I really need to do a combat log style output. Sometimes I discover horrible errors . Rend for example costs 10 rage when it ticks and not when it's used.

DPSSim updated

Added: Glyphs, Bladestorm (+few rotations using it), rotation info.

If anyone knows Java and wants to add new rotation feel free to send sources - I will add them to package for others to compare.

I can't get it to work at all for TG Tankietka, whenever I try to use a 2h weapon in the offhand slot the simulations just hangup when I start it (for a test period of 100 sec). If I set a 1h weapon it seems to work. What version of java is required to run it (perhaps there is a too old version in my work computer)?

Otherwise it looks very nice. I like being able to use it independently (something fun to fiddle around with when bored at work).

Hmm, I don't have any problems with two-handed offhand and Titan's Grip on java 1.5 :-/.

Can you use [Set current settings as defaults] from [Options] tab and send me (pm should be enough) the content of "defaults.txt" file from directory where you saved DPSSim.jar? Maybe some specific combination of talents/buffs and stats are giving problems.

Also which rotation are you using?

DPSSim updated with build 9095 changes:

* Rend base damage has been slightly lowered, additional damage changed - originally took 15% of base weapon damage, now takes 50%, originally took 2.14% from weapon attack power damage, now takes 7.14%.
* Sudden Death now has a 3/6/9% chance to proc on melee hits. (Old - 10/20/30% chance to proc on critical hits)
* Improved Rend now increases the bleed damage done by Rend by 10/20%. (Down from 25/50%)
* Titan's Grip now reduces your chance to hit with damage-dealing abilities that require a weapon by 12%. (Down from 15%)
* Bloodthirst cooldown has been changed to 5 seconds. (Down from 6 seconds)

Dr: Any chance you can add the Hourglass of the Unraveller? I didn't see that as a trinket option. Thanks for your work on this addon. I love the ability to make changes in game and see the effects.
Thanks,

Seems like a combat log was fairly easy to implement.
I'm also making the code way easier to read, so the theorycrafters of you can check values. It is easier than a spreadsheet . I think I'll continue it (I'm rerolling a DK and I make a module for DK's too).

I'm ditching all the lv70 stuff now. There's no real reason to put any more work into it.
The Hourglass is the best blue trinket.

I tried it a few month ago. It takes at least 15min, 30 are more likely. When comparing two items 1dps is insignificant and most of the time you see the difference after 15s.
SEP are interesting, but only possible with a spreadsheet (or a very specialized sim with multicore support under Windows).

How do you see the difference after 15s if they take at least 15m to converge? My point is if you cant get the accuracy for even basic SEP, what does the sim actually tell you?

An error margin of 1 dps is reached very quickly, but one stable digit after the point takes along time (and I'd need that for SEP).

Dr, I would try allowing the sim display more numbers to the right of the decimal place. While observing the simulator activity of the final dps number in the last 4 digits to the right of the decimal place, and if activity stops changing in the tens and hundredths place, you might be able to find a stopping point to make an accurate stat comparison, but if it takes 15+ minutes to do that one time, then the objective becomes kind of moot.

What do you think?

It would take at least 10mins per stat. You'd need a multicore windows simulation that can run in the background for tasks like this (and I don't know a suitable language well enough).

I've updated my sim, too. You've maybe noticed the empty lists sometimes. They show up now.

One question Tankietka, where is the difference between the Death Wish dynamic buff with normal fury rotation and the Death Wish rotation without the dynamic buff? Both seem to produce similar results.

I like your sim, by the way.

Why do you need an error margin that specific for general purpose SEP? Wouldn't simulating +10 str for 30s or so, +10 agi for 30s or so, +10 etc. for 30s or so, and then dividing the dps gain of each stat against Strength's give you a fairly good approximation? If you reach an error margin of 1 dps quickly, using 10x(or 20x or 30x, w/e) of any given stat should amplify the effect enough to make stabilizing at 1 dps differences sufficient.

It should give you a good guess, like e.g. hit is the best stats to aim for me now and not ArP.
I can try it, but what are SEP good for if you could just test the item in question for itself?

Thanks for bringing this up - DeathWish dynamic buff was one of the earlier features for testing mainly. It does not check whether there is enough rage to use it for example. I will remove it in next version as usage of any rage consuming ability should be incorporated in rotation.

Thank you :-)

I really like your ingame simulator. Is there any chance for you to implement a function to retrieve the equipment information from the character right there instead of manually filling it all in? That would be the real advantage I see in an ingame simulator.

It already exists, but isn't fully working.
Edit: Done, but I'll correct/update Rend+DeepWounds first before I release a new version. That one should have everything implemented working correctly.

I guess mostly just to save a bit of time. You can calculate SEP values, and then calculate the rough SEP strength of any number of items instantly to get an at-a-glance breakdown of gear choices rather than doing it piece by piece by hand.

Here's the common problem:

You have 2-3 items per slot of relative strength and each change in item changes the strength of the other items / talents.

Take for example the choices [Felfury Legplates] and [Leggings of the Immortal Night] combined with [Hard Khorium Battlefists] and [Gloves of Immortal Dusk].

Each item is on the same loot table, and depending on what you are already wearing, the choice may not be too clear. Finding the right setup wouldn't be too bad, but changing items in these slots might change the item of greatest power in other slots also. This is why SEP is very handy. You can quickly glance at the ratios to quickly determine what item would probably be best for that slot so you could test one against the other, instead of mindlessly guessing.

I hope I explained it better..

I'd love to do SEP, but as explained it's something spreadsheets are way better at. The shaman sim takes over 15mins to do SEP and it is very fast.

Updated my sim to 1.9, it's very good now.
You'll find the link and changelog in my signature.
There's even an Arms DW rotation in there, I think.