Which of these three point-allocations give the most damage? This is all assuming that 17/34 with Impale is the best single-target DPS build--is there anything better?
I'm pretty well set that my build (see profile) is the best one for single-target dps. The only thing that might be said is that I don't use imp OP that much on bosses so those 2 points could be spent elsewhere, but there really isn't much else to put them in and I like it for grinding/pvp. Other than that I'm pretty confident that build is the best there is. I was 12/39 for a long time and prefer that for utility/AoE damage, but 17/34 is objectively better for single-target.
Well I'm not Deathwing or Celendo but this spreadsheet was basically meant to answer these types of questions
The only thing it doesnt have clearly is a cooldown for overpower. I tried to calculate my own using the data thats in the spreadsheet already.
Gear set: my ctprofile: http://ctprofiles.net/2868856 (not the best gear but not the worst either)
wowwiki says that bosses have a dodge rate of 5.6
Used version 0.45 of the spreadsheet because the later versions seem unstable (cerc refs and dont adjust for taking consumables buffs off etc)
Basic logic was that I needed to work out all the attacks that I would do in the fight period (set at default 300 secods - minus 30 in execute phase because not overpowering (assumption))
Using the Final speed for MH and OH in the spreadsheet as well as the cooldowns for BT, HS, WW and Hamstring these are the number of attacks I get in 230 secods
Attacks
White MH 151
White OH 139
BT 45
HS 27
WW 27
Ham 108
Total 497
Dodge rate 5.6%
Overpowers 27.85823236
OP cooldown 10.7688096
This gave me an overpower cd of about 10 secs, which from game experiance seems about right although I dont use it that often due to having too much rage to lose. You can use cd's of 30 secs or whatever seems right to you but this was my approach
Total DPS figures for me using my gear, bok, bom, motw, zerg stance, no consumables (standard setup):
Other than resulting in very little differences, the highest dmg was build 3 buy a small margin. Of course doesnt take play style into account I love the fact that I can OP get a high probablity of a flurry and this is great when I'm struggling to get going or in resist fights where I have less than optimum rage gen.
Use the spreadsheet try with your gear, try different cooldowns, use maths or a good parser to see how often you use abilities in fights (my actual cd/s in reality are actually higher due to human button mashing ability, lag and increasing number of fights requiring movement)
@Khalim: Do your results reflect the fact that improved overpower greatly improves your chance to proc flurry?
I'd be willing to agree to Carnitine. His build was my "single target DPS no aggro limit" (aka Patchwerk) build too, before i had to put some points into Prot tree because of raid issues.
The one time i used it at Patchwerk i remained exclusively in battle stance (no +3 Crit bonus from zerker, but stance swithcing loses precious time AND rage) and overpowered every opportunity i got. Which in effect got me flurried and with a consecuting HS got me WAY more damage that a WW in berserker stance followed by an autoattack. For roughly the same rage cost.
And it could have been a streaky run but overpower messages lit up way more frequently than expected when i was DPSing PW yesterday.
Khalim, can you elaborate on those errors a bit more? I'm sorry the later versions have been unstable, but the spreadsheet has grown so complex that it's inevitably become one big circular reference. Make sure you have iterations on to at least 10.
As for the talk about best dps talent build, I think the 34/17 build is best. I'm doing some modifications to overpower currently. Battle stance + overpower actually compares very well with berserker stance + WW. Food for thought, it will be in the next release.
Question for the math wizzes out there. I can't for the life of me understand how to figure this out, and it feels like I should be able to. Anyway, since I've been going over overpower, I've realized I need to take into account it's increased crit chance. This also applies to yellow hits when your white hits reach the crit cap. So, basically I've got:
flurry chance = 1 - (1 - C)^X - (1 - (C + 50))^Y
where the second part is obviously the overpower part. I'm not even sure if subtracting it is right. Something tells logarithms, but I can never get an answer that feels right.
Just an FYI... I was playing around with my PvP gear in the spreadsheet, and noticed that Dragon's Blood Cape stats are shifted to the right by one cell... should be 9 str and 22 sta. :)
1 - (1-c)^X*(1-c-.5)^Y is what you're looking for, I think.
Melador> Incidentally, these last few pages are why people hate lawyers.
Viator> I really don't want to go all Kalman here.
Bury> Just imagine what the world would be like if you used your powers for good.
I think there is an obvious issue when you get up in the crit range. Your white hits are going to hit the crit cap and so will your overpowers. Think full buffed with 2 crit stones + onyxia + mongoose for example. There are different crit caps for yellow and white damage.
