Warrior DPS Calculation Spreadsheet - Page 73

danyson · 06/04/09, 4:50 AM

@landsoul i compared the dps value from build 405 to 406 and for me it looks like the bt change was a huge buff, but you have also implemented the nerfed rage formular. in my opinion the rage nerf should outweight the bt buff on the spreadsheet. because im getting a 6% dps buff compare to the 405 build and this although the rage formular was changed. thinking so the bt change would have at least increased my dps about 10% which i think cant be true

Tengen · 06/04/09, 5:46 AM

This was asking for help- but I've just figured it out ^^''

sorry!

I really appreciate the work you've put into this Landsoul, it's helped me a bunch!

Ferboten · 06/04/09, 6:15 AM

I'm getting a 10.5% increase in dps myself when I compare the two spreadsheets, changing the c Coefficient in the old one to 453.3 since the old value of 320.6 was incorrect.

One interesting thing to note is using the best in slot list (for horde) with the new sheet gives me 7300.9778 dps, while using the old sheet with a c coefficient of 320.6, a 11.11% (repeating of course) DIM and expertise rating bonus of 65.57997558 to simulate the 3.1 nerfs gives 7552.3152 dps with the same setup. It's like they're playing ping pong with our dps.

I have a question about how Imp. Berserker Rage is treated in the spreadsheet. Does it consider how long on average before a free GCD is available to use it? Or just give a flat amount of rage based on its cooldown? I notice Landsoul dropped it for Anger Management, even though the spreadsheet still suggests Imp. Berserker Rage gives more dps

Lantador · 06/04/09, 9:59 AM

Im having a small issue with the recently released spreadsheet. I just got Mjolnir Runestone, and replaced Grim Toll with that. DC:G is my other trinket, and im sitting at 46,77% passive ArP as a Fury Warrior, but getting increases in DPS by changing my few ArP gems for Str gems. I dont quite understand that, the sheet tells me SEP for ArP is 1,0519, and as far as I know the proc on GT and Mjolnir are the exact same, so why would I suddenly need less passive ArP?

Kalroth · 06/04/09, 10:34 AM

Originally Posted by Lantador

and as far as I know the proc on GT and Mjolnir are the exact same, so why would I suddenly need less passive ArP?

The Mjolnir Runestone proc is really 665 ArPen rating, not 612.

Lantador · 06/04/09, 10:41 AM

Ahhh. That explains it, thank you.

Also found a minor bug in the spreadsheet:

The values on Twirling Blades (BiS ranged) are slightly off, it still shows the values from before the Uldu10 hardmode weapons got buffed.

bo5r5osh · 06/04/09, 1:34 PM

I think there is a bug, when I turn off all the buffs, I get DEV/0

edit: Actually if I try to play around with the buffs, switching this and that off, I also get everything DEV/0

MikHaven · 06/04/09, 2:28 PM

Originally Posted by bo5r5osh

I think there is a bug, when I turn off all the buffs, I get DEV/0

edit: Actually if I try to play around with the buffs, switching this and that off, I also get everything DEV/0

Turning off buffs is a tricky thing, I should track this down for Landsoul, but basically it is still using the buff even though it is 0 Hence the d/0 error. In order to safely turn off buffs you have to turn off both of the buff ammount and the talented ammount. Basically their are two components to buffs, and you have to set them both to 0, so it doesn't think you still have the buff.

Vulgrym · 06/04/09, 2:47 PM

Originally Posted by MikHaven

Turning off buffs is a tricky thing, I should track this down for Landsoul, but basically it is still using the buff even though it is 0 Hence the d/0 error. In order to safely turn off buffs you have to turn off both of the buff ammount and the talented ammount. Basically their are two components to buffs, and you have to set them both to 0, so it doesn't think you still have the buff.

I'm also getting Div/0 errors. I went back and zeroed out "ranges" that were associated with a prerequisite value (eg. Master Poisoner 1-3 set to 0 and %up also set to 0) but am still getting Div/0's.

Trying to step through it now.

landsoul · 06/04/09, 4:46 PM

Yes, I know about the massive jump in damage when I finalized the changes, but keep in mind (and I am keeping in mind as well) the rotations are fairly advanced and they aren't nearly what the average player can sustain.

