The point is that there is no change when i let weapon without ench or i set mongoose on OH.When i put potency there is the 13 dps change. I cant believe that there is no change with mongoose.

I noticed this too, and I think it has something to do with the "USE EXECUTE" and "DO NOT EXECUTE" feature seen in the upper left corner of Warrior_DPS.xls.

For example. When I equip Talon of the Phoenix, the box shows "USE EXECUTE" but when I equip Gladiator S3 1h axe, it changes to "DO NOT EXECUTE" and my overall DPS goes down. Clearly S3 axe is better than Talon of the Phoenix, so it's a little misleading if you don't see it. What determines whether execute is used, and how can I change it?

*edit* on that note, is there a way to tell it to use different weapons for execute range when calculating overall dps?

May want to look at WowEquip also, its character frame seems a good fit to what one would want this to do.

Moved.

IIRC, the deciding factor is your AP. At a sufficient gear level (I think this is a bit over 3k AP and ~31% crit), your normal DPS cycle out-DPSs Execute spam. As far as the overall DPS drop goes, I have no idea.

Still not clear for me, why wouldn't the use bonus from Brooch be calculated by spreadsheet? Even if so, why am I still getting bigger personal dps with Sapphire than with Berserker's Call (and it's pretty big diff - 10-ish dps)? Hell, even Tsunami is way lower than Sapphire...

ADDED:
Allright, a bit confused now, what speciffically does the choose of trinkets in trinket window do? Is it only factoring special abilities like on-use buff or proc, etc, or does it calculate whole stats of trinket already. So, if i change to Berserker's Call, I have to add this static 90ap*110% into stat window, but the buff is automatical? If SO, back to Sapphire - do I have to add 70ap*110%*125% to stats, or is it calculated already?

ADDED:
I think I found answer myself, Sapphire has it's buff calculated already, so if you put it in trinkets window, you don't have to do anything, other trinket's has only their procs/uses/specials calculated and that's why they get lower dps values than sapphire, you have to add static stats from them to your stat window manually.

Everything should be being calculated automatically. You should not need to manually enter anything. If an on use or proc trinket is equipped, Warrior_Sim.xls is informed and adds in the effective AP or whatever stat it is.


As to the guy having an issue w/ OH mongoose, I'm not sure what the issue is. When I put mongoose on or take it off I see a difference, so I don't have any idea why you are not seeing one.


To the guy asking about being able to specify different weapons for execute, I am actually already working on that, and hope to have it finished and posted today. Feature creep ftw!!!


Also, the Solarian's Sapphire bug, causing it to be counted twice, is corrected as well, and will be in today's release.


For the guy being told not to execute with a DW fury set-up, I have to ask what is your MH? Are you using the slow s3 axe for the OH? If so, and your MH is 2.6 spd as well, you most likely will be told not to execute, as execute DPS with matched slow weapons is very poor. Also, execute scales with crit, but not with AP, so as your AP increases you will eventualy reach a point where execute is less DPS that your cycle. Exactly what amount of AP it takes depends on the weapons you are using, particularly the average execute spd they produce.


Further into the execute topic, I suspect that execute does less dps than the current spreadsheets indicate, since it is currently not taking into account things like MH and OH missing or being dodged, ect. I'm working on getting it sorted right now though.

Grim13, is possible to add Executioner enchant option for off-hand to your simulator?

Ok, I want to get my thoughts on the various factors affecting execute out here and solicit other people's thoughts on the matter.

There are four primary weapon options: (F = Fast, S = Slow, MH = Mahin-hand, OH = Off-hand)

FMH/FOH

• Lowest cycle DPS
• Highest execute DPS
• Both weapon swing every GCD making executes likely
• OH hit may not generate enough rage to execute if MH misses
• OH will always swing with the MH, meaning no waiting for alternate rage in MH misses

FMH/SOH

• Low cycle DPS
• High execute DPS
• MH swings every GCD, making executes likely
• OH likely to generate enough rage for an execute if MH misses
• Have to wait for OH to swing, as it is not synced w/ MH, resulting in a wait of between 0.0 and (OHspd - MHspd)

SMH/FOH

• High cycle DPS
• Low execute DPS
• Will have an OH swing every GCD
• OH hit likely to not generate enough rage to execute if MH misses or hasn't swung yet
• Fast OH is simply not an effective backup for MH-less GCDs

SMH/SOH

• Highest cycle DPS
• Lowest execute DPS
• Will not have rage for execute much of the time, if MH and OH are the same speed
• If speeds are slightly different, swings will be staggered enough to produce rage for executes on most GCDs, making execute viable.
• OH and MH both are capable of generating enough rage for an execute, for both filling in MH-less GCDs and backing up MH misses

