deleted: clicked post accidentally and too early.
deleted: clicked post accidentally and too early.
I think you're underestimating Scent of Blood in your guide.
First, we will never have enough RP that you can reserve 20 RP at all time while still bombing DC. That's just not going to happen.
In every pair of runesets (21seconds), we will acquire 110 RP from all our strikes (1PS, 1 IT, 2 DS, 6 HS), with 4 empty GCDs before we have to refresh diseases. If we use just 2 RS every 20s (very likely), we will only have enough RP for 2 DC. Also, reserving 20 RP is a very tricky thing - at least not as simple as you make it out to be. Imagine if you have 45 RP, and the next blood rune will only refresh 2s from now. Do you DC? If you DC, you won't be able to RS for 2s rune refresh + 2 more GCD to strike the runes to get 20 RP enough to RS. But if you don't DC, you're seeing a dip in threat for at least 2s. Without SoB, we will see more GCDs filled up with HoW, which generate very little threat if at all.
Second, don't just look at percentage of recount to determine the value of DC/SoB. That's because: (1) If I remember correctly, the DC proc-ed from Sudden Doom is not able to crit and (2) you're already DCing less in the first place. This goes back to point 1: do you actually DC at every opportunity? If you only do 2 DC at every pair of Runesets, you're potentially only DCing 6-7 times a minute - of course the threat value is very low. If that 6-7x a minute contributes 6-7% of your threat, and if 3/3 SoB can increase the frequency to 9-10x a minute, that is a 3% increase in threat, already equal to the value of Subversion without even counting theoretical higher percentage of RS over melee ratio.
Third, 40 RP per minute per point is an underestimation of its uptime. Its uptime is really high, and it can consume the charge from RS - which can't be blocked/parried/dodged, increasing the RP return from the talent itself. In any multimob tanking (2 and above) you can be guaranteed it has close to 100% uptime. With 3.6 speed weapon and 20% haste, we're expecting to hit a single target boss 20x a minute - with 80% uptime, you're expecting 160 RP per minute.
Now, it's possible that you're not seeing the benefit of extra RP in your raid. If you raid with 1x Disc Priest and 3x Resto Druid bombing you with PW:S and Rejuv, you'll probably have enough RP to the point that you forget having Scent of Blood. Or if you can get 100 RP every 45s from AMS (very possible) you'll probably forget having Scene of Blood. But that's anecdotal - not everyone is going to have that kind of composition, and not every boss has that kind of mechanic.
To prove the point, I created a template in Kahorie's DK Simulator, exactly same spec, gear is what I have in Armory, one with 3/3 SoB and one with 3/3 Subversion. The result is SoB template wins out by 64 TPS (8758 to 8694). No big deal obviously as the difference is <1%, but it shows that SoB is better than "giving more DCs and DCs contributes so little to TPS".
You seem very up-in-arms without understanding the point in the first place.
1.) RS >> DC for threat, and for threat per RP. Therefore you should only EVER use DC when you can do so without inhibiting your use of RS. Flat, inalienable statement. Because of that you should *never* be using DC, even if a rune blackout is available, if you have less than 60 RP. That's a choice to support your superior threat tool, and it is a choice. Using Death Coil is an option not a given or unavoidable.
2.) I don't base any of this on Recount. If you are using DC and losing RS procs, then you are sabotaging your own threat. Because of #1, if you get SoB you should only ever be getting more DCs, which *is* more threat, but it just happens to be one source of threat, and not one that I would hitch my star to at the cost of other things in just any build. Some builds make good use of that, and some of those builds even go so far as to use Glyph of Dark Death.
3.) The 40 RP per minute is presented with math in tact. Let's look at the extreme though. If you are using a quicker 2-hander (say the rare 3.2 base speed), 23% total raid buffed haste, that's a 2.48 sec swing time, or 24.16 swings per minute. If you could get 100% uptime on Scent of Blood you'd get 240 RP per minute.
