Melons

Same here, I have a bit under 10% hit, 3/3 precision. Saw several misses yesterday, and I'm quite sure it weren't parries.

 

Phrentic

8% style cap
5% extra miss from TG

->13%

against

3% Precision
a bit under 10%

-> a bit under 13%

Conclusion: You will miss a few attacks.

Style hitcap is at EXACTLY 8%. I tested it by hitting heroic dummy over 6000 times. And yes, the heroic dummy is lvl 83 which is proved by about 900 Thunderfury proccs which count as magical attack, which is proved too, because TF proccs missed about 17%.

Screenshot will be added soon.

 

cypher66

Did anyone figure out if execute spam was better than keeping bt on cd and execute off cd?

I have been doing my normal rotation and filling in free gcd with execute as a rage dump instead of hs, which seems to be doing the best dmg.

 

• landsoul

TF proccs missed about 17%.... that's what you used to prove that the dummy is level 83? You are a draenei too.. Did you know that you get an extra 1% hit on all attacks?

Visit my Youtube Channel for an increasing selection of warrior videos, including Undermanning, PvP, and LK Raids:
http://www.youtube.com/user/LandsoulWoW

 

Phrentic

I did this with my Warrior. Thunderfury not Windfury. Warri is human :-) Frentic @EU Nozdormu

[Thunderfury, Blessed Blade of the Windseeker]

 

Graul

Not really sure what this proves considering you had 10% spell hit and should have only missed 5% with the TF procs at 8% melee hit.

 

• landsoul

I read Vulajin's report as well as many others regarding new hit table theories. Apparently they have changed to some degree, so I ran some tests myself.

Unfortunately, on the target dummy which is a skeptical target but using it and recount is perfectly convenient.

I did 3000 white hits using [Hand Axe] x2 with capped weapon skill.

I had 6.25% anti dodge, and 13.06% chance to hit, and 38.77% chance (100% rampage uptime) to crit.
My crit attacks were around 34%
I was dodged 4 times out of 3000 ~.1333%
I had 14.1% missed attacks (which suggests DW 27% white miss rate)

Now, in the rogue threads, Vulajin posted some very convincing data that there was a -4.8% physical critical strike depression and a 3% spell depression versus boss targets. Others posted data with 27% white miss rates. My test only confirmed this.

What does this mean? Well it means the value of crit will go up very slightly, and that we must take this into consideration when doing calculations. The dilemma between having 6.5 or 6.25 anti dodge is still up in the air because if its under 6.5% you don't want to waste the stat budget, and if its over 6.25% you don't want to be dodged ever. As far as the hit is concerned it doesn't really matter.

For the miss rate on specials, there's some pretty convincing data for 8% as well, but I'm out to test that as well. I want to see it for myself. Update: got a combo with 8.0015% chance to hit (164 rating) and I'm afking for a while.
Update: 1300 attacks, 0 misses. Likelyhood of special hit cap as 9%: 0
Likelyhood of it being 8%: pretty damn good!

Visit my Youtube Channel for an increasing selection of warrior videos, including Undermanning, PvP, and LK Raids:
http://www.youtube.com/user/LandsoulWoW

 

• landsoul

I thought about this idea again of what would happen if you didnt slam or if you didnt WW. The thing that started to get me was, that if you started to delay a WW by .5-1.0 seconds that would also be lessening your chance to proc bloodsurge. Now, let's assume bloodsurge was a 100% proc. It would not be wise to delay the WW to get the Bloodsurge proc, because it would push the WW to collide with the BT sometimes, further forcing WW back to a 10 second cooldown even though it was not intended to be anyways. Also, if you prioritize BT over WW, their cooldowns will overlap constantly.

By your logic, you are delaying WW indefinately if there was a free GCD where slam was not up. This completely nullifies the use of the WW glyph, as it would only get used part of the time. What if you got 3-4 chain bloodsurge procs on your GCDs in between BT? Are you not going to use WW during that time?

The flaw in this logic is that you are calculating the loss of a slam on the chance it will overwrite itself. Okay, it has chance to overwrite itself in many many places other than after a whirlwind.

I don't know, maybe I'm searching of a more absolute way of looking at it.

