Sham

I would concur to this. The only way to effectively use EP would be to compare your stats (AP, crit, hit, haste, etc.) to those tier levels. While my character has the Sunwell crafting engineering goggles and a few pieces of Vengeful Gladiator and Vindicator PvP gear, my other gear quality is at the Heroic/Tier4 levels. I built a model using stats from the Tier4 Cyclone set along with Kara weapons, then built a Tier5 Cataclysm set with weapons and filler gear from Tier5 content. I learned that my stats came closer to Tier5 content, so until my stats fill into the Tier6 Skyshatter level of gear, I will continue to use the Tier5 EP.

I would highly suggest shaman with mix and match gear to use this way to effectively EP assess their gear. And when all else fails, using Yo!'s sim has been more than helpful in deciding gear, even for this mid range shaman.

 

Groombah

To answer the original question, I would say that it is more beneficial to use best in slot gear as criteria. At least it does feel common sense. If you have a scenario where item A is giving you marginally better benefit but item B will be better at the end of the zone, I would tend to go for the option B. Differences are not that big anyway.

At least it seems logical for me to pursue the best in slot items and base one's gear choices on best possible performance. The actual difference might be small anyway. For me, the lootrank link Yo posted works pretty well when trying to choose t6 gear and I won't bother checking my EPs after every upgrade I get. Unless I need to compare weapon performance which is pretty esoteric.

 

Pitbuller

I calculate dps 0%->5% and 5%->10%
Dps gain was almoust same(46 and 45). But Ep from haste was(0%) 1.85 and (5%) 1.3.
And without passive haste or pots swing timer was up >2s all time.

Slow, slower, shaman weapon.

 

Malan

One thing I'd like to see for ease of use in checking upgrades is for the sim to have an interface more like Rawr does. Import armory and wowhead data, allow us to choose upgrades per slot and check the differences in expected dps output.

Shitting up every single thread on EJ since '06

 

Nahela

I'd love to have something Rawr-based, or like it, especially since I have a heck of a time getting the Sim running if I'm actually in WoW, for whatever reason.

Question I have here, I recently got EquivalencePoints now that Enhancer has been split up; However the step for Armor Penetration is in increments of 0.1. Is there anyway to get this so I can input a value like 0.35? Right now I tend to just toggle it between 0.3 and 0.4 and see what I end up with.

 

Tornhoof

You can do that with my weo2 already and the error in the enh model with the new flurry modelling is well below 2%

http://code.google.com/p/wowequipoptimizer/

 

Cronax

Interesting, but is your weo2 cross-platform? On my mac I cannot run a windows executable, but anyone with a browser can run the sim making it available to pretty much all the wow players there are. As long as Yo's sim exists, as far as I care you can go ahead and ignore the people not using windows, but the percentage of people playing wow on a non-windows client will probably surprise you.



In any case, I think the most useful EP values for a "tier" of content are the values at the end. You're aiming to get all the gear you have access to, so it makes sense not to waste DKP/whatever token your guild uses on an item that is an upgrade for you at the time, but will be nullified by getting a different upgrade for a different slot (a so-called "temporary upgrade"). I agree that the chances of this actually happening in the field are slim if not nonexistent, but that slim chance is basically why we're using a simulator rather then a spreadsheet isn't it? If someone doesn't care and wants to loot as much as possible, they can just run to the simulator every time they get an upgrade.

Anyhow, whatever values are used, make sure links to the itemsets you used to obtain said values are added so there won't be a million people whining in threads everyday to say that the value that is up does not match the value they have even though they have that particular tier on farm. Not that that really needed saying, but better safe then sorry.

 

galzohar

The error in what is 2%?

Becuase when you try to compare items that are close together a 2% error in final DPS will make the items (or even items that aren't too close) impossible to distinguish. Say if I do 1500 DPS and have 2% error that's 1470-1530, so if I try changing items that are expected to change my DPS in the area of 30 DPS I will not be able to tell which item is better.