The spreadsheet results (Done with low (bug, fixed) AP), once again:




Some items of note:

• 2.7+2.6 is equivalent cycle DPS to 2.6+2.6, but has 68 more execute DPS even though it is a slower MH w/ the same OH. This is due to the effect of staggered swings with an OH that hits hard enough to power executes.
• 1.5+1.5 and 1.5+2.6 appear to generate the same execute DPS, but 1.5+2.6 has slightly more cycle DPS due to the harder OH whirlwind hits. the 2.6 is also more likely to be able to back up the MH in the event of an MH miss, though it is balanced out by the possibility that it may not swing when needed.
• Matched speed weapons slower than 2.0 do not work well with execute, as evidenced by 2.6+2.6.
• 1.8d is a 1.8 spd dagger. It does well for generating executes, but note the low cycle DPS. This is due to daggers using 1.7 spd for instant attacks like whirlwind, while other weapon types use 2.4 spd. Stay away from daggers unless a really really good one falls into your lap, as they are clearly not as good as their non-dagger equivalents.
• In order, the best combos indicated are FMH/SOH, SOH/SOH(non-matched). SMH/FOH and FMH/FOH, about 10 DPS behind the leaders.
• These tests were done when the spreadsheet had a bug in the AP calculation, resulting in very low AP values. With the bug corrected, the numbers would be relatively the same, but shifted slightly in favor of the load outs with better cycle DPS. My conclusions would remain the same, as the shift would likely put 2.7+2.6 and 1.5+2.6 even closer together at the top.

In terms of what weapons to go for, other that Warglaives, Swiftsteel Bludgeon + Arena 2.6 spd 1h or Wicked Edge of the Plains + Arena 2.6 spd 1h look pretty good. Arena 2.6 1h + Arena 1.5 1h is up there too, those weapons are just that good. Hell dual Arena 2.6 1h actually does pretty good without even using execute.


Here are some more quick numbers:

MH = Swiftsteel Bludgeon

• OH = Gladiator S3 1h Axe; DPS = 1436.7
• OH = Gladiator S3 1h Axe fast; DPS = 1424.2

MH = Wicked Edge of the Planes

• OH = Gladiator S3 1h Axe; DPS = 1432.4
• OH = Gladiator S3 1h Axe fast; DPS = 1423.6

MH = Gladiator S3 1h Axe

• OH = Gladiator S3 1h Axe; DPS = 1427.7
• OH = Gladiator S3 1h Axe fast; DPS = 1436.8
• OH = Swiftsteel Bludgeon; DPS = 1433.9
• OH = Talon of Azshara; DPS = 1430.7

MH = Talon of the Phoenix

• OH = Gladiator S3 1h Axe; DPS = 1423.1
• OH = Gladiator S3 1h Axe fast; DPS = 1414.3

The main thing to note is that in all cases except for when OH spd matched MH spd, overall DPS was higher when using the slower OH. Also note, the arena weapons are pretty bad ass.


My current execute Model:

DW Execute CD = 1.81

• We divide execute time by the sum of MH and OH executes and that gives us our approximate execute cooldown.
• In some cases the result may be rather similar to what one would get from (MHspd + OHspd)/2, but the key difference is that two slow weapons that are not the same speed are more properly modelled, as this way gives them a faster execute CD, while the (MH+OH)/2 method assigns them a slower CD. ex: a 2.7 spd MH and a 2.6 spd OH with the old method results in a 2.65s execute CD. The new method resultis in a 1.99s CD, which is much more accurate in this case.

Execute Time = 68.20

• First we determine how long, in seconds, the target will be below 20%

Max GCDs = 45.47

• Now we compute how many GCDs there will be in the amount of time available to execute. This is the maximum number of executes that can happen

Available GCDs = 23.05

• Based on MH speed, we subtract the number of GCDs that the MH will be swinging on

OH Crits = 30.15

• Since the OH does reduced damage, I am currently assuming that only OH crits will generate sufficient rage for an execute. This is where things can get complicated when I try to more accurately model execute. The current model does not account for things like OH damage, partial rage from MH miss/dodge, and myriad other factors. It seems to be a decent approximation though, I think.

OH Executes = 15.29

• Based on the frequency of OH crits and the "available" GCDs, we get an approximation of how many OH swings generate an execute.

Thoughts and notes:

Eventually, based on your gear, you will reach an amount of AP where your cycle DPS exceeds execute DPS. Once this point is reached, slow weapons are the clear choice.
Slow weapons are obviously the best choice for your cycle load out if your are swapping in faster weapons for the execute phase.