Visit my Youtube Channel for an increasing selection of warrior videos, including Undermanning, PvP, and LK Raids:
http://www.youtube.com/user/LandsoulWoW

 

Rocco

Macro: /castrandom Blood thirst, Whirl wind, Slam
is what I'll do!

nah but on a more serious note: People are talking earlier in the thread about when the 5% penalty is removed that its stupid to drop down 5% in hit rating, because then you will get too little rage from white attacks. What, why?

If someone calculated this, what is the optimum number when Hit starts getting worse and worse like 329ish was before the buff? Why wouldnt it be 165ish (5% hit and then 3% "free" from talents thus reaching 8%, the yellow damage hitcap)

What would be the diffrence on a warrior in 3.0.7 with 10% hit and a 3.0.8 warrior with 5% hit if you looked on their white damage and rage generation???

Maybe I'm retarded but I really don't see your point.

http://sig.gamerdna.com/quizzes/INFL...-jerks/rco.png

 

• landsoul

27%miss - 8% = 19% miss offhand
27%miss - 13% = 14% miss offhand

14%^2 = 2.0% chance to double miss
19%^2 = 3.6% chance to double miss

14%^3 = .27% chance to triple miss
19%^3 = .69% chance to triple miss

Someone with 5% more hit is getting rage starved 2% less!! I'd rather be rage starved 2% more often and do more damage on the stuff I do hit than have a tiny bit of insurance on rotation, when rotation DPS is only a portion.

Visit my Youtube Channel for an increasing selection of warrior videos, including Undermanning, PvP, and LK Raids:
http://www.youtube.com/user/LandsoulWoW

 

Frederic

Removing the penalty on the yellow attacks doesn't lower the penalty on the white attacks. Which isn't to say I don't think you're right. It's just you aren't gaining any chance to hit on your white attacks.

 

Shha

No you are right its a possible scenario, and delaying WW is a bad thing there. The thing is the whole math there was based on STATISTICS. The way this new fury mechanics work, there is no way to make "right" choice every time. I only calculate the chances - eg there is a lower chance of wasting a slam if you do A instead of B etc. Every time though if you make a decision, a bad RNG can "prove you wrong".

Yes you arent going to use WW, and you will LOSE damage following my order. But what is the chance of that? Even if we assume BT has 100% chance of proccing SLAM the probability still is low. It means you went with a BT,SLAM,gcd,BT,Slam,gcd and both gcd in this rotation got a SLAM proc from heroic strike. The "window of opportunity" for the HS to do it is 2 sec, so basically one heroic strike each time. Chance of that happening is 4%. So yes, in 4% of cases you will end up pushing WW back by 10 seconds. (and reset cycle of slam anyway, so you in retrospect would be better off doing a WW).

So in 4% of cases we exchange WW damage for Slam which is a damage loss (although slight rage gain). (2200 dmg loss)

In 80% of cases (1st Heroic DIDNT proc a slam), there is no problem at all because we just slam at 1st gcd).

In 16% of cases (1st Heroic DID proc a slam, but 2nd didnt), you delay WW by 5 seconds, but you gain a full SLAM. (700 dmg gain).

Thats how you have to look at the whole math there - if you follow my order you have 4% chance of losing , 16% chance of winning , 80% chance of breaking even. Even when your "winning pay" is lower then "cost of losing", still with 4 time more chance its worth the risk - but its just statistics - you still CAN lose out on the decision.

That was with assumption that BT has 100% chance to proc BS, normally the "odds of winning" are even higher etc etc.

 

Phrentic

It seems that my last post was a bit confusing. So I will post this again to answer any futher questions, which occured by my last post.

I did 6000 attacks @ the heroic dummy. I did this with prot-specc, no precision and no hit gear. As weapon I used [Thunderfury, Blessed Blade of the Windseeker]. During these 6000 attacks my Thunderfury (futher called TF) proced over 1000 times. This proc is calculated by the rules of cast mechanic. Casts again lvl 83 PvE mob have a chance to miss of 17%. So these over 17% misses of the TF procc prove that the dummy is lvl 83. If the dummy was lvl 82 the hit chance of the TF proc has been 7% higher. 1000 procs are enough to kill the RNG.