If the error in the EP is 2%, however, it's quite neglicible, since if 1 crit rating is worth 1.98 AP or 2.02 AP it doesn't really matter.


At the end making EP for the general public will never be accurate, so it doesn't really matter how carefully you pick the baseline. People that will want accurate data will have to run to the sim anyway to get their EP, and if they want really accurate data they'll have to run the sim with different item setups (to combat situations where items are so close that changing to them changes the EP in a way that makes you change the items back...). It's all a matter of how accurate you want to be, and since you'll never be accurate when making EP for people who can't use the sim, might as well just make the EP for various gear levels and post the average, with expected margin of error so people will know when items are too close to be compared with this EP, and need better inspection with either personally generated EP by using the sim themselves or running the sim with both items.

 

Tornhoof

The error is the difference between constant simulation and closed form equations. This still does not change the results at all.
If your gear does 1500 dps, and an item increases this by 10 dps, this item will be superior to another item.
It just might be that the same gear might score 1530 dps with a simulator and the item might be 12 dps better or similar. As long as you stick to one tool to measure the items the results will be consistent and correct.

The error is just the difference between both modelling tools, the simulator and weo2 in this case.

http://code.google.com/p/wowequipoptimizer/

 

Meso

That makes sense about the expertise and it was exactly what I was expecting, but the opposite happened to me tonight. I just got the T6 bracers and I had the belt since befor so I tried running the sim with T6 belt, bracers, gloves and shoulders to check the 4 set bonus. For some reason I got an expertise value around 0.9 AEP when its usually around 0.3.

Its possible that Im doing something wrong so I took screenshots of the stats I entered so you can see exactly what it looked like;

Base stats - ImageShack - Hosting :: 33973308bo1.jpg
Procs and stuff - ImageShack - Hosting :: 96524769xb4.jpg
Buffs and defuffs - ImageShack - Hosting :: 77243015cn1.jpg

Heres also a link to my armory - The World of Warcraft Armory
(I didnt have time to re gem the shoulders and gloves yet. Only had them for looks until now.)

Can anyone see if I did something wrong with the sim? I ran it 3 times and I always got an expertise value around 0.9 AEP and it just seems way too high to be right.

 

Yo!

Usually it is around 3.0 not 0.3 so 0.9 is indeed reflecting that you already hit 6.5% reduction with MH and is close to it with OH.

 

Meso

Yeah, my bad. Figures it was something really obvious. Ill stop running the sim in the middle of the night after many hours of raiding.

 

Hytoshu

Let me first start off saying that I have been a long time reader and that both me and my good friend Wun read your original post and quote it like a bible.

My take on this my seem to be a little bit different but that is just becuase after looking at gear that drops from simular instances to tear (for example Skulker's Greaves to t4 legs) show a definant improvement over tier. I was the one a while ago that asked what everyone would have thought of about EP values for sets and I was go told to run Yo's. I run Yo's constantly, that wasn't my request.

The big point is that the tier stuff is terrible except for the new items. Poor itemization in my opinion and that forces us to wear druid, rogue, hunter gear. So instead of trying to reach magic numbers of t6 gear, why don't we reach the numbers required to raid t6 quality raids.

I have to agree with most of the people in this thread actually on what they have all been saying and I'm still trying to work through most of it for retainability.

When I originally started raiding SSC/TK I was in solid Glads and I put out numbers expected of an enhance shaman. I followed the guildlines in Malan's original post and was pretty solid. I still get to raid.

I didn't however have the improved strength totem, I never have. So we had to calculate out different ep values depending on who was raiding with us for our other shaman.