Scent of Blood relies on incoming hits though. If you're up against a single boss target (which is where this will actually matter as trash will not be terribly reliant on RP generation), with a really fast swing timer of 1.0 second *after* de-hasting, that is 60 swings per minute. Scent of Blood procs off of anything but misses. Lets say the tank is only uncrittable with only 6% chance to be missed, so 94% of the swings that are thrown have the *chance* to proc. The ability has a 15% chance on each swing that can proc it. So 94% of 60 swings is 56.4 swings that have a chance to proc, and 15% of those will proc on average or 8.46 procs per minute. *If* in the best case scenario, those procs are equally spaced, they will come 7.09 seconds apart.
If you are getting 24 swings per minute (2.4 seconds apart) you can get 2 swings in between procs. That means that you won't get value from the third point in SoB as you will never get that third swing for 10 more RP. However, with 1 point you will get 84.6 RP per minute (1 swing, 10 RP), and with 2 points you will get 169.2 RP per minute (assuming you aren't missing, parried/dodged).
Typical weapons are 3.5 sec swing time though, or 2.716 sec with full raid haste. That's 22.09 swings per minute. Typical raid bosses will throw swings capable of proc'ing SoB closer to once every 2 seconds. 94% of 30 swings per minute is 28.2 proc'able swings, 15% of those actually will proc for 4.23 procs per minute, thus the 42.3 RP per point. Looking at it like above, the most ideal situation would have those procs equally spaced to make sure you get the most swings in for RP gain. 4.23 procs per minute is a proc roughly every 14.18 seconds. With a 2.716 swing time, you could reasonably get all three swings in between procs and get the full 126.9 RP per minute with 3 points invested.
The truth is though, that is a mitigated value because procs *aren't* evenly spaced. If you get two procs in a row, you've wasted the proc completely. If you get them one swing separated, you've lost 1-2 swings if you have 2-3 points invested. Etc. In addition, in order to get the RP generated you have to connect. If you are hit capped and expertise soft-capped, your melee auto-swings will still fail to connect ~7% of the time which will further diminish your returns. My estimation was actually a little generous, and did not try to factor the reductions, only the gains.
So, the point of all of this is, RS is your priority and you are not wise to use DC if it will make you incapable of using RS. It's simply never worth it. If you believe this, not sure why you wouldn't, then you understand that the only value of SoB is to use more Death Coils.
More Death Coils *will* get you more threat, there's no question about that, but there's a question of how much.
We'll use your rotation, PS/IT/HS/HS/DS > HS/DS/HS/HS/HS bi-phasic rotation, unfaltering. That's 10 moves per 20 seconds, 15 seconds of GCDs leaving 5 seconds open. Technically that's only 3 GCDs with a fudge factor, but that's not quite right, let's break it down properly. Phase 1: 5 moves, 7.5 seconds elapsed on rune abilities, 2.5 seconds remaining for the first to refresh. Those 5 moves generated 55 RP. If you RS once (likely, as you said), you have 35 RP. Not enough for a DC. If you have SoB, the chances of getting a proc in the first 7 seconds is slim, and in order to get that DC in then you'd need to proc it AND use at least one swing. That means you need SoB to proc in the first 3 seconds of the fight, which will be a rare occurance. Phase 2: 5 moves, again, 7.5 seconds elapsed, again, and 55 more RP generated, and we'll say one more RS. Now we have 2.5 seconds again until our runes come up, that's 1 GCD and change, most people will not stall that half second, will use a second GCD and delay their next rune set. Now we have 70 RP after two RSs, and conveniently that will fund 1 DC, not two. In order to get a second (and delay your next rune set by half a second) you will need to have had SoB proc and get in at least one swing before now. In 19.0 seconds that would hopefully be a pretty decent chance, but let's refresh that math. If your attacker swings once every 1.5 seconds (to be favorable), 12 swings, 94% to be proc-capable becomes 11.28 proc-capable swings, with a 15% proc chance becomes 1.69 procs, so maybe as much as 2 procs. That means we most likely have the RP to fill that extra GCD.
