Ihmes

There's this thing about imp sunder and HS:

I specced for both 3/3 once, and noticed that on teron, I was rage-starved way less. On the kill before, I was constantly rage starved for some periods (avoidance streaks?), but on that kill I think I dipped under 10 rage maybe once. I had a WWS on that but it's expired already, got only a few "normal hits", 90+% was HS. Sustained, WWS'd TPS was around 1100 (with this Kenneth Gant Niebuhr - Online portfolie)

So basically those talents are useless in infinite-rage-situation, but they help to make more situations like that. So hardly useless =)

 

Xerophyte

Oh, I'm not at all emotional, I'm sorry if I gave that impression. We're discussing the relative value of some of the most insignificant choices we could make in a virtual world, hardly cause for anger. I just like to debate: worst case scenario I get to practice my english and best case scenario I might also learn something.

Over half of your total potential rage expenditure is HS: specifically, keeping up an instant cycle and block is ~8 rage/second and adding HS to the same ups that to ~17 rage/second. There's a huge range of incoming damage for which you'll either be on a partial HS use cycle of some sort or almost permanently at 100 rage, certainly every 25-man boss I've ever encountered falls within that range. Kaz'rogal is in the higher areas and I've personally about 70% of my melee attacks turned to HS, but it's never much near 100%.

I'm confused as to why you see mitigation or threat as an either/or situation. Obviously no tank can get away with ignoring one completely at the cost of the other, or we'd be tanking all threat sensitive encounters in full dps gear just to put out 2k TPS (good fun to do on Solarian for kicks and giggles, though). We can thus infer that there is always some balance between the two to be struck and I personally don't feel comfortable making absolute statements as to where that balance is and what trade-offs anyone else should or should not make on one vs the other.

 

◊ Quigon

Xero: Typically when progressing survivability is much more of a concern than aggro. It is rare that you need additional talent help to stay ahead before a first kill, especially as dps often do not outgear your threat potential (and probably never will before level 80 now that expertise is in). As their gear improves, and your rage generation becomes worse through improving gear, you'll find aggro talents to be more useful - which is basically farm mode.

As for Enkidu, I never said imp HS was worthless. I'm not even sure where you're coming from with that.
But if you want better mitigation, there is a spec for that.
If you want better aggro, there is a spec for that as well.

You will simply not run into aggro problems if you're doing even marginal rotations on the bosses in this game that hit hard; this is especially true now with expertise. Mother Shahraz, Archimonde, Az'galor you can spam a macro if you like and afk and you'll be way ahead of even a warglaived warrior/rogue.

There are dozens of hybrid specs where a player can choose what they want as some "best of both worlds" spec. I'm not going to debate the relative worth of those dozen specs, other than to say they're neither here nor there obviously. I've use middle-of-the-road specs for a long time as well, they work fine if you feel you must have say: AM, or Imp SW, at the cost of 1% or 2% or 6% of this or that here or there.

The original argument came up over how 5/5 cruelty was assumed standard for Prot warrior specs, which I think it is. It wasn't my claim, and it wasn't even JamesVZ's claim to whom I was agreeing - it was some automated system to look for poor specs. You're free to disagree, but I stand by the original statement, and the math contained herein. I'm not sure arguing about it will change that. Yes HS is a huge part of your RPS useage - but that isn't a corollary to imp HS being one of your best threat talents.

The bottom line is 1 point in imp HS and cruelty are both insignificant. (Although IMP Demo Shout and Imp Def Stance are not.)

 

Xav

I don't know about the whole "Dps can never catch you now" thing. I consider myself extremely proficient at generating threat, and there have been a couple situations where you normally wouldn't think people could catch you, but they do. Teron is one: people are going absolutely crazy with DPS stacking, consumables, groups, timers, and it's a very short fight. People are now putting numbers like 2300 dps sustained out on it, and that's going to demand around 1400 TPS sustained from the tank. That's where you start greatly sacrificing avoidance and survivability in exchange for threat.

On Shahraz, I wear absolute pure avoidance/mitigation gear, which puts my threat output, paired with playing extremely defensively (making sure TC is constantly refreshed, shield block always up just to add extra mitigation, and demoing in the likely event the DPS warriors are choosing to not refresh it constantly), really low. Even in SR gear (although I'm sure some fo the rogues are wearing less and less SR these days), I've had rogues (warglaives) catch me, because they're not suffering nearly as much as the tank does in that type of scenario.

