I need help on my current pve gear as a hunter. I have full DS and so do 2 other active hunters in my guild... but I can also use the blue pvp helm+shoulder and T3 bracers+boots together with 4/8 DS. so with full DS I would loose 2.3 crit 2.3 dps +1%hit. on the other hand I would gain the DS 8/8 proc ... about 200 mana and some stamina. so my question is. is the DS proc worth still using full DS over the other build with 4-5 hunters in raids.

thx

Yes. The proc stacks.

http://forums.elitistjerks.com/viewtopic.php?id=7312

Does that help?

That doesn't account for the Expose Weakness proc, so it's not very helpful. EW is almost impossible to account for in a spreadsheet because there's an insane number of variables in how useful it is (where in the shot cycle it falls, how many hunters in the raid, how many hunters are MM vs Surv spec, etc etc ad infinitum).

I used 8/8 DS until I obtained 4/9 CS. I like the proc.

i use my 8/8 for bosses when there are 3+ hunters in the raid, where mixmatch for trash.

Good stuff Kreave, but could you fill out your profile? These guys hand out 3 day bans for no profile/fake profiles.

I tried to solve the problem by directly computing the probability that a single hunter would be under proc, and got a different answer for DPS contribution than you did. So, in the spirit of debate, here is my approach:

The weapon of the hunter is approximately 3.0 speed w/quiver. This means that if a shot procs, then the next two shots will gain the benefit of Expose Weakness since the proc lasts 7 seconds. (Of course, further shots may also gain if a re-proc occurs.)

From this, we can construct a Markov process with 3 states:

State 0= no proc up
State 1= proc up for the next shot
State 2=proc up for the next two shots.

Putting aside stacking damage contributions for the moment BUT recognizing that reprocs can occur, this completely describes the system if we can assign probabilities to the different states. Let p denote the probability of a proc, and bear in mind that this process works as follows:
from any state i there is a p probability of moving to state 2, and a (1-p) probability of moving to state \max(0,i-1). That is, you either proc (or reproc) and go to state 2, or you fall down a state.

Assuming the process has a well-defined ergodic distribution, which seems certain, let us define the following:

\gamma_0= (Long-run) Probability that you are in state 0,
\gamma_1= (Long-run) Probability that you are in state 1,
\gamma_2= (Long-run) Probability that you are in state 2.

to deduce these values, note that we have the following

\gamma_0=\gamma_0 (1-p) + \gamma_1 (1-p) ** (says "If you aren't proccing then it's either because you the proc wasn't up the shot before and also didn't proc on the most recent shot, OR you were proccing the shot before but it was the second proced shot and you didn't re-proc")

\gamma_1=\gamma_2 (1-p) ** (says "if you have one shot under proc to go, then the shot before you must have had 2 shots left under proc and didn't get a new proc on that shot")

\gamma_0+\gamma_1+\gamma_2=1 ** (say, "stuff adds to 1" :))

So, you can solve those. Doing so gives

\gamma_0= (1-p)^2 / [(1-p)^2+p(2-p)]

Hence the probability that the proc is up at any point in time is 1-\gamma_0 \approx 0.062 if p=0.032.

Assuming an 80 DPS gain from 450 RAP, the Expose Weakness gain for a single hunter is 80*(1-\gamma_0) \approx 5.03

This is substantially less the other approach suggests. Now, you might say "Well, you're not allowing stacking procs from the same hunter" and that's true but you can easily see that the probabity of two procs from the same hunter being up is very small and can't influence this result too much.

So, my approach seems more direct. I don't use this math often (ever) so I may have screwed something up. thoughts?

edit: typo fixed.

If you look at what I quoted you can reduce it to:

\gamma_0= (1-p)^2 / [(1-p)^2+p(1-p)]=(1-p)/[(1-p)+p]=1-p

So what you have is that your \gamma_0 is equal to the properbility that the next shot will not proc. But the proc will last an interval of 7 seconds. I chosed to just calculate it is a dps increase of 80dps in those 7 seconds, the reason for this is that I am not the only hunter, so chances are that other hunters cycle fit better in with the proc, or that is proc and I get a multi and aimed off in the interval. Therefore I think that is the best way to calculate it.

There was a typo. The denominator should have been (1-p)^2+p(2-P).

I must go raid, but certainly will get back to this thread.

What you calculate as \gamma_0, is what the proberlity is that the next shot is without the bonus on. Therefore you can't just take this (1-\gamma_0) and multiply with the increased dps. You need to multiply it with the increased damage each shot gets:

450/14*3.0 = 96.4

If you then factor in crit and RWS you get:

96.4*1.05*(0.8+0.2*2.3)=128damage / shot

If I take your chance and multiply with this I get:

128damage / shot * 0.062 = 7.9

which makes it closer to the result I got. Also I think you estimate of 2 shots pr. 7 seconds is a bit low. That is it if you are only auto shooting. In a standard dps circle you get 5 shots pr. rotation(which I set as 10 secs) which adds up to 1 shot/2 secs. Therefore I think that you should calculate it with 3 shots getting the proc, and especially because other hunters also get the proc. And chances that they get 3 shots are even bigger. So if you could do the math for 3 shots I would like to see it.

thx for all the reply's... nice calculating there xD. I'll stick with full DS now till I get 4/9 CS I guess. have fun in debating the most accurate dps increase from expose weakness... but I'm to lazy to make that kind of calculations :s

This math is way too high level for me, but remember that things aren't so cut and dry. For one thing, you should be able to fit in 3-4 attacks in an exposed window (2 autos and aimed, multi, or both) unless you're very unlucky. For another, don't forget that RWS, Barrage/8Giantstalker, and Slaying all multiply the 450 ranged attack power bonus.

