except on Halfus heroic you want to kill 2-3 drakes as fast as possible, not trying to AOE all 4 drakes + Halfus. Even when whelps are out, you should be letting the ranged / good aoe classes handle them, since BF cleave makes combat rogues one of the best drake killers.

It is better for RDPS for the drakes to die faster so the whole raid can do 400% damage to Halfus, than trying to AOE as combat in this encounter. Besides, if you want to pad the meters, you can turn on BF while attacking Halfus sub 50% on the drake(s) you are not killing

There's so many different posibilities on how you can kill Halfus, and how much different guilds does it. I've personally never been able to control my Blade Flurry in our kills, due to the fact we pull all the drakes and gather them ontop of Halfus, and AoE/cleave down from there. But enough about Halfus, I was merely mentioning him for the fact that there's times where you have to deal with boss-hp-pools where substained AoE dps with the amount of targets known actually can make a difference.

I haven't seen it mentioned anywhere in the thread, so I thought I'd point out that RvS effects the duration of Expose Armor, in the event that you find yourself being the only person that can provide the buff.

Not that there's any real reason to doubt this, but I just checked it out. I used a 1 pt Expose Armor with RvS Strike up ... EA was put on with a duration of 13 (probably actually 13.5) seconds and the RvS debuff was removed. Good catch.

Edit: Works on Kidney Shot, too. A 5-pt Kidney Shot removed RvS and lasted 8 seconds.

8.1 seconds to be exact, 8.7 with glyphed RvS. This makes Kidney Shot now the longest duration CC in PvP.

Have been doing some tests currently on which enchant to use on the offhand for Combat, seeing as Hurricane as a chance to also proc from spells and also as Haste is our strongest secondary uncapped stat.
For this post I'm assuming Landslide as a melee-only proc (basing on the tooltip) and having a ppm equal to Hurricane (until proper testing is done).

World of Logs - Real Time Raid Analysis this is a world of log fight from Cho'Gall, fight is (at least on normal) quite tank and spank for melees, at least with the strategy we use. I know you're probably going to say it's too short, but before jumping to conclusions at least read my post.

We see the fight lasted around 8:30 minutes, which is about 510 seconds.

Hurricane seems to be divided in a ppm proc and a spell proc (45 sec icd, 15% procrate). The spell procs should have been 510/50 (45sec icd + 5 seconds for it to proc) = 10. That leaves us with 26-10 = 16 procs from the ppm part, which a landslide enchant would have had.

Now, calculating the total EP values, Hurricane should be around 450 * 1.5 * 26 = 17.550, while Landslide would have been around 1000 * 1 * 16 = 16000.

Basing on this, a Landslide enchant seems subpar to a Hurricane enchant on the offhand. Of course as gear progresses we get more haste, therefore the spell/ppm ratio of Hurricane becomes smaller and smaller making Landslide overtake it, but I'm more interested in current applications.

Of course criticism is appreciated, as I'm not experienced at all, and I know this is based on a few things which we still don't know about Landslide, but I guess it was at least worth posting.

I've been trying to test this- the difference in stun time is tough to watch. I've been in duels glyphing and unglyphing this. I'm not sure the .6 second difference is really real at a 5 point kidney shot. On the other hand , a 4 point kidney shot definitely gets the vote of approval as being longer with glyph than without- but this could be subjective.

Is there a real way to test this besides mousing over and relying on a priest with a stopwatch? When I'm on the dummy, I'm almost positive that the duration is 8.7 seconds, but in pvp I'm wondering if there is a cap at 8.0.

I mostly try to watch my feet. I'm considering zooming in more. My goal is to stand in fireballs little to none, and also cleave the correct drake (the next in the kill order). I try to accomplish this by chosing a position where that drake can be cleaved with the tiny frontal flurry cone, but the other drakes won't be. Because it's a mass of flapping wings, ground effects, and fire poo, I don't have a perfect solution. If we had more combat rogues (or more rogues) then I'd probably call for the non-sequential targets to be tanked a bit further out, but I don't think with just one rogue it's worth reducing the aoe damage the other classes are doing (the splash damage on Halfus and the other drakes that some other classes bring still counts, after all). We're still new to hardmode, having only done it twice (25m).


