Most incorrect results were due to incorrect input.
Post sreenshots of first pages after running.

I'm not saying any of the results are wrong, just that with the exact same settings, I get very different values with 20K hours vs. 100K hours. I ran them back to back. I can still post screenshots if you'd like though.

I don't think I understand what's meant by "more viable" here, if not "higher DPS".

"More viable than it is now."

Same Problem here. With the exact same input, i got results for AEP values not even close together. Thats with 20k+ hours used.

Well, if program is unstable to the point where it produces very different results with same settings running for 1 hour of real CPU time - it is definitely wrong as it simulates 10 years of non-stop gaming and averaged over such long time results are expected to be constant. Yes, please, post or pm screens, otherwise I do not know how to help you as what you observe should not happen. 10,000 should be enough for EP calculation and 100-1000 for dps calculation. Please remember that you do not need to find EP values if directly comparing 2 pieces/sets of armor in the sim - simply compare dps. I am satisfied with sims stability after some EP calculation improvements few changes back.

I typically will see some variance between 10k hour runs (e.g.: hit changes from 1.65 to 1.72, haste changes from 1.89 to 1.75, whatever) and do what Malan does, which is to run a few different 10k sims, throw out any huge outliers, and take the average of the remainder.

I have noticed that armor penetration is very consistent between runs (always 0.29 with my stats), as is expertise (3.08-3.09).

Armor penetration isn't changing because it will affect all your melee damage exactly the same, whether they be crits, hits, or glances. expertise will change a little because it's changing some of those dodge/parrys to hits, albeit not that many, so a smaller change.

A few pages back there was a discussion about helms with/without metagems past t5 content. While I saw an increase from my T4 helm to the Zul'jin helm, the S3 helm with the RED is considerably more dps. I have started stacking armor penetration since then, so that could be part of the difference (I'm at about 500 with the helm.) Anyways, if you have the arena points for it and can't find anything else with a meta, I'd highly recommend the S3 helm.

edit: Once raids start up again I'll post some numbers.

I am debating if I should change out my Legs for either the Season 3 Vengful Legs or the Badge Reward Legs of the Shifting Camouflage Pants. I currently have the Forest Walker Kilt Forestwalker Kilt - Items - World of Warcraft with +8 Bold Living Ruby, 2 Sovereign Nightseye's and the +40 atk +10 Crit LW Kit on them. Any ideas would be great.

i dont want that malan has to say it again, so i do:


all you need is an calculator and the first page of this thread.

Read the first page of the thread, specifically the part on simulations.

I noticed there weren't any EP values listed for T6 quality gear in the first post. I ran seven 10,000 hour simulations recently, so if we would like to use my values in any way, here they are:

My stats for the simulation were:

Attack Power - 1782
Crit % - 36.5
Hit % - 17.5
Haste % - 4.63
Boss Armor Value - 2930
Armor Penetration - 336

Anyway, dunno if they're useful to anyone but myself, but there they are.

Hey, I can't decide for an enchanter, which is more valuable for PVE, +2 weapon damage or +4 stats for ring enchants.

2.6 speed weapons hasted to 2.0 mean each +2 enchant is worth 2 dps.

It would be nice if WWS would have like some option to show weapon speed over time, because DST/heroism/mongoose/LW drums increase the value of this +2 weapon damage in a raid setting.

Obviously +4 stat is superior for PVP, but in PVE, roughly +1 dps and some minor crit.

16 AP and .32% crit is worth about 8 DPS or so. Don't use the paper doll to determine how much DPS a stat gain gives you; all it tells you is how much base non-crit auto-attack DPS you gain.