In reality don't most of us have a special pair of weapons to switch over to at 19% ?
Would be really good to model this, specially as I don't know which enchant is best on my MH executing dagger.

Your way of modelling the number of executes looks pretty good, maybe could compare theory with some WWS parses.

Warrior DPS (modified) 02/15/08

• A few things: I'm in a hurry, so I may have missed some bugs. I moved a lot of stuff around, and that usualy messes stuff up, though I found all of it, I hope.
• SEP deems to be acting goofy at certain points. If an SEP value is obviously messed up, try adding 1 or 2 points of that stat in the manual entries section, then compute SEP again. This seems to work most of the time. Go ahead end remove the manual entry after SEP is computed.
• Off-hand executioner has been added, and I am pretty sure I modeled it correctly using the rule of addition for probabilities, since it just refreshes, rather than stacks. The resulting DPS is kind of underwhelming. It is a great OH enchant, so long as you don't have it on your MH, and vice versa. Dual executioner looks to be a no-go.
• As always, I am open to comments, feature requests, whatever...

2/15/2008
Added individual White, Cycle and Execute DPS to "Gear" sheets
Added "refreshable non-stacking dual-wielding" 15s/1ppm uptime calculation
Added off-hand Executioner enchant
Reverted PvE target armor value to being universal, but left PvP armor and resilience setting on a per-sheet basis
Moved some things around and changed some colors to try and get things a little more organized

2/13/2008
Added warning for when max haste (including procs, flurry, etc) can result in weapon speed being faster than the minimum time required to complete a slam cycle
Added SEP column to Gems in Warrior_DPS.xls
Re-inserted distance to hit and crit cap calculations in Warrior_Sim.xls
Corrected a bug causing Blood Frenzy to not be counted
Corrected an issue where PvP mode was not setting target level to 70

2/11/2008
Corrected error where PvE armor changes in gear sheet were not taking effect
Added "SEP" column to each sheet in Warrior_Items.xls. This sheet calculates an items total SEP, based upon the current SEP of "Gear1"

• This is intended only as a way to get a quick idea of what the best current upgrade is for a given slot it may not always indicate the best item
• For example, under certain circumstances, like when near a cap such as the +hit cap, it can be misleading

Corrected a bug causing Solarian's Sapphire to be double-applied

Just curious Grim, but what did you use for the execute cost when determining whether non-crit offhand hits would be enough to execute with?

Execute can cost anywhere from 15 (normal, untalented) to 7 (talented, 2t6) rage, which is less than half the original. If an OH crit is enough for a 15 rage execute, then an OH hit is enough for a 7 rage execute. With the new Sunwell pieces, I think 2t6 is going to become quite common.

Going from WoWWiki formulas:

Rage = 15d / 4c + fs / 2

Where
d: damage amount
c: rage conversion value = 274.7 (at 70)
s: weapon speed
f: hit factor = 1.75 (offhand hit), or 3.5 (offhand crit)

so, for an offhand hit to be enough for a 7 rage execute:

7 = 15d / 1100 + 1.75s / 2

with a 1.5weapon:

7 = 15d / 1100 + 2.625 / 2
(14 - 2.625) / 2 = 15d / 1100
(14 - 2.625) * 550 = 15d
d = 11.375 * 550 / 15
d = 417 damage

With a 1.5 speed weapon, I doubt that's going to happen. How about say, 1.8 though?

d = 10.85 * 550 / 15
d = 397

How unlikely is it to get 397 damage from an offhand hit with a 1.8 speed weapon? It's a while since I've looked at the average damage of my offhand (plus I normally use a 2.6 one anyway), but is that beyond the reaches of probability with the extreme buffed APs dps warriors run with these days?

[Vanir's Left Fist of Brutality] (likely from badges in 2.4) appears to be a decent compromise for players who want a different speed on their OH from their 2.6 speed mainhand so that rage intake staggers, yet don't want to sacrifice too much cycle DPS due to crappy whirlwind damage from a fast (sub 2.0) offhand. Certainly an option for non-raiding alts whose only other option is s3 1h weapons.

Is there a way to apply the execute model to the main dps cycle to account for speed differences there?

Edit: On second thought, this model doesnt account for misses at all unless I misunderstand something.

Concerning new offhand fists... suprisingly I'm getting more dps (not much tho, around 1.5dps) from the fast one than from the slow one. Only explanation that comes to my mind is that current spreadhsheet calculates average dps for the whole fight, including execute range. If that's the case then it's totally wrong, cuz most of us have special fast "execute weps" (and those who don't should have) and counting executes with those default slow weps has to hurt our dps bad.

