dr_AllCOM3

A quick glance at the todo sheet would have solved your problem.

 

dr_AllCOM3

Even a simple mainhand+offhand autoattack simulation without anything else results in very streaky dps (<1% but still).
If someone really wants to help, look at my flurry post 1-2 pages ago.

 

Wravc

Sorry if this is a dumb question, but I'm using both of the spreadsheets, the one usable with Open Office for at home, and using the one that works with excel while at work and I get a couple different things. For instance, on the simple one that works with OO, if I change out the stats and put in the info for putting Talon of Azshara into my OH(i use Dragonstrike) in replacement of my s2 fast OH, I get somewhere around a 7dps loss. But if I use the excel spreadsheet with all the drop down menus and such, I get a 16dps upgrade. My question is, why am I seeing that big of a difference between the two and which has the better chance of being correct.

 

Grim13

At first I was like, wtf, that doesn't match what I have, but then I looked back through my changelogs and saw that I had already adressed this issue in the 01/04/08 release, so if you're using my modded version of the sheet, you need not worry, this issue has already been resolved. BTW, according to my numbers, mongoose and potency are about equal on the off-hand weapon, with mongoose usually showing around 0.5 more DPS. Not 0.5%, but 0.5. In the case of my gear, it is only 0.3 DPS, resulting in a 0.02% increase. I'll be sticking with potency OH.

 

Grim13

Non-scientifically, we have:

Talon
10 AP
10 Hit
15 Agi

S2 Axe
19 Crit

15 Agi = 10 crit, so we can simplify to:

Talon
10 AP
10 Hit

S2 Axe
9 Crit

10 AP + !0 Hit, looking at the SEP, is slightly more valuable than 9 crit. Not by much, but the fact that the talon is 2.7 speed ups the DPS from more whirlwind dmg as well. (Compare s2 fast OH vs s2 slow OH; slow results in more DPS, though there are some potential issues with same speed weapons that I need to look into) All things considered, I think the Talon should be more DPS than the fast s2 1h. Specifically, assuming DS MH, my numbers are:

s2 fast: 1104.7
s2 slow: 1118.5
talon: 1121.2

Note, I used swords to get these numbers so that any racial expertise wouldn't be a factor. Ordinarily, I'd be using an axe to get the orc racial, and about 3 dps or so from the OH.

 

Grim13

Feel like a whore making three posts in a row, but whatever. This one is for the solicitation of ideas; Specifically, I am looking at the way execute is modeled for DW. 1.5 seconds is being assumed, but I don't think that is always the case in reality. For example, when using two 2.6 spd. weapons. Or, even trickier, two weapons that are sometimes hitting very close together and sometimes not. Or the possibility that on OH hit may not generate enough rage. It's got me thinking, but I've not worked much of it out yet, other than matching speed weapons, which is fairly easy to deal with. Any ideas, comments, etc would be appreciated.

 

dr_AllCOM3

It's (MH+OH)/2 in the sheet.


@Offhand enchants
I'm using 2x Executioner right now. UptimeMeter shows 60-65% uptime and that makes it better than mongoose, if I'm right (~+10 dps).

 

Grim13

Ahhh, I am still using an older version for my modded version. That must have been one of the "and other things" you mentioned. I did a little modeling and was looking at trying MH spd - ((MH spd - OH spd) * 52.5%), which is pretty close to (MH+OH)/2. I was also looking at (MH+1.5)/2, since the modeling I did is also seeming to indicate that the speed difference between MH and OH determines how quickly the speed difference for actual swings shifts over a period of time, but does not affect the average speed difference over a reasonably long timed sequence. Still have more work to do to figure that out for certain though.

 

Shha

Well that is also because your version of spreadsheet seems to value STR over crit much more. At my level of gear the new dr_ALLCOM3 spreadsheet values 10 STR at 9.1 dps and 10 crit at 9.8dps. Your's goes with 10 Str at 10.7 dps and 10 crit at 9.9. Thats actually a significant change, possibly affecting how we should gem (I tend to value crit more from personal experience and I lean towards new spreadsheet values). Either way thats why Mongoose seems less valuable for you . Since I stick with dr_ALLCOM3 values its significantly better then Potency for me.

 

Grim13

Hmmm, I just ran the numbers on the latest version of both with my T5ish gear and got 10 STR = 6.5 DPS, 10 crit = 6.2 DPS on mine and 10 str = 6.6 DPS, 1- crit = 6.2 DPS on dr_allcom3's.

Keep scaling in mind. Crit scales better than Str. If you're at a gear level beyond mine, or geared out for high AP then crit will be better, and mongoose will begin to outperform Potency. I failed to mention this consideration in my previous post on the subject. The two seem to break even on the spreadsheet around 3150 buffed AP, with mongoose better above, and potency better below. I sit at 3320 in my DW gear, hence the small benefit offered by mongoose.