So that story boils down to this. I had 31% crit and only 1240 AP which held me through. How I would recommend doing this assuming the previous is say you need x amount of EP in order to raid this instance. Assuming stat caps, consumables, solarian trinket, self buffs, mana effeciency and fights around your level it would be helpful to say that my EP is _____ and what I need is _____ therefore I'm ready/not ready for this fight. By maximizing flask/elixers and taking potions, popping drums, raiding with the right warriors, getting in with hunters...etc you can add more to your ep. Therefore it isn't soley based around your personal buffs but also other buffs within your party. Basically this would stop us from "throw it in the simulator" every time we wanted to try a different group break up.

Add those ep values to the existing system (it seems to still work fine for me) and then generate a number on a boss which you should meet.

Not sure what your take will be on this but much appreciated and hope it helps.

 

Malan

I don't see how we could possibly be asked to determine what someone's minimum stats need to be for XYZ encounter. Things just don't work like that. Every post I've ever seen by someone who says "In order to raid Gruul you need X crit and Y AP" has been pretty silly and just uses arbitrary numbers. Are you going to do 2000 dps in Black Temple wearing KZ and heroic gear? No, but that doesn't mean you can't or shouldn't go in there if your raid says you're worth the slot to buff the group.

Shitting up every single thread on EJ since '06

 

Pitbuller

YourFileHost.com - Free hosting for ALL your files S2
YourFileHost.com - Free hosting for ALL your files S2
YourFileHost.com - Free hosting for ALL your files S8
There is haste bug images.

Slow, slower, shaman weapon.

 

Skreekins

Maybe the best option would be to take a "Web 2.0" approach to the general public EP values. It would take a little work up front, but might be worth the payoff.

The basic idea would be to come up with rigid progression boundaries that would force everyone into a group. There are others more qualified to come up with the boundaries than myself, but an example might be "3 or more bosses down in TK/SSC, but not kael/vashj" or something similar. I thought that some natural gearing boundaries were posted somewhere in this thread that might work well.

Once these boundaries are set, create a simple web app wherein people indicate which of the established boundaries they fall into and then submit the EP values that Yo's sim generates with the gear they are wearing. Given enough submissions, a general trend should start to form based on the type of stats that are readily available in any given tier. A large enough sample size would also iron out kinks such as people with bad luck with the RNG, people with badge loot, ZA loot in an otherwise "T4" progression bracket, etc.

Ideally, the submission would be an option in Yo!'s sim that automatically submits the values that the sim generates. This would prevent people submitting bogus values. Also, if a person is using Yo's sim it is quite likely that they have read the shaman thinktank and are not making clear mistakes that might mess up EP values, such as gemming for +hit. Finally, if attached to Yo's sim there could be qualification boundaries set in place, such as only submissions that were generated with 5000+ hours will count, creating min/max "realistic" values for certain stats to prevent anyone purposely generating EP values with bad stats to taint the average value (i can't imagine there would be motivation to do this, but some people are douchebags, for lack of a better term). I guess to simplify things you could just have the web app interpret strings generated in the "import/export" tab of Yo's sim.

Take the average of the EP submissions for each level of progression and you have dynamic "general public" EP values that change and adapt with items and content that Blizzard adds to the game, as well as showing a realistic level of gearing rather that striving for best in slot for every slot in each tier which very few people actually reach.

Again, it would be a bit of work up front. I would do it myself if I had any programming knowledge whatsoever, but I do not. If it worked, it would be a real testament to the value of the theorycrafting community you all have created here. If it does not work and the EP values are obviously skewed and awful, the whole project could be nuked with no harm done.

Thanks for all of the work and theorycraft that has been done by this community. Shaman everywhere are grateful.

 

Hytoshu

Maybe it was a stupid idea. But I didn't mean x crit and y AP. I mean something like... this incounter would recommend you sitting at 2200 EP value so if all the stats on your gear + buffs = 2200 EP you should be fine. Either way this isn't a problem that is going to be solved because most people use the EP values for different things and unless we have a "standard" raid with "standard" buffs its not even worth running the sims on except for yourself.