6000 hits and 479 misses is a chance to miss of exactly 7.9833 %

1054 TF procs and 185 misses --> 17.5521 %

If the dummy was lvl 82, TF misses would be significantly lower.

 

Graul

So if RNG is going to cut into WW so much now with Bloodsurge, does anyone actually see the new glyph as a good thing? It a) lowers our DPS on trash and b) lowers our DPS on any fight where AE is required. Granted there aren't too many fights right now where you are normally hitting more than four targets, but there are a few. On single target fights with the glyph it seems like it will be a wash. I don't really see any of the upcoming changes as a PvE "buff" as much as PvP fixes. If anything our DPS might actually be lower vs single target slightly until we are able to add more HS without obliterating the GCD cycle.

 

Shha

Why would it be lower? There is nothing that lowers single target dps honestly. You can simply go with a

WW
BT
SLAM (48%)
BT
WW
SLAM (48%)
BT
SLAM (20%)

X5
Rotation of 80 seconds. vs old WW/BT/SLAM/BT/SLAM x8 for same 80 seconds. That means 15BT 10 WW and 5.4 SLAM on average vs old 8 WW 16BT 8 SLAM which agreed is slightly worse because of bloodsurge procs (assuming decent crit), but it compeltely ignores Slams off heroics. Over 80 sec you are bound to do ~8 heroics in my opinion pushing it to the same level as before WHILE maintaining and easy spammable "castsequence". Now what we discuss here, is mostly a way to make MORE from the changes.

The new fury build will be more demanding from players, but will have a bit higher "dps cap". I overall like it, although I dont expect some huge dps increase.

Either way its the BS change thats actually a nerf atm until we get a lot of heroic strikes in, WW glyph is ok on most part.

 

Graul

There will be plenty of times where you get more than one Bloodsurge proc between BT cooldowns. And that's only if you're doing one HS per BT. It will get worse once you can start adding two for every three BT or two per BT. It would have been tight enough with the old 10 second WW. If you start skipping Slams just to maintain a "one Slam per cycle" rotation, you'll be gambling the RNG and probably go just as many rotations where you don't even get a single proc. My point is just that if you're ever going to Slam ~1 second before WW is up, the new Glyph doesn't do anything for single target DPS. It comes out as a wash and is a nerf for trash and a handful of AE fights. Having said that, would the Cleave glyph be a better alternative at this point?

For the long term, assuming there aren't even more changes, the Bloodsurge change may be a sizeable buff simply due to the fact that you'll be able to get more instants in within a certain time frame slightly upping Deep Wounds and Flurry chances. Whether or not this will be a gain over pushing WW back I don't know.

Something else I hadn't really thought of until now. How often do you think HS and either BT or WW will hit simultaneously? You can't realistically purposely space them apart so that they *never* hit at the same time or 0.5 - 1 second apart, proccing Bloodsurge on top of a Bloodsurge proc. In this scenario, should HS not even be considered as a potential candidate to proc Bloodsurge and if you happen to get a proc then you do, but for rotation purposes not rely on it?

 

TheElephant

Both the WW glyph change and the change to Bloodthirst seem to be positive, and I think their benefit will be easier to take advantage of if the rotation (or prioritization of abilities) is kept as simple as possible. First, going to the 16-second rotation that landsoul posted is a boost to DPS assuming that Bloodthirst were to stay the same. Increasing the BT cooldown to 16/3 seconds instead of 5 is far outweighed by decreasing the WW cooldown to 8 seconds instead of 10.

The changes to Bloodthirst appear to complicate this analysis, but there is actually a relatively simple solution. We can use Heroic Strike to buffer out the times when Bloodthirst does not proc, increasing its overall rate to greater than what it was before. After the first WW+BT in landsoul's rotation, we can either Slam (36%) or not Slam (64%), and if we cannot slam, the we queue a Heroic Strike instead, so the overall probability of Bloodthirst there is .36 + .64*.2 = 0.488, 48.8%. We can use the same strategy on the next BT+WW chunk that starts at t = 6.5s, for another 48.8% Bloodthirst. Finally, at t = 11.5s we have a lone BT and we can queue a Heroic Strike if it does not proc Bloodsurge, which brings the rate of that section to the 36% chance of two attacks (20% + 80%*20%).

