I didn't include enrage uptime because I thought I read somewhere in here that if you're using berserker rage properly uptime is 99%.;

Since I can't find the post to reference it, enrage has a 9% chance to proc, lasts 9 sec.

in 9 seconds you should have 7 weapon swings (not even including flurry) and 6 abilities.
Assuming 16% miss you only lose 1.12 of those 7 weapon swings putting you at 11.88 swings per 9 seconds with a chance to proc enrage.

Probabilty that after 9 seconds you don't have an enrage up: 28.4%. Then ZR is used, chance of not having an enrage after ZR: 4.5%
And then I don't know how to find the probability of not having an enrage for the next 14 seconds until ZR is up again.
But if you include flurry I think that enrage uptime is high enough to be ignored, at least when considering averages.


You say my way of calculating available slams is wrong.
EDIT:
You were right, I was looking at the probability the wrong way. I was looking at the probability of X outcome happening once in 3 events. What I should have been looking at was the probability of Y outcome happening for all 3 events.

Original post has been edited

I don't see how that changes the rotation though.

Correct, but I believe what neekko was talking about was the ability to use RB when required in both rotations. If he's talking about the enrage that provides damage, I'm not sure how that affects the rotation at all.

About the enrage uptime thing. I took the numbers you posted and calculated the possibilities of not having an enrage during those 9 sec. The actual chance based on the numbers you provide is 32,6%. Also you didn't calculate correctly the chance that an enrage would proc during berserker rage, which is also a mistake.

If you decided to check log parses of actual fights, and tried to see the actual enrage uptimes (dw+enrage+ berserker's rage) you would see that it ranges between 86 to 90%. Far away from your estimated 4.5% and even further than the 99% you initially stated.

Teddy tauren : you edited both of your posts cause both of them had big mathematic mistakes in them. Plz refrain on posting about things your mathematic knowledge is not able to explain properly or correctly , which only leads to you editing your posts by stealing something that someone else posted somewhere else.

The question at hand is comparing what's best to do when both bloodsurge proc and RB is available (and enraged). I.e. should the first off-bloodthirst GCD be used for Slam or RB?

Before I get into my explanation, my answer is the driving factor for Slam vs. RB in such a situation is actually the remaining enrage uptime. As previously said, having an empty off-bloodthirst GCD is more of a dps loss than any re-arranging of RB vs Slam.

I'll assume a simplified control environment:
1. Only options for off-bloodthirst GCD are Slam and RB (ignoring battle shout)
2. Sufficient rage

Scenario A: Enrage has 4 secs remaining, or 2 off-bloodthirst GCD
Using Slam first = Slam + RB + 0.51 Slam = 1.51 Slam + 1 RB
Using RB first = RB + Slam + 0.3 Slam = 1.3 Slam + 1 RB
Result: Using Slam is the obvious winner. Delaying slam would lose some extra chance of a bloodsurge proccing.

Scenario B: Enrage has 7 secs remaining, or 3 off-bloodthirst GCD
Using Slam fiirst = Slam + RB + 0.51 Slam + 0.3 Slam = 1.81 Slam + 1 RB
Using RB first = RB + Slam + RB + 0.51 Slam = 1.51 Slam + 2 RB
Result: In most circumstances, 1 RB > 0.3 Slam, even without factoring chance for fiery attack. Best to use RB first.


Conclusion: If Bloodsurge has already proc'ed, it is best to use it first unless the delay in RB will cause the lost of a RB during the remaining enrage timer.

(Coding this is slightly more complicated, would need to take into consideration Deathwish cd, availability of Berserker Rage to extend enrage time, etc. Another factor to consider is rage level, but I'm sure ur spreadsheet already accounts for this.)

I got a question about fury priority list/rotation.

I usually do BT-RB-BT-filler, with CS replacing RB when it's up. I want to ask, what's the priority list for fillers? When I don't have Slam proc nor Battle Shout available, what should I put in there? Is heroic leap better then heroic throw? What is your 'priority' for that spot?

I'd say heroic leap is a better choice since is does not reset your swingtimer as heroic throw does. I see that you use TG and then it's even worse with the throw because of your long swing timer. Heroic leap does have a good aoe potential and you might want to save it if you are going to aoe soon, but otherwise that's the way to go.

In any case, I think the throw will always be a dps loss unless you are at long range and cannot intercept/leap.
Not a 100% sure how it is for SMF, but i might try it out and see if I can find any gain/loss in dps.

But I'd say leap is maybe the one and only useful dps attack to fill in there. But i always saves it unless I know that I won't need it for aoe any time soon.

Isn't there a minimum range for Heroic Leap ? Do you run away to jump back at the target ? Is there just something I don't know here ? Or maybe I just don't understand how to use Heroic Leap in the rotation.

From my personal experiance: If I am low on rage (20-39) I don't use cs even if I have it avalible since there's a chance that I'll get rage starved (less then 20 rage). Otherwise I use CS right away and delay the BT. Though, your aim should be to not be that low on rage when you are able to use CS, and that's quite easy to prevent. Don't use up your rage on HS when you have less then maybe 5 sec left untill CS is avalible. Unless you know you'll hit 100 rage or something.