Without expertise, there really is no way a tank could out-threat the top DPS potentials nowadays, and that's probably a huge reason why they added it. And even with expertise, you have to gear and play very well to not be passed by 2300 dps classes.

But yes, during progress raids, people are absolutely going to gear and play more towards survivability most likely, and that's when you definitely should be saying to your raid "Watch your threat, I'm not in threat gear!".

 

◊ Quigon

Xav, I specifically distinguished between top DPS potentials from farming, and progression mode in my post. Just reread it. I mean your final line is the only one that appears as if you read what I was writing.
All the stuff above it is the reason I wrote a 5+ page guide on aggressive warrior tactics.

Also, I don't see for the life of me how you can are getting caught on Shahraz... I'm usually 50% ahead of the warglaives on that fight.

Rogues have been sustaining 2000+ DPS since well before expertise - we had a rogue put up 2000+ more than a few times back when I was in a pure-mitigation spec as well.

Also, for an initial Teron kill, aggro should not be a concern either. You simply don't have the reasonable avoidance gear at that point to be having rage issues. Farming teron, yeah probably - cause people see the need to go balls out as if it is some proving ground. Cause saving 30 seconds on a boss really equates to time saved after 15 minutes of trash.

If you find yourself needing a hybrid spec and giving up some mitigation to achieve higher aggro, then by all means go for it.

 

Aurrior

I think one important thing to point out is that you can *almost* spin your GCD constantly (devastating and revenging,
shield slam and shield block makes things different) just from the rage you get swinging your weapon.

Then again, I run 12/6/43, which is a rage/threat build, so I get an extra 1 rage/3 seconds and 1 rage every few swings.

Ignoring that, dumping in HSes is almost entirely a result of getting hit. There's few reasons (refreshing debuffs, making sure you have enough rage for another shield block, and interrupting) to not have your GCD rolling.

If your doing everything right, and are still max raged, then you don't need imp HS. Otherwise, that bonus 400 threat (200 threat, 200 extra damage) per swing is one of the best threat sources you can get.

(Does anyone else bind HS and Devastate to the mouse wheel?)

 

Whiteknight

I think it's valuable to look at the frequency of attacks when evaluating the relative strength of the various talents. In particular when considering rage efficiency talents for threat builds.

I experimented a while back with UW, imp heroic, imp sunder in order to improve my threat and help out with rage starvation issues, and I found the value of the talents is (best to worst), imp HS, imp sunder, anger management, UW.

The reasoning is simple. Regardless of the fact that HS is the lowest priority move in your threat rotation (i.e. you drop it first) almost every single WWS I have for T6 bosses shows that HS is by far the *most used* tanking move. The only bosses for which this does not hold true are Gurtogg (not getting hit + avoidance gear), Rage (hits like a pansy if you're not in the AoE) and Akama (you tank adds, not the boss).

On most of the fights I find Heroic Strike is approaching the optimal ratio (hs:4, dev:2, rev:1, ss:1 per ~6 seconds). In low rage fights, HS is closer to a ratio of 2, but there really aren't many fights like this in the game.

This means that at worst, Imp HS is as efficient as Imp Sunder. At best it's worth 3*4/6=2 rage per second.
Imp sunder is roughly 3*2/6 = 1 rage per second.
Anger management is 1/3 rage per second.
And UW is 0.266 rage per second (for 5 points).

Note UW procs 40% of the time from swing based attacks only. Abilities like shield slam, devastate and revenge supposedly don't proc it (if you believe wowwiki: Unbridled Wrath - WoWWiki - Your guide to the World of Warcraft). I didn't explicitly test what it procced off, but I measured number of procs over the duration of the fight and found UW to be particularly low efficiency for tanking.


Anyway, practically speaking, my experience is that improved heroic strike + improved sunder armor will go really far towards turning those 'low-rage' fights into medium rage ones. Similarly a medium rage fight becomes full rage fight.
I found that anger management and in particular UW were not worth the talent points because there's already too much I want to get from my talents. It may sound counterintuitive, but if you find you're not using heroic strike much because of rage problems, you may find imp heroic strike will let you use it a lot more.
Anyway, when trying to decide what's right for you, go look at your WWS history and see what abilities you use most, and what you use least. This will give you a really solid foundation for basing your build tradeoffs on. In addition to pointing out possible weaknesses in your threat rotation.