So C1 = Min(whiteCritCap,crit) and C2 = Min (yellowCritCap, crit) and C3 = Min(yellowCritCap,crit+.5)
Then you need 3 different sections
something like:
1 - (1-C1)^X*(1-C2)^Y*(1-C3)^Z
As for the talk about best dps talent build, I think the 34/17 build is best. I'm doing some modifications to overpower currently. Battle stance + overpower actually compares very well with berserker stance + WW. Food for thought, it will be in the next release.
Awesome! This is why I was asking about fully modeling the dodge/Overpower frequency before. Some super quick math was showing that Battle Stance wasn't as bad as I thought.
Khalim, can you elaborate on those errors a bit more?
The errors i'm getting with the latest spreadsheet 0.461 is that on opening it says it has a circular reference on cell b44 (Flurry chance) and any changes to gear, consumables etc will not update any calculation. Thats why I was using 0.45. 0.46 has the same problems, not sure if its just me or if anyone else is also getting these. Not really suprising its the flurry field since its the most complicated and is driven by and relied on by so many things. I wonder if its worth taking out things like hoj and windfury to fix it?
Also does the way I calcuated the theoretical overpower c/d make sense? can that be incorporated in?
@Khalim: Do your results reflect the fact that improved overpower greatly improves your chance to proc flurry
I thought that Deathwing was adjusting for imp overpower at least in the DPR value of overpower but clearly he is now building a more complete adjustment. Hopefully he'll be able to keep enhancing this spreadsheet without completly breaking it.
I havent tryed dpsing in battlestance in instances ever, will have to give it a shot with like the bwl drakes this week since i ussually average about 700-740 dps (using rec but no consumables) on them it will be a good test since we're not up to patchwork yet.
The alternative which i havent also tried is not to spec imp overpower and go anger management and just stay in zerg stance. I wonder if the loss in overpower dps and lower flurry chance is offset but the no decrease in crit chance, increased rage from anger man and no disruption to rage flow (esp with 500 or so ping)
The complexity of modeling every element of warrior dps also raises the interesting question of whether bliz has a modeling / simulation software for each class or whether they just use acencdotal evidence and pot luck
Hmmm, I wasn't sure if the Tool options were controled by the spreadsheet itself or by the user's excel program. The whole spreadsheet is one big circular reference, there's no getting around it, and sometimes it makes it quite a pain to code. I'm sure I don't have to point out specific examples. I could make the spreadsheet only calculate cells in a once-through manner, but that seems sloppy and you lose out on "fine tune" dps.
I only just recently discovered Excel's ability to do iterations. That's why I went from White Flurry->Final Flurry to just plain Flurry Chance. I'm suspecting that you have iterations turned off. Go to Tools -> Options, go to the Calculation tab and check Iterations, set it to at least 10, 15 to be safe. And set maximum change to 0. This allows excel to calculate those circular references over and over until the change from doing them again is 0(or it did it 15 times). Download the new version, give that a try, and let me know how it goes. That's also why none of your gear changes are going through.
As for overpower, yes, for the most part that makes sense. It's actually pretty damn close to what I'm getting with the new equations I'm using. Only difference is that I'm using an equation that accounts for overlap, at least I assume there is some. Anyone correct me if I'm wrong: if you get a dodge at 0s, hit overpower at 1.5s, get another dodge at 2s, that second dodge is basically wasted because it was withing 5s of activating overpower.
Anyone correct me if I'm wrong: if you get a dodge at 0s, hit overpower at 1.5s, get another dodge at 2s, that second dodge is basically wasted because it was withing 5s of activating overpower.
You should have 0.5s to activate Overpower after it cools down. At 7s, Overpower can be used again. Within the 7-7.5s timeframe, you can use Overpower again.
Excellent problem solved thanks for that, perhaps put that in the instructions too if its not a default excell setting.
You're right of course, knew you would think of more scenarios like overlap - will be interesting to see how you code it. Using my figures posted previously I was calculating you would get approx 26 chances to overpower in a 300 sec fight. Some of those you wouldnt use because of the overlap or having too much rage or missing it due to human error/fight movement etc. Thats why gut feel or parsing is probably better for overpower than a formula. But if you do user overpower as often as you can somewhere between 10-30 secs seems about right.