About overwriting a bloodsurge procc, it's not the biggest problem in the world, the damage you get out of CS is way better. How I look at everything, CS>everything unless you might get ragestarved. If that's the case I might delay it with 1gcd, but not more. I'm single minded fury aswell and if I after my first BT I open with only have 20 (or less) rage left and I have a bloodsurge avalible I might use the slam since it's free, then I'll have a lot more rage to use when I finaly use a CS after the slam that can be invested in extra heroic striges with excess rage.

CS does realy improve the damage by alot and should be used on cd in 95% of the cases, imo that is.

I don't use it unless the guy i'm fighting have a very large hitbox. Most bosses have a very larg one and the aoe of the leap is quite big aswell. So most of the time you can stand a bit away from the boss and still be in range for the hitbox and still be able to leap a bit forward and to the left/right without ever be out of range to hit the boss.
The minimum range is just a problem on mobs with a small hitbox, and the you should avoid leaping because even if you hit the mob you might end up at the wrong side of it our facing the wrong way etc and miss melee hits

Don't misguide players. Here is what Landsoul posted a while ago:

This is how I would explain why delaying CS is not a big issue. If you start with BT in your rotation.

1st CS at 1.5sec (due to GCD after using BT).

The 2nd colossus smash can be used @22.5 sec of the fight after using 8th BT since 7BTx3sec = 21sec + 1.5sec GCD = 22.5sec (first BT was used at 0s of the fight)

The 3rd colossus smash can be used @43.5sec of the fight after using 15th BT since 14x3sec = 42sec + 1.5sec GCD = 43.5sec

The 4th colossus smash will be used @64.5 sec of the fight after using 22th BT since 21x3sec = 63sec + 1.5sec GCD = 64.5sec

The 5th colossus smash will be used @85.5 sec of the fight after using 29th BT since 28x3sec = 84sec + 1.5sec GCD = 85.5sec

The 6th colossus smash will be used @106.5sec of the fight after using 36th BT since 35x3sec =105sec + 1.5sec GCD = 106.5sec

The nth colossus smash will be used at x sec of the fight after using 7(n-1)+1 BT since 7(n-1) x 3 sec = 21n - 21 + 1.5 GCD =

21n - 19.5 Using this formula you can find how many delayed colossus smashes you could use per fight.

Then the colossus smash time loss would be calculated by comparing a rotation where colossus smash is not delayed and used on every 20th second interval, i.e.: 0,20,40,60... so the to know the time of the fight you would use 20(n-1) where n is the number of rounds in your rotation.

So for example if the fight lasts 10min = 600sec, then not delayed number of CS would be 600 = 20(n-1) ; n = 31
Delayed number of colossus smash would be 600 = 21n - 19.5; n= 29.5

31 - 29.5 = 1.5 colossus smashes delayed

On a different example, if the fight only lasts 340 sec, then the number of undelayed of CS would be 340 = 20(n-1); n = 18
Delayed number of CS would be 340 = 21n - 19.5 ; n = 17.12 18 - 17.12 = 0.88 CS delayed

The number is not a whole and thus you have not lost your full colossus smash yet, it's only behind.

Since I believe this proves my math, then I can claim that

1) you will lose 1 colossus smash when fight is > 399sec
2) you will lose 2 colossus smashes when fight is > 820sec
3) you will lose 3 colossus smashes when fight is > 1240sec

Scenario 3 is impossible, because no fights last over 20 minutes

Agree. There is always a chance of refreshing enrage on any of the subsequent attacks during the enrage. What is your approach on this then? Are you running a binomial expansion to calculate the average duration of a continous enrage period, and using said average enrage duration in place of the current duration from my explanation?

My approach is a bit simpler. If you agree with the rest of my argument (less the duration part), then you'll probably agree that the solution of the problem is really binary. Using Slam/RB depends on whether there's an odd of even number of GCDs remaining in the enrage, not necessarily the length.

The chance to refresh enrage is 9% per hit. From a foward looking perspective, the chances of reproccing enrage is 9%, 8%, 8%, 7%, 6%, 6%, 5%... on any subsequent hit, pretty even in my opinion. Further, this means the chance of increasing the length of enrage by an even or odd number of GCDs is really close, probably +/- a few %.

Considering the total chance of refreshing enrage at all over 7 hits (or approximately 7 secs assuming 1hit/sec) is 49%, and the chance of enrage extension affecting Slam/RB choice is minimal, it is simpler just to use the current remaining time of enrage for determination of Slam/RB. Once there is a refresh of enrage, then the Slam/RB choice can be redetermined using the new remaining enrage time.


Having said all this, Landsoul, may I ask are you planning on creating an advanced version of some type of Slam&Awe addon? These resuilts are near impossible to replicate in real time gameplay without a very very effective addon. I've thought about doing it once, but too lazy to learn lua with job and all.