 

dr_AllCOM3

I don't know, if the Str field is 100% correct in my sheet.
Crit should almost always be worth more, since you get so much AP from buffs.

 

Liar

The Devastate bug (see posts 451+) seems to be back in Grim's newest build. Hit is valued higher than crit even past 11%.

 

Grim13

I checked it out, adding 10 STR upped AP by 26.62 in the sheet. Long-hand, it's:

10.0 STR
*1.1 Kings
------------
11.0 STR
*2.0 Conversion Rate
------------
22.0 AP
*1.1 Imp. 'zerker Stance
------------
24.2 AP
*1.1 Unleashed Rage
-------------
26.62 AP


Looks good to me, though I think I am gonna write up a line by like for the stats, kinda like it was done in the old version, to make things easier to track.

 

Grim13

Took a quick look. There is a problem with prot, probably in devastate, but not to do with crit factor. I'll check it out tomorrow.

 

Shha

Well the str difference might be also because im really spoiled in my raid (well all melee are).
I usually run with LotP, enh shaman totem twisting imp agi, ret paladin, imp faerie fire (boomkin). This obviously boosts strength. Ill try running more tests in spreadsheet when I come home, and check how it behaves with different buffs.


Edit: Just started with grim's spreadsheet , and came to very strange results with SEP.

1) With ALL the buffs checked suddenly SEP for str and crit equaled at 10.7.
2) I looked over buffs and noticed that i turned "improved" FF on. Upon turning it off the crit value dropped drastically to 9.8. It made me think that maybe spreadsheet thinks im close to crit cap for white attacks. While my crit full raid buffed is massive (I can pass 50% crit rate with stacked group/major agility), still it shouldnt be possible to reach crit cap. However further testing showed that it probably has to be a bug somewhere in the spreadsheet. There were "magical" numbers of crit and hit when crit suddenly spiked in value (381 crit 200 hit rating showed crit valued 1.0 more then 380 crit and 200 hit + same other stats). Either way something in SEP in grim's spreadsheet bugs out at certain values. Ill keep testing though

 

Buanna

In Grim's sheet, everytime I stick two weapons in with any non-Bloodthirst spec, it picks my cycle to be WW>Dev>HS even when I have zero points in Protection, for example, any Motal Strike build.

Boni and Bonii are not words.
Bonuses is.

 

Grim13

I parsed out some theoretical data on executes and came to some interesting conclusions. Foremost, it looks like faster OH doesn't really mean more executes. A lot of what I'll refer to as "off-hand executes" are eaten up due to occurring near the same time as the MH swing. Plus, after looking over off-hand damage, it doesn't look like anything but an OH crit will generate enough rage to use execute.

The result, to sum up, after I factored all of this in, is that OH weapons of different speeds will not result in very different execute rates, so long as the OH and MH are not the same speed. If they are the same speed you take a big hit in execute DPS. I parsed a range of offsets from 0.7 seconds faster than the MH up to 0.1 seconds slower than the MH, in increments of 0.1 seconds, over a period of ~30 seconds. After looking it over, I worked out a multi-stage formula that takes what I mentioned at the top of this post into account, and ends up fitting very well with the data.

The end result was that the "synthetic" execute rates ranged from about 1,9 for the fastest to about 2.0 for the slowest. It seems the single biggest factor is whether or not the two weapons are the same speed or not. A bigger speed difference will "average out" faster than a slower one, but it shouldn't be a problem for boss fights, as even a 2.7 speed OH will average out in around 30 seconds.

As far as which is better, it isn't a big difference. I ran some numbers comparing s2 fast vs s2 slow in the OH, and the s2 slow was 0.8% more DPS. To get to the guts of it, the s2 slow was 17 more DPS in the 3xBT+2xWW cycle and the s2 fast was 19 more DPS in the execute cycle. Since you're only executing 20% of the time, the extra OH WW damage was better over the full duration of the fight. Using same speed weapons drops you execute DPS by well over 100, and overall DPS by around 20.

In addition to this, I'm working on a lot more features and tweaks. Some of them are:

"Fix It" macro now corrects most, if not all, errors in Warrior_Sim.xls
Sep now goes to 3 significant digits and is computed slightly differently
Added "normalization" column to SEP table so you can compare SEP values as a percentage
PvP/PvE mode now selected on the individual gear sheet
Corrected another devastate bug
etc....

 

Grim13

No MS cycles for DW was ever done. The choices are between the 3 BT cycles and devastate I'll look into adding MS and 31-less cycles.


