◊ Quigon

Is this just poorly edited?

At high rage levels you can't use heroic strike?
At low levels you don't gain threat?

Also, explain to me how improved HS improves your theoretical max TPS?

Also, I think it is obvious that imp HS is a good thing if you need threat.

Imp HS is used precisely for issues of marginal rage - save 3 rage here and there to potentially allow 1 additional skill down the road, or in a cut-off situation. It will increase your practical threat, but not your theoretical.

Which is why modeling it properly is extraordinarily difficult - almost impossible since rage budget is random.
Hence why I take question of a graph that attempts to do so without showing the math or foundation. Imp HS shouldn't provide additional threat unless you somehow are modeling threat per rage, which is not a fair model due to caps at 0 and 100 rage.

Plus, trying to make a statistical model of when to use what ability at each point of rage is something that I attempted to do in the vanilla version of the guide. Your rage determines what abilities you use - the boss rarely changes that, as you cannot rely on an expected random outcome. The decision making time in that regard would come down to saving rage in 1.5 to 3.0 seconds for an upcoming shield slam or whatever skill and doing the math on the fly of the cost of HS and whether or not you're going to be left with enough rage for the SS if all else misses (including the boss).

This is why GUI's matter. You'll gain rage in seeing "Hey I've got 19 rage with Revenge and SS coming up - best not to burn the HS until that gets to at least 28".
Certain numbers tanks just see and know how to break it down.
17 is a big one - 2, 7, 9, 26 are a few others - don't underestimate shield bash either.

 

galzohar

The graphs assume constant rage generation (which we all know is far from a "good enough" approximation), in which case the GAIN (TPS with improved heroic strike minus TPS wihtout heroic strike, aka benefit of improved heroic strike) from imp heroic strike is 0 at both very low and very high rage situations. In very high rage situations you're doing your max tps and the graph agrees that in those situations improved heroic strike does nothing (gain from imp heroic strike is 0). When you start to not being able to do full heroic strikes without improved, while the improved still allows you to use full heroic strike, you start getting more and more out of it - until you can't even use full heroic strike even with improved, where the value starts dropping again until you don't use heroic stirke at all in which case imp heroic strike has a 0 value again. This is basically the same explanation as above only going on the graph the other way around. So these graph results are pretty predictable but their base assumptions are flawed beyond the point where you can say it's a "good approximation", as we all know dodge streaks are just as a significant factor to your threat as the actual dps the boss does.

Also you can't just go and say "with >X rage you should heroic strike, otherwise don't", as I will be able to come up with (possibly unrealistic) scenario of certain avoidance and boss dps levels where heroic strike above that rage would actually not be a good idea, or that heroic strike with less than that rage is a good idea. Fact is while you can always "leave enough rage for 1 special" (or 2, or 3... etc), you need to also take avoidance and boss dps into account due to the chance for you to have rage for "enough" specials but once those spcials are done you dodge *again* and don't have rage for the next, making the heroic stirke a tps loss.
Say in a simplified example hits happen every 3 seconds and your weapon is 1.5 speed and each attack gives you exactly enough rage for 2 heroic strikes and 2 specials. If you have 50% avoidance, if you HS you'll have 50% chance to be out of rage after 2 heroic strikes and 2 specials, and 50% chance to keep going. Don't do heroic strike and you lose threat for sure, but don't lose as much if you dodge.

Overall I think this is way beyond something you can do with a spreadsheet, and requires a simulator that will take as input your character (talents, gear), the boss (how much he hits for and how often), and provide the "breakpoint" where you should use heroic strike by simply running the simulation with varying numbers of that "breakpoint" and giving you the one with the highest TPS. For example it could tell you that with your gear your best TPS would be 1000 and you should HS if you have at least 50 rage (by simply running all the options and reporting the highest TPS one).
anyone up for actually making a threat generation simulator?

 

lythrdskynrd

I came on to find out what the expertise cap was - but couldn't find it right away.
In doing some research on my own I made it to WOWWIKI that explained.

