" is really a tremendous waste of time, as people can have wildly different values based on what kinds of gear they have been able to receive, such that any calculation will always come with the footnote "Go to the simulator and run it yourself for accurate EP values." What's the point of spending all that time coming up with EP ratings for different levels of gear if we just end up telling everyone "Go calculate it yourself, because the values may be off?"

What I was suggesting, even though I may not have been completely clear, is that a more useful set of guidelines would be answering questions such as "If my gear is well balanced, which stats will give me the most benefit? If my gear is AP heavy, what gives me the best returns? If my gear has both comparatively low haste rating and crit rating, what is the best way to increase my dps?" Whether or not it is worth the time to come up with the answer to such questions is unclear (my gut feeling is that it isn't.)

I really don't care enough to argue semantics with you.

Generating a common set of values isn't a 'tremendous' waste of of time. As I've shown before, even given wildly different values of EP weights I can produce nearly the exact same ordering of gear in every slot, across the board. Not just 'best in slot', but even among the intermediates, with only 1-2 items per slot really in different positions. The point of that is that a provided set of values will still lead someone to the right gear in the end. Can someone run the sim and find that their Agility weight is 1.54 when the standard set says its 1.48? Yes. Will it change the item rankings? Very very little.

Deleted because this conversation isn't going anywhere.

I think one of the biggest factors is as soon as you open up the 'heroic geared' then you'll see Expertise drop down, as everyone should work their asses off to get a SoC ASAP. At the very minimum, it should be assumed for the T5 and above EP Values.

I think another factor will be if you're going to spend your badges on a single piece of gear, what should it be? If you don't PvP, I can't think of a better way to go than the MH Fist weapon from badges. Sure, it's 105 badges, but you'll be hard pressed to EVER replace that MH weapon.

Chiming in with the approach I use for Roguecraft...to do, for example, "tier 4" EP, I assume the best-in-slot items a rogue can choose from tier 4 content (Karazhan, Gruul, Magtheridon) and from items roughly equivalent to tier 4 (pre-2.3 badge rewards, season 1 arena). 2.3 badge rewards are included with T5 gear, 2.4 badge rewards with T6 gear.

Now the obvious point is: "but the 2.4 badge rewards can be purchased even if you're in Kara!" That's true, but the EP weights will usually be determined by the overall quality of your gear. Thus, you should include qualifications that the EP weights that apply to you are not determined by your current level of progression, but rather the average level of your gear. A person with a whole crapload of Kara gear and one 2.4 badge piece can use T4 EP weights just fine, be it a shaman or rogue.

How Yo's sim calculate haste ep?
If I use only drums without passive haste Ep value go down to 1.3 from 1.85 but if I calculate it manually using 5% haste(4/4 drums) and then raising haste% I get 1.85 again.

[(Dps gain) / (one ep dps value)] / haste rating

When using +5% haste.
(46 / 0.3226) / 78 = 1.82

This is not same problem that Sebudai have. No haste pots, no passive haste, no haste procs.

You calculated benefit of going form 0 to 5% not beyond 5%, try checking what will happen with your settings with 5% haste in place and increasing it to 6% and beyond. It is possible that you hit the same wall as Sebudai did - the point where further increasing haste brings greatly reduced benefit but without the gear that brought those amounts of haste and crit (pushed % of strikes hasted by flurry to 90%+) you would get EP values suggesting back to this gear. I guess this point is the speculated sub 1.5 haste where you are guaranteed to miss 4 windfurries and wait additional 1.4 sec after WF cooldown for the chance to proc it instead of missing just 2 WF and having a chance to proc it immediatly after WF cooldown. It is similar to what people start to observe while reaching expertise limit (3xT6 + shard): EP value for further increasing EP becomes 0 and those items drop down in lootrank not reflecting great use of this stat in place already. The way to deal with it - use earlier EP weights for used portions of expertise or (better but takes time) check gear combos directly against each other in the sim.
Here is another reproducable point where haste drops:
the best available gear set
(all buffs on except for imp. faeri fire, trueshot aura and ferocius inspiration)
Corresponding lootrank link
Changing Band of Ruinous Delight to Hardcore Ring brings just enough haste to produce
Link with haste EP = 0.7