For reference (because armory is showing my pvp build right now), I tank with this spec currently Talent Calculator - World of Warcraft
It's almost exactly the 'threat' spec Quigon describes in his OP post. When we were doing progression content, I dropped imp sunder/imp heroic and 2 points from 1hand spec to gain imp TC and imp Demo.

 

◊ Quigon

I'd agree with your order in threat, but I'm not going to attempt to justify it with math, only by experience it sounds right:

Imp HS > AM > UW
UW just never felt noticeable at all as a prot warrior.

I also really liked imp sunder before devastate was implemented.

Also, I am not sure one should take Imp HS over Imp Sunder with the modern implementation of the talent. Even if in theory it provides more rage saved over time, Imp sunder allows you to have 9 rage devastates (and sure, 9 rage HS's), but you need a selection of low rage abilities to fit into a low rage situation... like a game of tetris sometimes you have to choose the appropriate ability for the rage that you have at that moment. You would typically not choose an HS in this situation.

And yes, spamming HS is hugely important. I typically just spam the hell out of that button unless I feel a strong need to get into a perfect rotation to get ahead of our top DPS'ers. Imp HS is a very nice talent. But that doesn't mean I'd say its the way to go in all situations.

The problem with playing smart with HS is that its not immediate - and it must be pressed very often... therefore, most tanks simply spam the button to death. Thinking every HS out is actually borderline impossible, or at least not worth the massive distraction of thinking about each weapon swing in a real world situation. However, that isn't to say you shouldn't let up on the spam when the red bar gets low.

 

Xav

I read your post, I just think it was wrong to say that people will never catch a tank now in threat. It's just simple math really, sure dps may not catch a tank if they succesfully shatter/vanish and don't slow down, but that's kind of the point. On longer fights, some classes just will catch the tank since they can out-threat them. I don't "spec" for threat, I don't have imp HS etc, but I would if it was possible for me to get it safely. It's just a gear thing, and is for epeen - people are going to mash the buttons like no tomorrow with no care in the world for aggro on Gorefiend nowadays, and you're going to want to be able to hold it still. I'm completely aware of the differences from progress to farming!

 

◊ Quigon

Are you having negative ratios against DPS on Shahraz? Shahraz even with 70+ avoidance gear should be giving you plenty of rage to work with. Further melee uptime isn't 100% unless they're extraordinarily lucky, and they're wearing resist gear. A good melee does maybe 1500 DPS on shahraz?

You know that you can accidentally be facing a wall on Shahraz, land melee attacks but not specials. I've had that happen before.

I think perhaps you're only talking about Teron though, and to that, I agree that Teron becomes an aggro fight on farm mode.

Teron however, was not a fight I listed as having trivial aggro concerns with - but you can simply stop blocking on a farmed Teron and use the extra rage for threat. A tank stands little risk of dying there in a farming situation where it is over quickly.
I believe it was the tank in Furi's guild that got crushed a lot on that fight... I don't think I've hit shield block outside of T5 2 piece bonus on that fight in a few months.
I know it is crazy, but sometimes taking crushing blows on purpose is perfectly viable - like say: Gruul, Leotheras, Teron, Najentus (I'm sure there are many others).

I guess if a tank is afraid of dying and wants to wear avoidance gear and block on Teron, yeah they're gonna need to consider aggro talents - but hopefully not at the cost of CoR - which will hurt you more than a minor threat cap.

You don't want to make the mistake of dropping Imp Demo in order to take aggro talents to appease a 2-5% raid threat cap increase and suddenly drop CoR to compensate for incoming damage at a more significant DPS loss.

Further, I think Teron is a bad example. It can be done in aggro or avoidance gear trivially. The problem with Teron is people see it as an epeen fest - usually at the loss of control and time.

A better case for discussion of aggro would be reliquary phase 2 where threat truly is limited, and aggro is a huge concern on that fight - you probably shouldn't debuff in RoS phase 2, no demo, no TC, no blocking, full aggro gear - and then switching into an optimal burn rotation for P3 with SB/oh-shits. Or perhaps offtanking on Bloodboil, although you won't be using HS there, or aggro gear for that matter.