That's an opener though. Down the road, using your rotation, we're talking 10 moves per bi-phase, 15 seconds of GCDs used, averaging 5 seconds of open GCDs, 3 GCDs (4 will ruin your ideal rotation if you use that last 0.5 seconds on another GCD). If you are looking to use 3 GCDs per bi-phase, you never use them on HoW or a Defensive cooldown, and you only use Blood Tap for non-GCD abilities, then where does that put us with SoB vs without?
Without, we're saying (anecdotally, but it works) that we get 2 RS per 2 rune sets, and generating 110 RP per two rune sets. That's 70 RP to work into 3 GCDs, meaning we'll only get 1.75 DCs on average into 3 slots. In order to get the full 3 slots perfectly in our ideal rotation, we'll need 1.25 more DCs or 50 RP more. To get that we'll need 5 SoB auto-swings. In 20 seconds with our average 3.5 base speed 2-hander we'll get 7.36 swings on average. In order to get the RP to fill every GCD in that ideal rotation you'd need SoB up 67.9% of the time. With our generous 1.5 sec swinging boss, that's 13.33 swings per 20 seconds. 94% of them can proc, or 12.53 proc-capable, and 1.88 will proc on average. That's less than 2 procs per 20 seconds. *If* you get those 2 procs early enough, *and* you have 3/3 in SoB, *and* you do not waste swings (by proc'ing twice within 5 swings of the boss), then you will be able to fill that last GCD.
All of this becomes horribly representative because we are *tanks* not dps. We have defensive moves eating GCDs and preventing an ideal rotation. We move around, we miss swings because of that, and we rarely keep a static rotation (nor should we) as we adjust to suit the needs of the situation.
It is for all this combined that I don't recommend Scent of Blood to Blood tanks. And all of that was presented (in a little less detail) above.
Do you still think I'm under-valuing it?
It's not the size of your DPS output that matters, it's how much threat it generates. Hands down, RS wins.
It's also not just how much threat you generate, but how much when.
I think it would be really interesting to create a simple red light/green light indicator addon that will let you know when your threat cannot be passed by damage threat. In other words, when the boss has a certain amount of health left, and you have a certain lead over the second highest threat, that highest dps threat could only generate so much threat from damage without bonus threat abilities.
It would be laborious to make one that included all the non-damage threat factors, but a simple one could be nice. I've been tanking and gearing up a second DK on another realm, and plugging my warrior a lot, and laughing at the surge of threat that leaves average pug dps in the dust. Most targets I can just stop doing anything and get a drink while the dpsers finish them off (but what fun is that, then I wouldn't bump dps down the damage meters too). An addon like that could be interesting feedback.
That would be cool, Me i know my raid group and if I wanted I could probably gauge that off of experience. However for pugs it would be great.
No one is doubting that the threat/RP for RS is higher than DC. Let's get that out of the way first. If this is your point - well, I actually got it.
The reason I brought up comparison with DC frequency is that, by sticking to the priority of RS -> DC, we will see higher RS to melee ratio and higher DC frequency on top of that if we have SoB compared to without. The gain from both factors (higher RS and DC frequency) is a lot - and you don't seem to realize it based on the low contribution of TPS from DC (which is skewed on Recount). The higher frequency of RS is not something we can see from Recount, it has to be observed from (1) numerous combat logs with external factors properly considered or (2) simulator - neither of which you mentioned.
When SoB first came out, a lot of people have pointed that the third point of SoB is *probably* not needed, because the gain is small - however it all depends on the experience. A simulator could probably point out the estimated number for each point on single target fight. Still, 40 point/RP is underestimation - because the gain from first point is probably higher than the third, as you showed it yourself. As you RS more often, the benefit of SoB increases, because we remove parry/dodge from the combat table of Main Hand strike.
Also, VB is the only defensive move that consumes GCD - and it cost a rune, so it should never interfere with the flow. IBF, AMS, Blood Tap do not use a GCD. We should have enough RP for IBF and AMS by doing HoW when everything else is down - and we have a lot of blank GCDs. The biggest factor to disrupt our flow is miss/dodge/parry.