Overall, the the average rate of Bloodthirsts has increased. There is an average number of .488 + .488 + .36 Slams every 16 seconds, or 0.4175 Slams every 5 seconds, which we only would have achieved before with an effective crit rate on BT of 41.75%. Finally, any extra Heroic Strikes tossed in as rage dumps could proc Bloodthirst as well, and although there is not much room to use those and it would be a shame to ignore those procs, using the above strategy--landsoul's rotation with Heroic Strike to buffer out Bloodsurge misses--is a pure DPS increase in every sense.

 

Bronwyn

The lazy player could of course make a castsequence makro that inserts slams in all the convenient free GCDs. Then spam the makro while you strafe behind the boss. You'll lose some rage and others will think you've rerolled pvp rogue, but then ... your rotation will be flawless.

/castsequence Whirlwind,Bloodthirst,Slam,Bloodthirst,Whirlwind,Slam,Bloodthirst,Slam

Not counting HS, you have 3 chances to proc the first Slam (~50%), 3 to proc the second (~50%} and then only 1 (20%) for the third.

 

Shha

1st of all I was wrong about some stuff in my post. The whole problem proves to be more and more complicated the more I think about it. Yes the Slam priorities seem in order, but ironically enough it now seems that WW should take priority over BT. So overall WW > BT, BT>Slam but Slam>WW, which becomes less and less obvious. Im trying to actually make a simulator model now to see what its all about lol.

Right now it feels like the right way would be a

0.0 WW
1.5 BT
6.5 BT
8.0 WW
11.5 BT
.......
16.0 WW
17.5 BT

and if Slam collides with WW at 0.0 /16.0 etc , you stil WW and sacrifice Slam use, BUT if it collides with 8.0 WW you give it priority, push WW forward and at same time "reset" rotation to 0.0 point. So the pushing WW at 8.0 actually loses WW dps, but ups SLam AND BT (since you avoid pushing back BT by 1 sec in that rotation).

 

• landsoul

I've spent some time making different rotation models based off of a decision making process inside sized time-slices that are the differring sized of global cooldowns that are 1.5, 2.0, and 2.5 seconds. I would take basically the calculated average strength of an ability (counting crits, deep wounds, misses, dodges, et all) and divide it by the sum of its global cooldown and the amount of time you have to wait to hit the skill. Multiple scenarios came up. Up to Slam/4.5, Bloodthirst/6.0, and Whirlwind/4.5. I set up all these values into a table and began to compare them all. At first glance, several decisions became apparent, an instant slam was better than waiting .5 seconds for a bloodthirst, a Bloodthirst was better than waiting .5 seconds for a WW, and several other comparisons. This however, did not take into consideration any rotation overlaps or losses in the chance of overwriting a Slam! proc.

So then, I began to construct multiple scenarios of possible rotation periods and compared the totals of the values above and I would modify slam's value/(cooldown+wait) value by its percentage chance to be available at that global. There were 4 different percentages: 20%, 36%, 48.8%, and 59.04% to have a slam available at a given GCD. I havent yet figured out how to use these percentages in comparison to subbing them for a rigid BT/WW cycle yet.

Visit my Youtube Channel for an increasing selection of warrior videos, including Undermanning, PvP, and LK Raids:
http://www.youtube.com/user/LandsoulWoW

 

Stuntspike

Landsoul, Is is possibly time for us to move from a spreadsheet to a DPS sim? Seems to me that it might be easier to calculate RNG that way as per bloodsurge procs.

Changing WoW one MS at a time.

 

angi

Doesn`t work because it will cast Slam until the casts finishes. But the makro /castrandom Whirldwind,Bloodthirst,Slam postet above will do what you want. Bind this to your mousewheel and be happy.

 

• landsoul

Nah, there are some sims out there. They are difficult to update, take too long to calculate your results, and can't calculate stat equivalency.

Visit my Youtube Channel for an increasing selection of warrior videos, including Undermanning, PvP, and LK Raids:
http://www.youtube.com/user/LandsoulWoW

 

Tankietka

As for that: no they are not. At least mine :-).

 

◊ Cireena

Since the Sim thread was locked we need to start one to continue testing and allow other to begin utilizing the Sims.