@lvl 70 Versus Bosses
You need 23 expertise SKILL or 91 points of expertise rating.
As you get 6 Points SKILL from 3/3 Defiance that leaves 17 Points (or 65 Rating) to be at Cap

In pre Kara Gear that is:
[Bracers of the Ancient Phalanx]
[Brooch of Deftness]
[Shapeshifter's Signet]
With 2 points of rating left to get.

Effective +Hit though stays at 142 in all this.

One question I have (unrelated to the above) is what are the recommended stats for starting to tank in ZA?
My tank is uncritable, uncrushable and thanks to this thread - generates really good aggro. The question is:
am I fooling myself with only 12k health and 12k Armor (unbuffed of course).

Those are reccomended stats for starting Kara - but could I tank through Nalorak & other early bosses?

What point (unbuffed) would you say was comfortable?

 

galzohar

12k HP for ZA would be very challenging. Heck kara would not be easy if the rest of your raid isn't overgeared for it. But I've seen a warrior hit 70 and have over 12k HP on that very day with blues and cheap gems, so I don't see how you can have such low HPs when you're going to start ZA unless you just hit 70 and decided to give it a go, which probably isn't going to work.

 

Liar

23 expertise gives you -5.75% dodge/parry which barely removes the dodges of a level 73 raid boss (which is 5.6%). Technically this is not the cap though since raid bosses have alot more than 5.6% parry (I think noone knows the real value yet, guestimates range from 10% to 15%), but after that point expertise becomes less effective point for point as aggro stat since you only reduce parry after that point.

12k armor is pretty indicative of your gear level since armor is heavily tied into the ILVL/quality of your gear. So even without providing an armory link I guess your tank is still in blues which just won't cut it for ZA unless everyone else outgears it and can compensate for this (even then it's prolly too risky).

 

Tharia

I went into ZA with my prot warri alt (15,5k hp, 16k armor, about 50% mitigation total, values unbuffed) with three T5/6 Healers. Lynx almost instagibbed me with frenzied Saberlash+Blocked Hit+Crushing , Bear gave me some trouble, luckily i've got stone form. Hexlord is crazy too when a rogue gets siphoned near the end. For the sake of the nerves of my healers I drank a fort-flask and ates some Crawdads.
Sorry, but there's no way you're gonna make it with 12 k hp, at least not as MT.

 

Liar

Lynx can't crush, can he?

 

Tharia

If I recall correctly, he does. Could be wrong though, will look a bit closer next time .
But well, not even needing to crush you to give you hell doesn't make him any better, does it?

 

tangent

Disarm him when he chooses a rogue.
I don't remember disarm being useful preBC but I've found a few places in BC where it can help stop incoming damage.
Hex lord can be disarmed, lynx can be during the split phase, the big gladiator guys in BT and some other trash as well.
It isn't something I use a lot but when healers are dieing or things are going wrong it's nice to be able to buy a few seconds of lower incoming damage.

 

Uglesh

I think anyone thinking they can even OT in ZA with 12k HP and armor is fooling themselves. My stats are marginally better then Tharia's I have experience similar results. I think that any 12K tank won't make it thru the bear boss trash let alone the boss himself.
It also VERY much depends on your healers and having a good mix of heal types. Lynx has potential to deal A LOT of damage especially without a hunter to Tranq Shot.

 

◊ Quigon

12k HP and armor? I think we have shamans at that level. What the hell are you wearing?

Also you're not hitting the parry cap by a longshot with the available expertise gear, so just keep stacking expertise as you see fit. And I can't recommend that you read or trust anything from wowikki.

 

• Mjollnir

With regards to dodge vs. parry- since I failed to find it in the OP.

Tonight I got into a discussion with a druid who stated that raw dodge (a feral druid's only avoidance) was better than a warrior/pally's dodge+parry. His argument that the roll system worked as follows:
We will assume a warrior with 25% dodge 25% parry, and a druid with 50% dodge.