Your analysis of the opening is excellent, and I won't argue about it, but it doesn't reflect what I'm trying to say. I'm not saying that we should DC *every time* the GCD is blank, nor am I saying that I'm *only* trying to fill empty GCDs with DC. That is *not* the point of SoB, although to be fair, I should have elaborated it in my first post.
What I'm trying to say, is the point of SoB, is not just to give you more DC. It allows you to:
1. Increase RS to melee ratio
2. Increase DC frequency
3. help fill the GCD blanks to smooth the TPS curve by reducing the chance that unfavorable cases happen.
I'm pretty sure that these cases happen to a lot of us (albeit anecdotally):
1. 50 < our RP < 60
2. 10 < our RP < 20
In case (1) and (2), our GCD is blank - TPS will dip, and by a lot, especially if the next rune abilities can only come in more than 2s (usually at the second half of every rune set). The point of having SoB is to reduce the chance that such cases occur. By having SoB, let's assume that we get additional 10 RP compared to without, you could DC AND RS on first case, and RS on second case. The benefit of the reduced chances for both cases occuring is higher than you give it credit to.
Let's look at other RP windows:
3. 30 < our RP < 40
4. 0 < our RP < 10
In the case (3) and (4) above, it *looks* like having SoB or not wouldn't make a difference. By having additional 10 RP on both cases, we still can't do anything - the GCD is still blank. However, having SoB or not will make a difference for case (4) in the *next* GCD. If the next GCD we do a HS/PS/IT, we can immediately RS on next GCD in case (4). Also, having SoB will make a difference for case (3) in the *next* *next* GCD. If you have SoB, we're likely to immediately move from case (3) to arrive at case (1), where you can immediately fill all next 3 GCDs while *still* maintaining 20 RP for RS just in case it procs - notice that I still don't change the priority: RS comes first, DC second.
The only case where having SoB or not wouldn't make a difference is where (5) 20 < our RP < 30. With or without SoB, we can only reserve it for RS. Even if the next 2 GCDs generate RP, it's not likely we can immediately arrive to case (1) even with SoB.
The cases where our RP > 60 are trivial. We always DC and reserve the next 20 for RS.
All these cases don't happen *suddenly*. It's a cultivation of a long fight. Having SoB makes these cases cycle faster, and thus increasing the chances of favourable RP window to occur.
How much benefit it gives us in terms of numbers, I'll leave it to simulator - because a simulator considers every cases through its random number generation. The simulator values SoB higher than Subversion in a single static tank/spank fight. The reason I'm using Subversion as comparison is that Subversion is the closest threat oriented talent nearby that is dubious (IMO). 2H Spec is awesome, DRM is excellent, Bladed Armor is pretty much mandatory, whereas Butchery is not worth it.
At least I'll admit that there is a case where the simulator shows that 3/3 SoB is inferior to 3/3 Subversion. That is when we have only 30% avoidance - pretty close to our avoidance level at starting ICC (I'm sitting at around 31%-32% now, nothing 251 or above yet except the ring, but fully gear/gemmed for stamina and ignoring socket bonuses). The reason is that with the reduced chance of RS proccing, the value of having more RP is reduced. However the difference between the two is about 1% TPS difference, with the gap closing as we gain avoidance (at 35% 3/3 SoB is about 0.5% inferior than 3/3 Subversion, at 50% avoidance, 3/3 SoB is about 1% ahead of Subversion). At 30% avoidance level (and without SoB), the TPS contribution from HS is quite close to RS, so it makes sense to buff HS and DS compared to RS and DC.
So yeah, I still think you under-value SoB. On single target fight with 30% avoidance, it's about 1% inferior to Subversion. In multi-mob, the uptime should be close to 100% and RS is procced more often, so it should be higher, though it's probably harder to quantify. If we find the third point of SoB is unnecessary, we could have specced 2/3 SoB and 1/3 Subversion.
Ok, we aren't connecting on one point: if you always value RS over DC and make a point to always have that 20 RP available to use it, then SoB will never increase your RS usage.