Looking forward to the next version if it takes dagger normilisation into account, does weapon skill by weapon and models using recklessness. At the moment by best MH weapon would be deaths sting and harbringer > iblis. Also would like to compare using MSA and an axe/dagger/sword off hand with and without edgemasters
Also can you put a field where we could put in the time spent under 20% health rather than programed at 1/5 of the fight? 60 secs seems bit long on most fights
I don't want to seem like I am riding coattails of Deathwing's success but this isn't worth a new thread. I'm done with my simple stand-alone calculator for warrior DPS: http://www.noobschoolbus.com/wow/WarriorDPSCalc.rar
I took a bit of a different approach, instead of putting in all your gear slot by slot you just put in a few stats and the amount of hit and crit from items and your weapon information. It's not as comprehensive as Deathwing's sheet, but it's not really intended to be. It's more of an at-a-glance comparison between weapons and talent builds than an extensive look at your DPS, so there are a few fudged roundings like assuming Overpower isn't on cooldown and basically a lag-free environment. It should be accurate to well within 5% for most people. This is the first "release" of it so there may be bugs, let me know if you find something that's out of whack.
The app is a stand-alone .exe that requires the free .NET 2.0 framework, you can find it on Windows Update under optional updates if you don't have it. If ye be talkin like this and use Arrr at the end of your sentences a lot there's a stand-alone download for the framework out there as well.
Originally Posted by CheshireCat
Eh, my nostalgia goggles aren't as good as they used to be.
Moogle, my spreadsheet isn't supposed to be an extensive look at DPS. As I've said before, I could have the final number be in bananas per minute, and which has more bpm, that's your better gear. My point, bananas per minute is just about as much relavance as my spreadsheet has on your real dps in game.
I'm trying to make my spreadsheet a colaborative project. I try to remember who gave certain ideas and credit them on my notes. Is there a certain feature you requested that I didn't implement?
Also, from a debugging standpoint, releasing a program makes it very hard for your users to give you feedback. The biggest part comes from accurate modeling equations, and there's no way for the end user to check that. I know my spreadsheet would have stopped dead without the help of people in this thread.
Deathwing: i think the most useful feature of his program (which I won't use for security reasons) is the ability to specify what your stats are instead of what gear you are wearing.
Ive also been giving serious thought into making an addon version of the spreadsheet. Lua isnt that much different than excel.
Is there a certain feature you requested that I didn't implement?
No, I just needed something to do at work between code availability. :|
If anyone wants I'll post the source code but it's uuuuuugly. I had to redo a lot of it at one point to get the flurry calculations right, then do it over again to get Windfury right so things are pretty disjointed from the perspective of keeping the code clean.
Originally Posted by CheshireCat
Eh, my nostalgia goggles aren't as good as they used to be.
The banhammer would come down on me hard and I learn too much from these forums to do something retarded like try and plant a virus or worm. I wish I had some way to better validate it than that :/
Originally Posted by CheshireCat
Eh, my nostalgia goggles aren't as good as they used to be.
Deathwing: i think the most useful feature of his program (which I won't use for security reasons) is the ability to specify what your stats are instead of what gear you are wearing.
Ive also been giving serious thought into making an addon version of the spreadsheet. Lua isnt that much different than excel.
I'm puting that feature into the next release.
You can make the spreadsheet into an addon, that's fine, as long as credit is given where credit is due. I don't know anything about lua coding, but if you want some help understanding and translating the equations into that form, let me know.
Moogle, even though you present a good reason to use an executable...I'm still quite wary of downloading an exe from a source, for all intensive reasons, that is anonymous.
Moogle, even though you present a good reason to use an executable...I'm still quite wary of downloading an exe from a source, for all intensive reasons, that is anonymous.
As you should be in this day and age, though I have to note with some irony that excel spreadsheets are "trusted" when some of the biggest viruses have been through MS Office macros :p
If there's some way to validate the file to show it's not a malicious one I'd be all ears. Until then, use it or not, /shrug.
Originally Posted by CheshireCat
Eh, my nostalgia goggles aren't as good as they used to be.
I'm sure I'm not the only one posting from work where even the slightest chance of a virus isnt worth the risk. You wouldn't even have to put it there, the host you put it on could get infected and serve up infected exes. I wont open excels with macros either for what its worth :).
As for an addon, I would have to run it by the chief marital officer and I'm pretty sure it would get the veto.
Hmmm, should have thought of that, since excel macros are based on visual basic. Anyway, you're right, you're free to be just as suspicious of the files I upload.
Celandro, can't you disable macros when you open the spreadsheet? it asks me everytime I open the version I'm working on(just put in checkboxes). It would be a shame if you couldn't help me anymore on this.