Make sure you have correctly set your latency and human reaction time as these have a rather large effect on the rotations (lag increases bloodsurge overwrites and increases the time of each rotational period). If you want it as accurate as possible I suggest to put your latency to what it is in game and make sure your average reaction time is not too low. I know bloodsurge representation may be too high but it's my first shot at the new system.

I'll check on the div/0 errors with buffs.

foopeemoa · 06/04/09, 5:22 PM

The ability damage split of the fury spreadsheet seems much more accurate now (for me at least). From last night's IC parse (only fight I ended up dpsing on):
25.6% Heroic Strike
18.6% Melee
17.1% Bloodthirst
14.9% Deep Wounds
13.7% Whirlwind
7.1% Slam

Predictions using the 2.406 spreadsheet:
27.1% Heroic Strike
18.3% Bloodthirst
17.6% Melee
16.2% Deep Wounds
12.4% Whirlwind
8.5% Slam

Previously the percentages in the spreadsheet were always too high for HS for me, but they were really close now. The main issue I see, is IC has a lot of incidental damage so I was getting a lot of rage from that, and there is a lot of movement. I was prioritizing WW over Bloodsurge procs, but Bloodsurge over BT unless Bloodsurge had plenty of time on it. I found myself popping Berserker Rage very seldom now, since I rarely needed to, and there just weren't many free gcds to pop it anymore. It's still nice to have in case you get a string of misses, or when Stormcaller Brundir keeps facing you so you get parried.

And "Treads of the Invader" is in the spreadsheet as "Treads fo the Invader" currently. (I have never seen Melancholy Sabatons drop, sadly)

Inukenshin · 06/04/09, 6:52 PM

Last night on XT(hard)

26.1% - Heroic Strike
19.7% - Bloodthirst
15.9% - Melee
14.5% - Deep Wound
13.0% - Whirlwind
7.40% - Slam

Which is within 0.5% of what the spreadsheet predicts, pretty accurate if you ask me. Keep up the good work :P

Bluefish · 06/04/09, 8:25 PM

Landsoul, would you consider adding a survivability metric to the spreadsheet that would influence the order of the lookup pages? While most of the time SEP is the most relevant factor, there are a couple places where the sheet recommends upgrades that trade 40-60 Stamina for 1-2 SEP.

catch22atplay · 06/04/09, 11:27 PM

I think you have a winner here Landsoul. I haven't tested in a raid setting yet. However i did test on the 70, 80 and 83 dummy. I get at best within 0.25% for each ability and at worse 2% on Bloodsurge (RNG screws it up). I actually beat the spreadsheet 4 out of 5 times on each of the 3 dummies. But looking at it RNG was on my side with crits, bloodsurge and trinket procs when i beat it. The times i lost was when RNG was against me and or when i screwed up my rotation. That's the way it should be and was. 5 15 minute tests on each dummy.

BTW my priority is Bloodsurge over BT if BT has 0.5 seconds or longer to come off cooldown. So i am pushing BT and WW 0.5 seconds down the ladder every time i do this. After testing for quite a while this seems to do better for me. Waiting to only use Bloodsurge when there's an open GCD every 8 seconds was a loss in dps. Ended up being 1 bloodsurge every 12 seconds or worse and a lot of do nothing wasted time. Prioritizing BS over BT raised it to 1 BS every 10 seconds or better. If fact sometimes RNG really liked me and i was doing 1 BS every 6 seconds. BS is just to random for my tastes. I'd suggest more people test the rotation and see how they do.

I must say i'm extremely pleased. This is amazingly accurate given RNG with Bloodsurge.

HS use looks spot on. I'm sure HS use will become a bit inflated fully raid buffed etc as you'll exceed 100 rage and waste potential rage. Hitting a dummy unbuffed doesn't have this effect or rarely did in my tests.

Claertes · 06/05/09, 2:58 AM

I actually beat the spreadsheet 4 out of 5 times on each of the 3 dummies.

Deathwish-timing (with other cooldown e.g. passive and active trinkets) isn't calculated anywhere in the spreadsheet, but a very simple and good way to push your dps significantly.
But maybe we'll se a model (which means averaging probabilities => still place for "doing it better than the sheet modeled") for that in the future.
short version: a sheet is a sheet is a sheet, not sim.