 

Xav

Yeah I take some crushes on Gorefiend of course.

Shahraz is aggro sensitive when it's a dual warglaived rogue wearing very little if any SR, vs a tank in full avoidance playing extremely defensively - abundantly refreshing TClap + Demo and not getting very lucky at all with landing abilities. I'm not facing the wall or anything stupid, it's just an environment where I'm not there pumping out massive tps. I'll easily go 10+ seconds without getting hit, so that's a huge factor in streaky, and poor TPS, whereas the rogues aren't suffering anything like that. After looking it up, the rogue's doing just under 1700 DPS on it, which is I think more than 1000 TPS. (And that's with an FA, so it was probably 1800+ before the FA hit). After seeing it happen once it probably won't ever be an issue again since I'll play a little less defensively, although with vanish the rogues would probably never catch again on that fight either. It's sort of the opposite thing here - tanks can never downgrade their gear on Shahraz or they'll get owned, but as healers and DPS get better at surviving FA, they'll probably take off their SR and start to creep up on you ever quicker.

Funny you mention RoS phase 2 because I never have aggro issues on that phase, and only ever lose it in late phase 3! I don't do anything absurd there either, I shield block through most of phase 2 and just save a bit of rage when Deaden time is up while still doing a normal rotation. It's probably the most focused tanking you do next to Illidan, since you really have to be dead set on building your threat but not mashing your buttons so mindlessly that you're screwed come Deaden time. Fun stuff.

 

◊ Quigon

I see.

I have one of the protection warriors who is doing nothing do all the debuffing on Mother Shahraz.

The aggro gains from not having to debuff are quite large - 2 abilities every 30 seconds, usually I cast them way more than necessary because of the insanity of if they drop. I've certainly seen strings of demo resists well over 3-4. This is something I mention in the guide, about having OT's debuff - but should probably be noted for the mitigation warrior specifically.

I can absolutely see aggro loss happening in that situation, although I assume your rogue wears more than "if any?"

 

Xav

The rogue is probably wearing 150ish or something. I haven't raided with a warrior offtank in some weeks now, and the DPS warriors aren't fond at all of refreshing Demo, and on a typically irrelevent threatwise fight like Shahraz I won't get on their case about it. I also always have to Tclap for myself as well, so yeah, it adds up.

 

Xerophyte

Thunder Clap isn't too bad to self-refresh if it's improved, according to the spreadsheet it ends at about 340 threat/clap after resists, compared to 440 threat for Devastate. Quick napkin math puts it at a loss of about (sorry, couldn't resist abusing the for once...)



...to swap one Dev for one TC every 25 seconds. Given that the stance swaps needed for DPS warriors to TC are pretty disrupting with resists and so on and it doesn't cost me much threat I'm honestly not to fussed at keeping it up.

If the dps warriors desperately want to threat limit themselves by HSing for the honor of doing more threat/damage rather than covering demoshout on a sensitive fight then I'd first tell them to shove it, though :p

 

Enkidu

UW is indeed a terrible 5 pointer for a tank, it generates less rage than any other rps/efficiency talent by far (barring a 0 HS situation). I took AM over Imp.HS because it provided pretty similar rage in the example I looked at (about 55%-60% HS use) while being more beneficial in other situations where HS use is lower.
Are your casters not giving you more threat issues than your rogues these days? Mages not using invisibility while doing near 2k DPS as fire can be a real pain in the ass at times.

I tanked Shahraz last week with 73% recorded avoidance, while rage was pretty streaky HS use was still around 60% so it's not all that bad. Though swapping out a bit of stamina for more avoidance was, really, really noticeable in spike potential, did not think Moroes' Lucky Pocket Watch was that powerful until it finally dropped.

 

Fellwraith

I thought they changed it to a PPM model so that it didn't provide bigger benefits to faster weapons vs slower weapons. Pre-TBC it was a set value, and that was the argument for faster offhand weapons for a fury warrior, but they changed it more recently. Wowwikki might be out of date.


