The source of damage is not the problem with the data.
I'm done getting trolled, it's clear you're trolling.

The source of the damage can be roughly calculated using three points of data --

If casting time takes 3.5 seconds to spawn tornadoes and whirlwind attacks alternate between two hands, then our maths looks like this:

Cycle Damage = [(0.7 * MHDPS) + (0.4833 * {MHDPS+OHDPS} / 2)] * Ticks

The 0.7 coefficient is the weapon damage of tornado ticks multiplied by the casting time. The 0.4833 coefficient is the weapon damage of Whirlwind attacks. In other words, it's a rough estimate of the damage per cycle you do. Ticks can be calculated here: https://us.battle.net/d3/en/forum/topic/6037344497

More information on the math above can be summarised here: https://us.battle.net/d3/en/forum/topic/6864416737

For an automatic calculations, you can check here: http://www.d3rawr.com/d and review the "tDPS" tab that uses the calculations above.

This post was edited by acrimonious on Nov 13 2012 11:04am

I guess u have yet to read my last para of Post#1.
"Pls bear in mind that when doing this test, I'm using a much lower Dps Offhand in relation to my Mainhand. Hence the 30% damage output is more of an underestimation. Skorn Barb or DW barb with similar MH & OH dps weapons will definitely have more than 30% output damage from their ww. "


Doing replicates wouldn't eliminate the fact that Im using a different set of Equipments as compared to other people - DW weapons (1160dps mainhand, 760dps offhand). This in itself would have already meant that 30% damage is an underestimation.

OFC there's some variance, n test won't be perfect as what u said, RNG nature.

But since my point is to show that ww dmg has in fact more than 30% output, then this test is reasonable. Even if u account for those "RNG factors", the conclusion will still be the same.

Wow. Is this thread serious? This is a great test; I agree with that. But my question for all of the individuals arguing this test has a large enough size size is have any of you ever taken Statistics 101 even?

The OP used a sample size of 1. This sample size is not changed by the number of hits or anything else. There are too many unknown and random variables to conclude this test after only one killing of Azmo. To get sufficiently reliable results, one would need to perform this test over a larger sample size.

For example, one mistake could lead to a missed nado. What if this happens multiple times in the fight? Same as if RLTW or WOTB had run out at any time in this fight. How about Carpal Tunnel Syndrome? The circles around Azmodan cannot be considered perfect and therefore there could be a fairly large number of misses unknown.

The margin of error in this test is simply too great with the small sample size. It's like what I believe HDaqua was saying previously in the thread about flipping a coin. The smaller the sample size the greater the chance of variance. Variance needs to be minimized to get a proper result. Otherwise, this data is meaningless.

To OP: please continue with your tests and post the results after many runs. If the result is the same, excellent. At least then you know you have eliminated any bias and skewing from your test results.

e/ I am very interested in the results of this, however, I cannot accept one run as sufficient to get promising results.

This post was edited by FlufyNudel[LTK] on Nov 13 2012 11:13am

I can accept and understand your concerns about this. Guess I have assumed that at least 95% of the hits landed well on Azmodan but i could be wrong.
Ty for pointing this out for me. You can totally get your point across unlike HDaqua.

Will post up my Results later

This post was edited by infallibleleo on Nov 13 2012 11:19am

it makes my head hurt to read some of these comments..

can people really be this stupid/ignorant?

what is the world coming to..

You're welcome, and thank you ;). Always glad to pass knowledge on and help pursue greater understanding. Can't wait to see the results! Have been extremely interested in seeing something like this done for quite some time. Keep up the good work!

As requested, I have extended my sample size to total of 5. I understand that this is still a small sample size.
Would definitely extend this sample size in future to give a more reliable results.

My bad for my initial assumption that sample size of 1 is sufficient & reliable. And ty those people who gave me valuable input to improve the reliability of this test.
Feel free to contribute feedback in making this test more reliable.

Just a brief description of my stats again.
I have 2228str and in
In WOTB mode, my mainhand APS is 2.17 (6.67 ticks per second) n OH APS is 2.51(8.67 ticks per second), (source: http://us.battle.net/d3/en/forum/topic/6037344497)
with 63.5critical chance, 519critical damage.

N for consistency for the test of RLTW only, I ensure that my OH is the last weapon swung before using Sprint. I.e. consistency 8.67 ticks per second.
And for all the runs, no follower was used.

1st Test Run(as per Post#1):
ww & rltw: 1min 45secs
rltw only: 2min 30secs

2nd Test Run:
ww & rltw: 1min 41secs
rltw only: 2min 28secs

3rd Test Run:
ww & rltw: 1min 37secs
rltw only: 2min 27secs

4th Test Run:
ww & rltw: 1min 50secs
rltw only: 2min 23secs

5th Test Run:
ww & rltw: 1min 43secs
rltw only: 2min 28secs

Average time taken for ww & rltw is 103.2seconds
Average time taken for rltw only is 147.2seconds

Computation
X = ww dps, Y = rltw dps, HP = 138,425,408 (azmodan's mp10 health), T1 = time while doing both - 103.2sec, T2 = time while doing just rltw, 147.2sec.
T1(X+Y) = HP --> X+Y = 1,341,331.47, the total Damage Per Second you were doing to kill azmodan in 103.2 seconds.
T2(Y) = HP --> Y = 940,390.00, the dps you were doing to kill azmodan in 147.2 seconds.
X = 1,341,331.47 - Y --> X = 400,941.47, the dps that using ww added.
X/(X+Y) = 400,941.47 / 1,341,331.47 = .2989 or 29.89% of your total dps came from WW

Take Note
This 29.89% is never the 100% accurate answer, n should only be used as an reference. This test should be further extended in future so as to give a more reliable results.

Also bearing in mind that My Mainhand dps is 1160, and Offhand dps is 760, the actual ww damage output will be lowered as compared to Skorn Barb or Dual Wield Barb which has a more similar MH/OH dps.
On top of that, OH is the last weapon swung before using Sprint for my RLTW only test, i.e. consistency 8.67 ticks per second which is higher than my average ticks of 7.67 per second.

Taking into account of these factors, this 29.89% result is an underestimation. My opinion is that the proportion of damage of WW will be more if one were to use Two Handed Weapon, or a More similar Mh/OH dps Dual weapons.

Hope u guys will have a clearer idea on the importance of WW now.

Ww increases damage? No way! ......

It should be quite obvious that if you can generate enough fury then whirlwinding will do more damage than nothing...

This post was edited by Dontrunaway on Nov 14 2012 10:33am

HOTA >