Here's some info I gathered on the percentage of windfuries eaten by the main hand when wielding Syphon and Tide. For those of you with dual season three weapons it's kind of a moot point but some of us just can't get the crayon out

Weapon speed and Windfury
Theory states that, given the choice between two comparable weapons, putting the faster weapon in your main hand will result in higher damage than putting it in your off hand because (allegedly) fewer WFs will be used by the OH, which is subject to the 50% OH dmg penalty. This is because the faster weapon is more likely to attack first once the 3 sec WF cooldown is over, and will attack more often while outside the WF CD, giving it more chances to proc WF

Below is data from Servant logs recorded while using Rising Tide MH (2.6) and Syphon OH (2.8). Both weaps had mongoose, I had 5/5 flurry and also parry flurried a number of times during combat. Both weaps were passively hasted to 2.32 (Tide) and 2.5 (Syphon) from gear. With 90% flurry uptime the avg speeds became (2.32/1.3)/.9 = 1.98 for Tide, and (2.5/1.3)/.9 = 2.14 for Syphon

Here's what you'd expect for the % of WFs eaten by the MH when both weapons spend an equal amount of time hasted:
# MH swings over 10 sec/# total swings over 10 sec = (10/1.98) / [(10/2.14) + (10/1.98)]

[top] 2.14/(2.14+1.98)

51.87%

This is what occurred in testing:
MH WFs = 367/678 = 54.13%

Switching hands to put Syphon in MH and Tide in OH you'd expect the % of WFs that are MH to be:
1.98/(2.14+1.98) = 48.13%

The results were:
333/649 = 51.31%

I'm not gonna try to factor in the expected values accounting for parry flurries and mongoose procs. Theoretically, if the haste effects are distributed evenly, the faster your weapons are the higher the % of WFs eaten by the MH will be. This explains a bit of the discrepancy between theory and test, but not 2-3% worth. 2-3% isn't a big deal though.

What's most remarkable is that when the Syphon was MHed, the percent of WFs eaten by the MH was still over 50. So either there's a mechanic I don't know about, the tests were flawed (not enough data points perhaps), or the theory's bunk. If I had to guess I'd say they coded in some mechanic to queue WF procs during the cooldown and give precedent to MH WFs within that queue, so that when you come out of CD there's a higher chance a MH proc hits. That's probably just wishful thinking though. Whatever the case, OHing the slower weapon came out with the higher % of WFs eaten by the MH, which supports the theory

Data:

Yah I lost the link to the original post that discussed that stuff, the guy who did it ran some pretty extensive tests on the simulator and in the game to verify, but in the end the result was 2.x/2.y = 2.y/2.x.

Updated the T6 content EP weights with the averages of Sebudai's values.

[e] Good news for Pawn users, the latest version now has a GUI configuration so you don't need the strings anymore. Simply type /pawn and then input your EP values directly.

Considering how important they are to stat weights, I'd like to know which buffs Seb receives. I assume BoK / BoM / MotW / Relentless / Haste Pots / SoE / Battle Shout , but is he twisting (and thus including GoA)? Getting LotP?

Buff configuration the best part about EP weights from Yo's sim -- my guild typically runs with only two paladins and I'm often grouped with a feral, which decreases the relative weight of strength and agi/crit by a worthwhile amount.

In case anyone missed this, the former feral weapon skill ring from Lower City, [Shapeshifter's Signet] is now ranked (according to my personal EP values) as a Top 10 ring, better than almost every ring available in T5 content.

It's on the top twenty list for me, Malan. My EP values, due to my group makeup, undervalue agility (~1.6 EP) and as such T5 rings and even the Violet Eye ring are better. Still, that and the [Vindicator's Band of Triumph] are easy grinds for any shaman looking for guaranteed upgrades to their raid gear.

Alright, I know I should know how to do this, but I'm still unsure of exactly what level of "buffed stats" to input when I check "BoK' on Yo's simulator. Much, much earlier in the thread I remember someone saying, "Buff yourself with kings and mark of the wild and input those numbers"; is this accurate? If not, what am I missing?

Edit: also, is the "use 30% of AP as spell power" button working yet? If not, how should I handle that?

Because Yo's sim does not have STR and AGI inputs it can't know how much your AP and critical strike chance will increase with BoK. If you wish to account for BoK, check the BoK box and add STR*0.2 to AP and AGI/250 to Crit %.

I believe (hope?) Yo's sim will already account for the extra 10% increase in STR and AGI given by SoE/GoA with the BoK box checked. If it does, it probably also accounts the extra increase for MotW, and for buffs such as Roasted Clefthoof, Elixir of Major Agility, etc. But since you mention MotW specifically, if you want to get the contribution of Kings on MotW you may also need to add 4 AP and 0.01% crit. This should not be enough to significantly affect the output and I don't bother with this step.