A druid has only 1 roll, if it's sub 50, he avoids completely. If above, he gets hit.
A warrior rolls once, if it's sub 25, he avoids, If above, he rolls again. (dodge)
A warrior rolls twice, if it's sub 25, he avoids, If above, he gets hit. (parry)

With this argument, a warrior with 25/25 dodge/parry is actually only worth 44% combined avoidance.

I realize this is more of a hit table question, but this given scenario should be dis/proved, and some insight given in the OP. I feel it's highly pertinent.

 

Lookit

The druid misunderstands how the hit table works.

There is one roll - for the warrior in the above situation, picture that roll as a 100-sided die with dodge on 25 sides and parry on 25 sides. The remaining 50 sides will be miss, block, hit, or crushing blow (if applicable).

50% dodge is the same amount of mitigation as 25% dodge and 25% parry.


Edited to add miss to the remaining 50 sides

 

• Mjollnir

Perhaps OP could include a brief mention of the hit table, and the order of precedence given to each event taking place:
dodge
parry
miss
block
crit
hit
crush

Edit: spelling

 

Lookit

From the OP.

Don't feel too bad about missing it, as it is literally the most monstrous post I have ever seen.

 

Garithras

While the druid in question indeed misunderstands the hit table, an equivalent percentage of dodge is better than parry simply because dodge also affects many spells, etc, that are not parry-able.

 

◊ Quigon

What? Other than the 360 cleave which was removed isn't that completely incorrect?

A decrease in swing time is superior to not I'd think...

 

Liar

Yeah, agreed. There is so little difference between dodge and parry that is usually doesn't matter much which you go for although parry gives you a slight threat boost.

The best stat for avoidance is clearly "miss" since that works even while stunned or when not facing the mob. But considering you cannot stack miss chance because it is just a side effect of raising your defense rating, this point becomes moot.

(And yeah, I know that there is atleast one boss, Illidan, that negates your chance to be missed on Shear but even then it it still the best avoidance method there is)

 

Chardonnay

Lynx does NOT crush.
That's why the best option to tank him is two ferals Or feral MT + warrior OT.
His saber lash hits for 9.5-10.5K on 14K armor as I had the pleasure to experience Definetly not doable with fresh Kara tanks.

 

Reliknom

Feral MT and warrior OT on Lynx might be preferable where speed is not an issue. But for those who are close to the time limit and need a fast kill to secure the bear mount, if they DPS the Lynx in split phase, the feral OT will be better for faster threat generation.

"Morituri Nolumus Mori!"

 

• Mjollnir

Apologies for my earlier oversight, and thanks for the quick post.

Examples of this? Parry as far as I know remains the more valuable avoidance mechanic due to the hastened swing. After all, it is far more expensive to accumulate. Dodge and parry have equal stake in physical avoidance as far as I know.

It's been my experience that threat is never an issue on Lynx, regardless of the threat generation caliber of your tanks. Not sure what you mean by faster OT threat generation, and where it applies.

 

Reliknom

My example for the Lynx is specific for a 4th chest timed run where every second might matter, and so higher threat generation is preferable. I was also talking about the lynx in its split phase, where you should be able to go all out on the kitty to minimize corrupted lightning totem uptime.

"Morituri Nolumus Mori!"

 

• Mjollnir

Ah, the way we do it is ignore the kitty and just dps the sham- splitting to the totems when it pops. Hence my confusion.

 

◊ Jameson

Most people kill the cat since it has half the HP of the shaman. But hey, whatever works for you.

 

Lambach

Are you sure "most" people kill the kitty? I read on here a few weeks ago that killing the shaman first is faster(can't actually find the post), and since then we switched to shaman first and its been much faster. Remember that all dots and threat transfer between phases. Usually when we used to kill kitty, we'd notice the shaman at about 60% just from tank dmg and dots from the previous phase.