I was just wondering following on from this, 5/5 Strikers garb has a rapid fire buff and 2/8 Cyptstalker has an additional 4 seconds added to rapid fire. Would these bonus's be enough to ditch 8/8 DS?

It will probably be worth it on fights longer then 3 mins .. cuz you get at least 23 seconds extra rapid fire then + extra stat bonuses... but I'm not gonna be the one to calculate it :s

Not even close. With Ashjre'thul and a 15% quiver (2.96 speed), you normally get 6.41 autoshots in 19 seconds. With rapid fire (2.11 speed), that changes to 9 in 19 seconds, an increase of 2.59 autoshots. Over 3 minutes and assuming that 1 shot hits for an average of 500 with a 25% crit rate (787.5 average autoshot damage), that yields 1976.62 damage divided by 180 seconds for 10.98 DPS. However, 8/8DS rapid fires for an extra 2.04 autoshots every 5 minutes for 1606.5 damage over 300 seconds, or 5.35 DPS.

Net gain from 5/5 Strikers + 2/9CS bonuses: 5.62 DPS.

That's not even equal to what 8/8DS adds to a single hunter, let alone the raid as a whole.

Well, 50 attack power (neglecting the pet for simplicity) is roughly 7 DPS with a 15% quiver and using a 10 second rotation (ie 1 aimed and multi per 10 seconds). Just speaking of the set bonuses, no it isn't enough. I'm not gonna take the time to factor in the stat bonuses from the most-improved 4 pieces or to run the numbers on the mana regen of those pieces, though. I personally would wait for 6/9 except on fights where you need all the mana you can get, or on fights that are too mobile for the proc to be effective.

For striker 5/5 effect
with striker you get the following dps increase( n = the number of times you can use rapid fire(expect the first time):
(15+15*n)/(180*n+15) * 40%
Fight length: 3 min and 15 secs n=1: dps increase= 6.2%
Fight length: 6 min and 15 secs n=2: dps increase= 4.8%
Fight length: 9 min and 15 secs n=3: dps increase= 4.3%
Fight length: 12 mind and 15 secs n=4: dps increase= 4.1%
without the Strikers set bonus( n = the number of times you can use rapid fire(expect the first time)):
(15+15*n)/(300*n) * 40% :
Fight length: 5 min and 15 secs n=1: dps increase= 3.8%
Fight length: 10 min and 15 secs n=2: dps increase= 2.9%
Fight length: 15 min and 15 secs n=3: dps increase= 2.6%

This is of course without the CS 2/8 effect.

Then you can calculate the damage bonus your self. Depending on the fight length

Errr... no. Rapid fire is not a 40% DPS increase. It doesn't affect multishot at all, and even with 19 seconds it doesn't allow for an extra aimed shot. It's a 40% autoshot DPS increase, and autoshot is only about half our damage. So, halve your given percentages and you've got around the right numbers. That means striker's vs not is a 1.2% dps increase.

Unfortunatly I never had access to T2 pants so strikers + cs is the best I can do. =[

You have some points here, I don't know exactly how to calculate the 40% haste effect. But if you say it is around 20% dps increase I get the following:

For striker 5/5 effect
with striker you get the following dps increase( n = the number of times you can use rapid fire(expect the first time):
(15+15*n)/(180*n+15) * 20%
Fight length: 3 min and 15 secs n=1: dps increase= 3.1%
Fight length: 6 min and 15 secs n=2: dps increase= 2.4%
Fight length: 9 min and 15 secs n=3: dps increase= 2.2%
Fight length: 12 mind and 15 secs n=4: dps increase= 2.0%
without the Strikers set bonus( n = the number of times you can use rapid fire(expect the first time)):
(15+15*n)/(300*n) * 20% :
Fight length: 5 min and 15 secs n=1: dps increase= 1.9%
Fight length: 10 min and 15 secs n=2: dps increase= 1.5%
Fight length: 15 min and 15 secs n=3: dps increase= 1.3%

This is of course without the CS 2/8 effect.

Then you can calculate the damage bonus your self. Depending on the fight length

Using a 3.00 speed weapon and 15% quiver and my stats, I get a 40% haste bonus to yield the following (lag / delay / armor reduction / iAotH not included):

Regular clipped cycle: 569 DPS
Regular full cycle: 540 DPS
Rapid clipped cycle: 626 DPS -- 10.5% increase from regular clipped
Rapid full cycle: 709 DPS -- 31.3% increase from regular full, 24.6% increase from regular clipped.

Rapid cycles are just cycles with a 40% speed increase factored in; weapon speed in this case drops to 1.86, Aimed Shot casting time drops to 2.14.