The only time, without SoB, that you should be unable to use an RS proc is in the first couple rune sets and SoB will very rarely improve that as it is just as difficult to get the perfect timing of an immediate proc AND a couple swings in before the RS is available.
I made no comment on the simulator directly because that falls under the blanket of, simulator works in an ideal world, does not factor your usage of non-threat rotation items or account for moving, network latency, or any number of other things that will throw you off. The simulator can also perfectly manage things that you will fail to simply for the issue of perfectly timing these GCDs and rune CDs.
Any simulator that puts SoB ahead of Subversion though has me immediately suspicious. Each point of Subversion is +3% crit on HS, which with Might of Mograine represents a 4.35% increase in HS threat per point. I'd be stunned if the conditional extra DCs can really surpass that value in real execution. Perhaps in a perfect simulated tank on a non-dynamic fight, but even then I'd be a little surprised. Simulators are nice, but never make the mistake of trusting a simulator over real execution or the deviations that will result in the quirks and conditions of each fight.
Otherwise we seem to agree in all but conclusion. SoB will give you more DCs and fill more GCDs and we get more threat. Period. It's a gain, it's good. I just add that it is one of the last places I would fill in for adding threat, while you seem to prefer it.
(Horn of Winter takes a GCD, too, by the by)
How are you/we feeling about UH as a tank in ICC. I read your reply to Kensho, and while it seems that the make or break it factor for Bone Shield has always been the ability to avoid hits long enough to extend the duration of Bone Shield well beyond the duration of the other DK tanking tree-specific cooldowns to make UH a competitive tree.
As I see it, threat-wise, UH just isn't monstrously better than the other two, EH is lower, and hell even avoidance is lower, so aside from a desire to rebel, or boredom I'm struggling to find a reason to go UH at this early stage in ICC. Maybe later when more gear becomes available and a solution to getting the dodge rating back into a place it needs to be is arrive up.
I've digressed a bit. Thinking purely by the numbers, with preference aside, does UH still have a place?
Guide looks good, I tend to do all my own things but I do recommend this guide for new players or players who don't have the time to delve into their own theorycrafting, sim running or just beating their heads into the wall to figure it all out on their own.
I want to track Rime / Killing Machine procs a lot easier (specifically Rime).
Does Power Auras do this or do you have another recommendation for that?
Another question... is there a particular reason that the character sheet doesn't show the 3% hit from Nerves of Cold Steel?
I just want to make sure that my assumption is correct that if I'm already hit capped (263), I can reduce my hit gemming/enchanting by 3%?
I'm focusing specifically on special attacks.
[EDIT: Nevermind. I have a shadow priest and just zoned out and focused on this and ignored that she's in the same situation with the hit not being reported.]
Great guide, all the information that was scattered all over the place is finally in one place. Thanks a lot man.
A couple of things,
Firstly the new glyph of Icy Touch with the 20% disease damage increase, is this worth taking for a frost tank? While not significant - it is a flat out TPS increase, however i have not yet seen any numbers to quantify if it is more valuable than other glyphs available for selection.
Another thing - quite often i see a lot of references to macroing your RS into all the other non-RP abilities. I personally never did that and do not intend on doing so for the simple fact that i like to be in complete control of my RP. It is used for other tanking utilities and not just RS + Dump for example, IBF, Death Pact, Mind Freeze. As a tank, you might have to interrupt or pop your cooldowns quite frequently on certain fights. Having your RS macroed into all the abilities might prevent you from being able to do the remaining stuff. I personally find that it might be a much better habit to just get used to clicking it on your own just like any other skill in order to not miss your CDs or interrupts which may lead to a wipe.
From personal experience - unholy tanking is just not worth it at the moment, especially with the Chill of the Throne debuff giving a massive reduction in the avoidance in ICC. After a full clear of the ICC first wing i found that the Bone shield uptime is at the maximum of 15-18 seconds, and that is after i've put on every piece of avoidance gear available to a DK in the current content.
Generally, the idea is an excellent one, however given the current numbers we have in the ICC - it just isn't worth it in terms of suvivability for a tank. Frost or even blood offer much better rock solid EH.