The hardest thing for me to see is at what point do those EP values fall short. Does windfury not follow the AP cap? Does crit? Sitting in Hyjal yesterday I was sporting around 3200 AP and 45+% chance to crit, without mongoose. So, do the EP values change once you hit the cap? Stacking 200 Expertise doesn't sound like a good plan.

But by the same token I've always used the EP system as a guide line. If it is 20+ EP then it is a definant upgrade. If it is only 10 or less I take the EP values into consideration but at that point I see what I have too much/too little of and go from there. Why not just create a guideline EP for each tier and not make specifics?

 

Rhaegal

This is absolutely true for comparing two different pieces of gear, and most of the time when calculating upgrades on the fly I'll just use a generic set of EP values, knowing that they're close enough to the right ones for me. The real benefit to knowing exact EP values for various stats isn't so much in working out upgrades, but gems to put in them. Is a +4 Hit socket bonus worth going for, if it means you have to use a crit/str gem? Or if it's two gems? Admittedly, these types of decisions are very minor, and "gemming properly" vs. "gemming decently" in every slot might make a difference of a whole 10 EP overall, but I think most of us still care that it's just right.

Anyway, the question at hand... With BT-level badge gear available for some slots, T5-level badge gear available in other slots, timed drops in ZA that are excellent (and even onnormal--shoulders, belt, and chest come to mind, with cloak from the first timed chest easily available to people barely wearing Kara gear) , etc., I think calling the EP levels "T5/T6/SW" is relatively meaningless, at least compared to the way it was before all this became available. The line has blurred too much. Picking a set of "average ilvls" might work, but it would require people to understand ilvl and go to the trouble of finding out the ilvl on each piece of gear they have, and doing some math to figure out which gear level is closer to theirs. If the goal is for a quick, easy-to-use reference, the ilvl idea doesn't really seem to fit that. If people want to go to that much effort, they're probably amongst the same population who is happy to use the sim to generate their own--the sim is probably easier.

I do really like the idea of having sets of EP relative to what stats are stacked on your gear, however. For example, I could have made different choices with badge items or ZA drops and have a lot more Armor Penetration than I currently have. The gear is the same ilvl, the other items that I did not choose simply have different stat allocations (e.g., int/mp5 on the more huntery pieces). I've lost a fair chunk of crit over my last few upgrades due to stat allocation, but still have relatively high hit from the rogue leather, whereas going for different choices could have changed that balance as well. Those choices are certainly going to have an effect on my personal EP, at least as much as relative ilvl.

Stand back! I'm going to try SCIENCE!

 

Rhaegal

I'm confused by what you mean here - is there some sort of cap on AP where it's not as useful beyond a certain point? That seems strange, as AP should scale linearly across the board unless I'm missing something.

Stand back! I'm going to try SCIENCE!

 

• Daler

The problem I see with a flat EP sum being used as a benchmark is that some doofus could easily argue that by stacking nothing but "of Strength" drops, he hits EP values ready to raid X boss. And that's simply not the case. Such a toon could lack the crit needed to buff his group properly with UR, if spec'd poorly lacks the hit to cap his specials, etc., etc. Too many ways to game such a simplistic system that it could be rendered useless (which I think the idea is already).

Shoot, an X crit + Y AP guide would be more useful. And even then it's, as Malan noted, completely arbitrary.

EDIT: And of course EP values change as you approach the cap. Run the sim with extraordinary values for hit, crit, or expertise and you'll notice EP values approach and then hit 0 once you reach the cap. This is nothing new, and is hardly a shortcoming of EP systems. It's a shortcoming of the user.

Plus, AP cap? Yeah, there's no such thing. And the crit cap with only 9% hit to cap specials is (100-25-19-6.5) = 49.5% crit. And I seriously doubt there's a gear set out there that hits 49.5% crit without having a substantial chunk of hit on it.

 

Hytoshu

Ah, then I must be wrong. I thought that the soft cap for AP was 4k at which point AP just stops before expose weakness and windfury. I could very well be wrong though and probably am. I think I read it on a european post on arms warriors and the AP cap was something they justified going arms because slam goes above the AP cap. Once again probably wrong.