My apologies, I can't browse WoL from this connection. The estimated 1ppm on landslide (which IIRC has no ICD) would yield a 1.66_% proc chance from a 1.0 speed weapon- so each offhand attack should have a 2.33% chance to proc landslide. I'm not sure if this changes your calculation any, but the chance is normalized per hit, so extra hits (from haste, and slice and dice, and killing spree) each have a chance to proc landslide- again, assuming I'm as up to date on this mechanic as I think I am. Since hurricane has an ICD of 45 seconds (and can proc from spells), I think the question becomes sort of like, will we often see a second or even a third landslide proc from offhand during that time? Do you have a count of offhand attacks over that same fight? I would suspect that the best method would be via calculation, and I would suggest that this has already been accounted for, but if real world isn't bearing it out, it's pretty relevant.

Ok, I see it now. If the melee part of hurricane is in fact a 1 PPM proc similar to landslide, then your method looks sound, and the spell proc portion makes up for the lesser normal proc. You are saying that of the 26 hurricane procs, 10 were from the spell proc, and 16 from the melee portion, and you are assuming that landslide, given the same sets of rolls, would have procced the exact same 16 melee procs (and only those, as landslide doesn't have a spell piece), and that the total value of the buffs would then favor hurricane.

How much uptime was in your example? What are reasonable melee uptime examples this tier for us? These questions then become relevant, because if you only get to touch a boss for 20 seconds every 50, you would have really favorable uptimes for hurricane. Again, and sorry to ask a stupid question that because I can't access the site, but what combination is that rogue in the parse running? We are looking for parses with rogues running hurricane offhand and mainhand anything *besides* hurricane. Was his run simply unlucky for melee proc numbers (in other words, is that 26 a good estimate, or should it have been 10+20 instead of 10+16)?

Had to stop for around 20 seconds in the whole fight to avoid going into phase two with adds, but nothing more than that.

If we account for one proc during that timeframe, landslide goes to 17k and hurricane to 17550 + 450*1.5 = 18225.

The math is actually fairly simple. Assuming the 1 PPM proc has a x% uptime, the spell proc on hurricane will give an additional 24% uptime (12 / 50). Thus we compare (x + 24) * 675 vs x * 1000, which has x =~ 50%.

Thus if a single offhand landslide enchant has less than 50% uptime, which needs to be checked, hurricane should be superior. We will need to look at logs to see in practice what this uptime is, by finding logs with combat rogues using an offhand Landslide, and something different in the mainhand. (Note this is under the assumption that if a spell proc occurs, it will always generate a new stack of hurricane buff)

editted with more thoughts:
According to the mechanics thread for Hurricane
"On spells, it is a 45 second ICD and a 15% proc rate. If a spell proc occurs while a melee proc is active, it will refresh the melee proc. If a spell proc occurs with no melee proc, it will create a separate stack of the buff, potentially resulting in 3 stacks being up at the same time."

So according to that, we only get a second stack if a spell proc occurs before a melee proc, but we do not if a spell proc occurs while a melee proc exists. This would drastically decrease the 24% uptime estimate above, as well as your 10 procs from Hurricane estimate, since both assumes the spell proc will generate a new stack every single time.

On the other hand, looking through my logs around 3-4 weeks ago, when I used Hurricane/Landslide, landslide only had an uptime of around 15-30% if used only on the offhand.

So the big question we need to figure out is that, when a Hurricane spell proc occurs, what is the chance it will generate a new stack, as opposed to refreshing a melee stack.

Also note that a Landslide offhand enchant is much better than Hurricane enchant if you want to multi spec Mutilate or Subtlety for different Hardmode bosses.

Mainhand is expected to have a great uptime due to istants (especially during adrenaline rush), which offhand doesn't have.