Speaking of execute range... isn't it best to BT>Exe at higher gear levs, rather then choose between "only execs" and "no execs, just standard cycle"?

ADDED:
And maybe BT>WW>Exe, anyway below 20% "continue normal cycle without execs" can't be best... ever... for sure, even if your specials are all stronger then execs, u will get free gcd from time to time to exec anyway, so modelling no execs at all feels bad.

PS Sry for bad english.

Same issue here, i finally got 1850 personal rating and it shows a 2 dps loss for me going from Merc 1.5 speed to Vengeful 2.6 speed in offhand, feel rather lost to be honest, but when I add a bunch of random hit rating vengeful comes out ahead...maybe because of rage issues? Not sure to be honest

At lower gear levels, where executes dmg exceeds other specials it's true, but at bt level thing's getting complicated. First, I thought I'll be best off equiping weps with most hit and crit having in mind flat execute spam - hit to ensure rage each gcd and crit cuz of counting both into whites and execs. So my actual weps for execute range are: [Tracker's Blade] with executioneer and [Swiftsteel Bludgeon] with potency (I had an option for rolling for first dropped sword from Mother but eventually passed them to rogues from obvious reasons). But if you're at lev when doing BT>Exe is better, then ap gains a bit (altho not that much, since you won't be doing BT every cd anyway). That's the point where good, deep spreadsheet simulation will get you answers. For example how good would [The Brutalizer] for execute range be (had serious dilema if to use it or not, b4 i got my winterchill dagger).

Played a bit with grim's ss and found out that best weapon to pair my mh tracker's blade for execute range is... [Boundless Agony]... and by far! (obviously because executes benefits from armor pierce, crit and expertise only and that dagger has high amount of 2 stats from these)

Some thoughts about exe range:

1. 2x slow (diff speeds ofc) gives me 40-ish more cycle dps, but 80-ish less exe dps than 2x fast. Because pretty big part of cycle dps comes from BT, and weapon speed doesn't affect that, then doing BT>Exe spam with 2 fast weps should be best. In addition you have a chance to negate one "mh miss + oh normal hit" streak, and that is after BT, where you can have few rage points left and your normal oh hit can give you just enough to exe.

2. fast/slow ain't better than fast/fast for exe dps. Seems that occasional oh normal hits, generating enough rage to exe are balanced by less crits giving execs too, because of lower speed. I tried some extreme settings, threw off all stats from weps, leaving only speed and dmg range. W/o talent and t6 bonus the difference was nothing (0.4dps), ok, thought that 15rage is too much for even a slow oh to reach on normal hits. Tried talented exe, no t6 - the same. Finally got to 7rage execs, where oh normal hits should generate an exe, but again, difference was non-existant. Thought that maybe my gear lev was still too low to make oh hit hard enough, so finally I added some big ap numbers to make oh dmg bigger. But still, almost no difference. Either the model is lacking something then or difference between fast/fast and fast/slow is really that small (example: 898.2 vs 897). But I expected some border point, where rage generation for execs would change rapidly.

ADDED:
Some numbers I found playing with ss (for my gear) showing which stats should one be seeking, when using my preferable execute range settings (2x fast, bt>exe spam):
10ap: +0.5edps (execute dps) +0.9btdps (bt dps) -> +1.4dps (overall gain for bt>exe spam)
10hit: +2.1edps 0btdps -> +2.1dps
10haste: +2.1edps 0btdps -> +2.1dps
10crit: +4.4edps +2btdps -> +6.4dps
10exp: +5.4edps +2btdps -> +7.4dps
100arp: +6.8edps +2.2btdps -> +9dps

All recent discussion and ss tests about executes will prolly make me ninja that azgalor's dagger for my exe range, even tho I never thought about him in that sense

It's a good feeling, when things finally work like they are supposed to .

Hi Grim, OH executioner seems to be bugged. I don't think dual exec can be so bad that an offhand not enchanted provides more dps

***Addition***
Found it. In my hurry Friday, I forgot to copy the OH Executioner stuff to the rest of the sheets. It currently only works in Gear4, so if you want to mess around with OH executioner, do it in Gear4.


Ok, got a good chunk of feedback on the Execute Problem, which is what I really needed. I don't have much dual-wield raid experience, being a 2h DPS warrior, so I want to make sure I understand things correctly. Anyways, to address the posts that came in over the weekend:

As to the cost of execute, in his original sheet, Doc wrote in execute cost taking talents and 2x T6 into account, so that part was already done, and is not an issue. It is also not being factored into execute frequency right now, only the damage calculations. At the moment I am simply using a generic assumption for OH rage generation. I would very much like to get it modeled more accurately, particularly that aspect, but it is going to take a bit of time to work out.