I find AM to be a useful tanking talent, but I think it's mostly a subjective argument. There's something to be said for getting a set amount of rage while trash is CC'd (warlocks have a nasty habit of banishing at the wrong time), a boss is in the middle of a transition (Leo, Kael), your rage income is interrupted (Naj'entus, Rage) or you're offtanking. Before you get some of the higher end weapons, the talent also ensures that you can always shieldblock, even if you miss an auto-attack and get a big string of avoidance. This is particularly important for Morogrim where your attack speed is reduced. It's one of those talents that doesn't look very exciting on paper, but is very useful in practice.

Assuming you're already 8 points into arms to get imp thunderclap and parry, the opportunity cost of going 3 points further in is relatively low. However, I also expect a fury warrior to be able to demo shout inbetween whirlwinds and bloodthirsts, so I wouldn't expect a tank to have to provide his own demo shout in the fights where you'd really need it (Maulgar is probably the only hard-hitting boss I can think of where I wouldn't have had a DPS warrior to provide demo, any other hard-hitting boss is usually a single target, so the other warriors are on dps duty).

 

stampy

The problem with Imp. HS isn't its efficiency in its best possible situation -- HS every swing -- its that it is a talent that becomes better and better the less you need it.

If you have the rage to heroic strike on every swing, as in the "this is why imp HS is really good" example above, you just plain don't need it. In an encounter where you have the rage to be using heroic strike on every swing as well as an optimal threat-per-second instant rotation, you aren't really going to miss that one heroic strike out of 7 or 8 swings. And, as long as were going to talk about "full rage," that is only a factor if your rage generation is *perfect* to maintain all HS and instants. Any more than that, and you aren't losing rage to heroic strike cost, your losing it to capping out at 100.

As you move down from the perfect rage situation, you need to consider the actual cost of heroic strike and the benefit. As the character of an encounter leans more towards where you want better threat efficiency, you are using heroic strike less and less, and the returns from imp hs decrease.

A tier 5 or 6ish tank is going to gain 6 to 9 rage per white swing, making a heroic strike cost a total of 18 to 21 rage per. Full on 3/3 imp HS means that, every 5 or 6 heroic strikes, you save enough rage for one additional heroic strike. In a less rage generous situation, at a 50% mix of white hits and heroic strikes, that will come out to be about one heroic strike every 18 seconds, or 16 TPS... for 3 talent points.

There are better places to spend those points, unless you are absolutely dying for AM. I used to be a big proponent of AM when I was still being weaned over from my 2-hander, for just the reasons listed -- but with a little more experience, its much easier to identify the situations where you are going to wish you had AM, and conserve bloodrage for the moments where I need a boost.

Between imp demo shout, imp defensive stance, the new dont-leave-home-without-it imp sunder, and (perhaps somewhat luxurious) customization talents like tactical mastery, imp revenge, and imp taunt, I can't see fitting more than the necessary 8 points into arms as a full-on protection warrior.

 

Fellwraith

Technically it's also affected by focused rage, which could reduce the net cost of heroic strike by another 3. You can get heroic strike down to a net cost of 9 rage + the cost of the white swing. Realistically, HS would cost about 15 rage in that scenario.

I don't think it's possible to generate 9 rage a swing with a standard tanking weapon. You'd have to crit, wield a very slow weapon, or have an inordinately high amount of attack power. With 1200 AP and a brutalizer, you'll get about 6 rage on a swing that doesn't crit, whiff (miss, dodge, parry), or glance.

The main benefit of having imp heroic strike is that you can maintain a max TPS cycle when you are taking less damage (Shield block, shield slam, revenge, devastate x2 and HS on every attack). I think most warriors understand HS is a rage dump, but it's also the only way to push more threat out in a shorter time frame (and sidestep some of our GCD limitations). It's horribly inefficient from a rage perspective but incredibly useful from a time perspective. Just because the ability isn't rage efficient to begin with, doesn't mean talents spent to increase that efficiency are a waste.

The threat and TPS contribution will vary based on your gear and the encounter. In my case, the two points I have in imp HS mean I can take 824 less damage every 6 seconds and still maintain a max TPS cycle. Anger management means I can take another 220 less damage. For 3 talent points I can maintain a max threat cycle and take 1,044 less damage every 6 seconds (174 less incoming dps). That's 8% less incoming damage required than what I would need if I didn't have those talents.

This probably only matters to you when you overgear the encounter and are farming the content, but that's the time you'd be looking at builds that are geared more toward max threat vs max survivability anyway.