Also, I had to turn up my stats before raid buffs to see the caps get hit. Maybe Yo's could put if you are capped for such things as Crit, expertise, hit...

Also, I guess my suggestion was pretty stupid now since I read it fully awake and not as intoxicated. \

I agree completely and said it so much less elegantly than you. Thanks for putting that out. Also, Shoulder Pads of the Stranger are better then anything at t6 level and won't be replaced till sunwell to throw another number out there.

 

Rhaegal

If this cap exists and can be modeled in the sim (I would argue that if it exists, it should be predicted by the sim), you could try entering in some absurd value for AP, like 5000 (well, absurd for us anyway), and see how it effects the EP of other stats. If AP is no longer useful at that point, crit/hit/haste/etc. should have values completely off the charts. However, your reasoning for there being one (how are expose weakness and windfury related at all?) makes absolutely no sense to me.

Stand back! I'm going to try SCIENCE!

 

Toots Hepcat

49.5 - 5% (druid buff) - 3.5% (Goa) - 2.3% Crit (Major Agi) - 9.6% (double Mongoose proc) = 29% crit from gear to be at least partially crit capped when mongoose is up. Very attainable if you're completely ignoring hit and expertise.

Which is one reason I absolutely HATE the idea of a static EP. ANY static EP. The best gear will always be the best gear, but while gearing up you want to know if a given piece of gear is an immediate upgrade. It does you no good to say "Well, this upgrade would be great if I had [Mounting Vengeance]" unless you actually HAVE that weapon. The essential gear-up question 90% of shamans have can't be answered based on "tiers," because the answer is based on the relationships of buffs and numbers individual to the shaman and his raid. It can only accurately be answered by personalizing EP.

So if there is resistance to using the simulator, or questions of the stability of its output, I think our goal is simple: find the equations that can be used to generate EP so folks can build their wretched spreadsheets and pretend we're rogues or hunters or something.

BTW, there is only one thing standing in the way of us having flat equations: no equation to calculate DPS from Windfury. This can be turned around: there is no equation to calculate the average number of windfuries in a given unit of time. It can further be reduced to this: given that there is a minimum 3s between windfuries, what is the average time we will be waiting outside of that 3s for our next windfury? And how many of those windfuries will be from the main hand, vs the offhand?

Average Windfury Lag. MH to OH ratio. Find these, and you've got your damn spreadsheet.

 

◊ Shalas

Average WF lag isn't quite meaningless, but it's pretty close. If you have no haste effects other than flurry, 2.6/2.6 and 2.8/2.8 will do the exact same white + non-SS WF DPS with the same MH : OH proc ratio, but 2.8/2.8 will have a higher WF lag. Unless you change the number of swings in a 3 second period, the only effect speed has on DPS is SS's nearly linear scaling.

What is useful is the actual WF proc rate for each hand. Just brute-forcing this isn't really practical, and while the regression-fitted equations convinced me that a closed form for this should exist, I was pretty unsuccessful at finding it. If you did know the WF proc rate for each hand, you could pretty much just take a fury warrior spreadsheet, unnormalize WW and make WF apply to both hands and you'd be done.

Technically you also need the actual Flurry uptime for each hand, but the estimation that ignores flurry's interactions with WF is so close that it isn't really a problem.

 

Malan

Where have you even seen this? I have never had an item just sitting in my bank waiting for the day that my crit EP hit 1.8 so I could equip it. People keep postalizing that this is case and I've yet to see evidence of it. And if 2 items are so close in value between 2 different values of EP, I would submit that its really not that big of a deal.

We talk a lot about these hypothetical situations where someone would have to suddenly start swapping out gear because XYZ stat suddenly plummeted in value because of a recent upgrade, but seriously I've never heard of anyone actually finding this necessary.

Shitting up every single thread on EJ since '06