Dude, sorry but I have no clue what you mean, so I don't think I can answer you.


Yes, it is calculated as an average over the course of the whole fight. I large part of this entire thread, and much of the recent discussion is with the intent of getting things more accurate. Warriors are fairly unique in that our DPS mechanics change drastically once the target reaches 19%. It is fairly easy to calculate that into an average, but unfortunately it isn't that simple, as you have pointed out. There are so many variables and options that it is going to require some serious work to get execute phase into something I would feel confident calling an accurate representation of the facts. Really, I'm not sure how accurate it will be, even if I take it as far as I can. The way execute works, a simulator, like Doc is currently working on, would be able to do this much more accurately. I'm just gonna try and do what I can and see where it takes me.


That is actually a fair summation of the direction I am thinking of taking this. It will be fairly involved, but I'm hoping to have something by Friday maybe. Adding the new cycles will be easy. Adding support for special execute weapons will be a bit of a pain but not bad. Modeling execute in a fashion that takes all the variables into account will be a bit of a pain. I guess I'll just have to start at the beginning and work my way through to the end.


This is definitely true. Because of the current way the sheet works you will only see an average. In the most recent version, I have added a display showing white dps, cycle dps and execute dps, so you can look at those numbers to try and max out for both pre and post 19%.

Sounds good. I'll do that for the next one. I appreciate the offer.


What I am intending is to have a whole new set of cycles, just for sub-20%. The ones that come to mind are:
Exe
BT>Exe
BT>WW>Exe

Exe
4xSlam, 2xMS/BT,1xWW,1xExe
4xSlam, 2xMS/BT,2xExe
4xSlam, 4xExe

Some of those cycles may be worthless, but the point of doing all this is to make certain that all possibilities are looked at, in order to better make the best decision possible. If there are any cycle possibilities I missed, please let me know. Don't think I'll get any work done today though. It's a holiday, so I think I am going to duck out of the office early, and I don't work on this stuff while I am at home.


I modified the execute formula in the 2h section to have it's cooldown take slam into account (as though one was doing a 4xSlam, 4xExe cycle) and for it's dmg formula to not take time spent casting slam into account. The initial data is interesting. It shows execute at about 33 to 50% more, gear dependent, single-cycle DPS than WW. The first thought was damn, maybe I should do 4xSlam, 2xMS/BT, 2xExe! But after thinking on it a bit, I am not certain. Sometimes it is nice to have that extra 20 or 30 rage in reserve, to protect against starvation, but but mixing execute into a cycle would be to give up that luxury. This will require some thought.

I have also noticed that, depending on gear, execute seems to range from about -10% to 15% more rage efficient than the other instant attacks, namely MS. Not so much BT. This is telling me that if your having rage issues as a 2h warrior, then switching to 4xSlam, 4xExe may be a good move. MS is more DPS, but only if it can be used every cycle that it is supposed to be.

And of course 4xSlam, 2xMS/BT, 1xWW, 1xExe may be a good compromise, especially if rage generation is good. You're only dumping your rage reserve every 4th cycle, and adding DPS to an otherwise empty GCD.

Lots of interesting things to scope out. This is going to be a busy week. I'm off to go hit up some EVE Online right now though.



***Addition***
I just checked into it, and I'm not sure what would be causing that. I am not able to reproduce the bug. I'll keep my eye on it though.

***Addition***
Ok, now that I try it at a higher gear level, I am seeing the bug. I'll check it out.


***Addition***
Found it. In my hurry Friday, I forgot to copy the OH Executioner stuff to the rest of the sheets. It currently only works in Gear4, so if you want to mess around with OH executioner, do it in Gear4.

Two things to note:
In endgame equipment, I saw highest Execute DPS with both Warglaives shortly followed by [Boundless Agony] and either slow or fast S3 Offhand based on the spreadsheet.
And another thing:
Your comparisons are lacking, if you really want to know what Blizzard should spend, you should really compare them depending on how much item value they take, a useful comparison would be(taking 10stat points):
10 Strength
20 AP
10 Hit
10 Crit
10 Haste
10 Expertise
70 Armor Penetration (Blizzard values 7 Armor Penetration as 1 stat point)

I'm really tired and going to bed now, if you haven't done it by tomorrow I'll recalculate your comparison based on 10 stat points for each stat and edit it in tomorrow.