 

stampy

To clarify, I was starting with the total amortized cost of a heroic strike, with focused rage, without imp HS... focused rage is pretty much a given. And yeah, I may be off on the rage... I don't tend to pay that much attention to how big my white hits are; and I'm on the road without WoW access... and the armory is once again useless.

But, my point remains that imp hs is marginal improvement. I cooked up a spreadsheet to see what the actual TPS gain is in a very simplified environment, assuming that the tank is hitting exactly every (weapon speed) seconds for the same amount of rage, and is taking a perfectly even amount of rage per second from incoming damage. The tank maintains a shield slam/revenge/2*dev rotation, and optionally a shield block every 4.5s, but will heroic strike whenever it is swing time and he has enough rage that he wont starve out the instants (and block if so desired).

The improvement just looks at how much threat you get from burning off excess rage with 12-point base cost HS vs. 9-point base cost HS, but quite importantly, capping the HS rate at swing speed.

For the sake of example, I assumed a 1.6 speed weapon producing an average of 4 rage per swing (that means the average swing, including misses/glances/crits, is somewhere around 90 damage). I put it on the fast and weak end of things, because that favored imp HS the most.



The first problem is that heroic strike caps out at about a 33 TPS improvement, and these results are based on a very weak-hitting weapon, which reduces the rage-burn factor and actually increases the TPS number. Just upping the speed to 1.8 and the rage per swing to 5 (still only 135 average damage done per swing) drops the peak TPS improvement to around 27.

Furthermore, the gain drops off sharply in out-of-perfect conditions. Too little incoming rage, and your heroic strikes are far enough apart that the HS'es come rarely, and you always need to HS a couple of times to enable an extra HS for a threat boost. The extra heroic strikes you get from that saved rage are few and far between.

And, while that nice steep line is nice when you are going up, it ends too fast... in the rare situation where you really do have all the rage you can use, imp hs is totally useless -- you are capped by weapon speed.

Obviously, this is not a perfect model. But, it does give an accurate shape for the real scaling ability (or lack thereof) from imp HS. In live conditions, I can't imagine seeing more than a 20 TPS increase from spending 3 points in improved heroic strike. If I was just getting started and I was hobbling along at 500 TPS, it would likely be welcome; but in fairly good gear, I cant justify spending 3 talent points on something that ups my threat less than a fairly inexpensive glove enchant.

I have to imagine that tanks that see significant threat gains from picking up improved HS, consistently over a variety of encounters, are likely using heroic strike too often and starving out their instants.

 

Xerophyte

I don't think anyone is saying that IHS is very significant as such, but neither is anything else we can spend our last few talent points on (however as Quigon says, there's a huge, glaring exception in IDS if you can't consistently get that AP reduction from another warrior or a hunter pet). I agree it's about the same order of magnitude as glove enchant, however the same goes for 3% crit, 3 rage/second or 3 rage/devastate, which are really the only talents we can directly compare to IHS in a meaningful fashion.

Side note: this could just be me but the threat value you use for HS in the graph looks a bit lowish, checking the TPS spreadsheet (and correcting cell X6 to read "=(V6-V3)/W6" to get the threat/rage right) seems to get a peak of about 3/1.6 * 23 ~ 43 TPS for a weapon of that speed. I could be missing something, 'course. Initial guess at potential source of discrepancy: are you accounting for threat from the damage increase caused by HS removing your chance to glance? Or am I just missing something else here or in the sheet?

 

Whiteknight

I'd be interested in seeing the results of your spreadsheet applied to improved sunder armor.

 

Xav

Why for? For curiosity's sake? Imp. Sunder Armor is a talent that is going to be benefitting you pretty much constantly, in any rage situation. There really isn't any reason someone wouldn't always max it out.

 

Sepulture

The simplest way to compare imp hs to imp sunder armor (or to gauge the value of imp hs) is to look at rage conservation.

On the surface, that would seem to be a pretty simple process where rage conserved = # of heroic strikes x 3 (or x talent points spent). This might even be something you can easily model in a spreadsheet and multiply by some tpr number (such as 23).

The problem is that the applied mechanics don't fit well into a spreadsheet. There are just too many variables; anything from an ancestral healing proc to trinket use will affect incoming rage-not to mention boss mechanics.

When you simplify things, an improved Heroic Strike really has two states, efficient and inefficient. Yes, there can be partial efficient states, but let's just exclude this fact.