**edit: Added a WW > HS cycle and a 5x MS, 3x WW cycle, but don't feel like screwing around with imp. MS or imp. WW, so too bad if you wanted to use them with a DW MS build, too bad.

 

dr_AllCOM3

I think, two weapons of the same speed ending up at 1.5-2.0 speed are the best for execute. They should give you 30 rage, so you can choose between BT/Execute. It would be the best use of the GCD.
Like you said, in the end it isn't that much of a difference. Same as slow/fast weapons for WW.


edit: fixed mongoose in offhand, see first post 2.08

 

Gruntle

Exactly what I was thinking. Would be nice if you could run a test with matched 1.5 sec weapons Grim13.

 

Grim13

Now this is fucking interesting. The results I'm getting so far. I need to get a bigger sample of weapon combos tested, but thus far, the results seem like they may be indicating a solid case for twin fast weapons. I'ss post numbers in an hour or two.

 

dr_AllCOM3

The ghosts behind the Ogre King in DM are best for testing dps. They never turn around.


I have autoattack+flurry now in my addon. No way to tell if it really works correctly, but I think it does.

 

Machinator

Maybe I am misunderstanding, but do people not normally switch to fast weapons to execute? The goal is to get 13-15+ rage every 1.5s. I didnt think about the matched speed part before, but it makes sense. With matched speeds you will get 2 or 4 swings/global. With different speeds it would be less spikey.
I dont know how detailed your program is, but in addition to 1.5/1.5, 1.8/1.3-1.5 might be useful as well. 1.8 with flurry is 1.44, just under a global cooldown.

"Information is ammunition."

 

Grim13

Ok, I ran some combos and recorded the results. Not sure how much I like the result. I am still dissatisfied with the way execute is handled. I think I might try to write something up that takes rage generation into account, specificaly as it relates to having enough rage to use execute at a given time. For now though, here is what I have:

Oh, and before I forget, I am a 2h blood frenzy warrior, so can someone tell me what the current popular dual wield setup is, as far as weapon speeds? I want to focus my number-crunching a little more.







A few notes on the data...
Obviously the a, m, d, s, whatever after the number represents weapon type. I mainly added that in order to highlight the difference between dagger and not-dagger.

cDPS and eDPS represent the DPS when you are using your cycle and your DPS when using execute. The rest should be obvious, even to a retarded hippopotamus.

As to the weapons, I created testing weapons, all with the same stats, and all are exactly 100 DPS. The selection of speeds and weapon types was done by doing a Wowhead search first for main-hand and one-hand weapons, then selecting the speeds and types that had the biggest selection. I did the same thong filtering for one-hand and off-hand to generate the off-hand list.

Also, I filtered all weapons to be epics of 90 DPS or more, and to have no +int, +dmg, +heal, +def, +parry, or +dodge in order to keep caster and tank weapons out of the list.



So.... According to the current modeling, the big winners are 1.5 MH w/ 2.6 OH, 1.4 MH w/ 2.6 OH and 2.7 MH w/ 2.6 MH. An interesting trend I have noticed is that it seems like the best overall DPS comes from combos with slow off-hands, particularly when paired with a fast main-hand. I think the reasons are as follows: First, the slow OH does more whirlwind damage, increasing cycle DPS. Most of the GCDs are going to be eaten by MH fueled executes, which will often eat up whatever rage the OH had previously generated, so you don't lose many chances to execute, despite the slow OH. Plus, when it does hit, it will have a proportionally better chance of generating enough rage to fuel an execute. A fast OH will lose many of it's execute opportunities due to not generating enough rage. I want to dig deaper into this eventually. I am beginning to thik fast MH, slow OH has some real potential.

Another thing, I had forgotten that daggers use 1.7s for instants instead of 2.4, meaning they generally suck because your whirlwinds will be extra shitty. the 18d and 14d weapons, as shown on the sheet, tend to suck compared to their equivalent non-daggers.

 

Volkor

Im afraid someone has been telling you fairy tales my friend. The only time i get anywhere near topping the meter is in
MH with the pack and i can get a good SS/WW/Cleave going. Fury warriors are not great at single target dps, we can hold our own but the likleyhood of out dpsing rogues on a boss that know what they are doing is slim.

You do realise that your friends fury warrior would have to do about 13,000dps in order to get 10% of a bosses health down in 45 seconds. Absolute crap im afraid, i wish it could be true. So to answer your question yes rogues can beat him because as ive said 1. Rogues are better at single target dps and 2. You friend is mistaken.

Oh and if he is having to pull out of combat then either the tanks no good or he's forgotten to ask for BOS. Last point if he was doing anywhere near 10-13k dps as the story goes he'd have been 1 shotted after 2 seconds or sat on his arse for 50% of the fight until the tank had generated enough threat.