We can call efficient 1 and inefficient 2.

There are three timeline events which you would need to record, along with two variables.

Timeline events:
A Reach 100 rage
B Reach 9 rage (not enough rage to shield block, and therefore a good point to assume interruption of 4x cycle)
C Heroic Strike performed

If C happens between A and A, then C = 2 (inefficient)
If C happens between B and A, then C = 2
If C happens between B and B, then C = 1 (efficient)
If C happens between A and B, then C = 1

So, you would need to record just two variables: warrior rage level and a timestamp any time rage levels changed. You would also need a combat log.

Find your A's, B's, and C's in the timeline, Add up your 1 states, multiply by 3 (or talent points spent), and divide by the timeline's total seconds and you have your approximate effective rps gain from imp HS.

Now, what that's worth in threat is anyone's guess. You can't assume that this conserved, efficient rage was used to swing another HS. If that conserved rage was realized right before an avoidance streak, it might have meant a shield slam or revenge, or perhaps even a shield block or spell reflect.

I'm not sure that there is really a way to objectively answer this question without combat logs and a lot of hard work. The logic above might get you close enough to make a rough guess- and the result would still be biased based on your playstyle (how trigger happy you are with HS-do you tend to screw up and sit there in a 100 rage state? Do you overdo it with HS and end up rage starved frequently?).

Edit: Come to think of it, if I had a mod/addon that recorded rage levels, there are better, more practical applications. You could name the mod "bad tank" and have it track how often you sit/hit a 100 rage state, or dip to the dangerous 9 or less area. Shield block uptime % anyone? Hmm..

 

◊ Quigon

Stampy where did you come up with those graphs? I'd like to see the equations you used.

Imp HS doesn't improve your TPS unless you have modeled for real-world circumstances, of limited rage and burn-off or some pseudo rotation with a cap.



I realize that in the game imp HS will improve your TPS, but it DOES NOT change your maximum TPS.

Furthermore - the cost portion of an HS is already well modeled in the excel threat calculator spreadsheet... it takes the cost of an HS and subtracts it from the negative value of the rage generated by an incoming white damage attack.

What it does not do appropriately is subtract the threat from the HS from the white damage that would have hit otherwise.

This adjustment was something you did in an earlier update Stampy.

I just don't think it is appropriate to show a graph without math to support it. Even if you explain it loosely - I could draw a graph that shows me winner Powerball.

I mean that just BEGINS to sum it up...

 

galzohar

The general idea of his graph was pretty much what I was expecting, really. It's pretty obvious.

Note that the graphs are "additional" threat due to imp heroic strike.

At low rage levels, you don't use heroic strike at all, and thus don't gain any threat from it.
At very high rage levels you have unlimited rage anyway and can't use heroic strike.
In the middle, the more rage you get the more HS usage you get, up to a point where every hit is heroic strike. Past that point your TPS with heroic strike remains the same while your TPS without heroic strike goes up, quickly, until they're equal at the point where every hit would be heroic strike without having imp heroic strike.

What I would like to see modeled is how avoidance and dodge streaks affect the optimal threat cycle. A boss might have enough dps for unlimited rage but no matter how much dps the boss is doing you could possibly have a dodge string that will dry your rage off. Therefore avoidance has an additional effect on improved heroic strike beyond simply taking less overall damage. This also might change the optimal threat cycle at certain rage levels - if the boss isn't hitting *too* hard but enough for you to generate rage faster than you can spend on a no-dodge streak, is it really good to hit HS when you're at 100 range with high avoidance? It probably is, but at some point of very high avoidance it could stop being true if you're trying to do as much threat as possible... And improved heroic strike may have an effect here, even though on the graph it would have no effect obviously as the rage/sec you receive is higher, on average, than what you spend.

As you can see, the DPS the boss does is far from the only relevant factor for how effective rage reduction costs are and when you should actually heroic stirke. After all there is *always* the option for you to have an avoidance streak and run out of rage no matter how much dps you're taking which means improved heroic strike will always provide additional tps, even though I would expect the added TPS to be extremely small with realistic avoidance levels on very hard hitting bosses.

The end is that no matter what situation you're in, if you use heroic strike then improved heroic stirke will increase your tps. By how much is a compeltely different story.