There is, however, in my opinion a good way to use BS - on bosses requiring tank switching. Bosses like Northrend Beasts, Saurfang, etc where the amount of time you are actively tanking is minimized - the effective uptime of the BS rockets really high making it a very valuable tool. The same thing is applicable to offtanking in ICC - having to absorb lashes or cleaves - it minimizes the number of hits you take, thus increasing the uptime of the BS making it an extremely valuable tool.
Since 3.3 the SS revamp, the ST TPS of the unholy build is very solid, i didn't have any issues holding aggrom but overall, i just don't see the worth in it as an end game tank when compared to frost.
As it goes for the huge amount of theorycrafting and stuff in your guide - excellent work, accurate to the point.
One question, here and in many other places I've seen it advised to "macro" rune strike to everything, but is that really a good idea? It seems to me that for interrupt fights in particular (say, Jaraxxus/deathwhisper when the melee interrrupter is running), you would want to have the choice of using rune strike or not. Same with an emergency IBF; if you know "Big Bang 3.0" is coming, you would want to make sure you have the RP ready without worrying about halting your rune ability rotation?
Moreover it seems like frost in particular might like to use frost strike more often instead if they have a large threat lead and are trying to contribute larger damage; I'll also admit to having hungering cold in my frost build since it's mostly my "trash+onyxia" spec while my mainspec blood is for bosses.
Rune striking to everything I agree is a bad idea. I have it macroed to oblit and blood strike only. This way my two primary single target skills have it but I am still able to build rp without problems.
I dont have that problem? I dont feel thats a swell idea. rune strike being the most threat gen attack, should be Spammed as much as possible. Things like blood tap, ERW, horn of winter, along with talents such as scent of blood, chill of grave also generate runic power.
Blood tap 10 runic power, can be used for vamp blood, UB, or bone shield, Horn is good when runes are on cd, or u need extra runic power.
Druids which revitalize, which regens "energy" such as 16 runic power on chance, is usually procing.using UB VB or Bs also generate 10 runic power, seeing how a kick is only 20 runic power which u can prob skip out on a DC or a FS to compromise.
i personally save ERW for things like a O SHIT button, which is comparable to death pact. My cds are generally rotated, with ibf being situational.
So I want to know, via an addon that allows better placement, when I have BLOOD runes available (not to be confused with death runes).
Does Power Aura Classic do this? What about that rune addon in the guide?
I don't use any rune modifications and so I have a tendency to keep using death runes to make death runes. I want to be able to track exactly when I need to BS to convert a blood rune.
I'll catch up!
Fely! Power Auras is great for tracking those, that's what I use, but it doesn't do it all on its own, you have to set up the tracker yourself (though there is an export option, so maybe I'll add some of my auras to the guide). The character sheet only tracks miss chance reduction from hit rating, it does not add in any of the hit chance improvers otherwise (and Spriest's spell is a debuff on the target). Crit chance will also only reflect universal crit chance, but that *is* updated by talents like Dark Conviction and Ebonplaguebringer (note, it is not updated with Annihilation which is melee only, even in the melee tab). Power Auras does not track rune states, or at least I haven't figured out if it does (it was just updated very nicely by the way).
Pavor! Glyph of Icy Touch is nice, and it wouldn't be terrible for Frost or Unholy, however, your diseases will still be a relatively small portion of your threat, so the value won't be as competitive with other options, especially for Frost (with OB and FS). If you're going pure threat I expect you will still find 3 better for each spec.
Unholy is still plenty strong as a tank spec, even in ICC. The thing is it *feels* like you're investing in something that cripples your strongest tool without positive reinforcement. Yes, Bone Shield will not last quite as long, however, you may find it to be more effective at key times. The rapid strikes and stacking damage sources actually really benefit from Bone Shield's coverage. Do I think it's a terrible idea to be Unholy in ICC? No. Will it feel perfectly matched? Probably not. Don't let it make you feel like you *have* to play differently if you like Unholy tanking, but if you aren't Unholy, you probably won't feel much gain from switching to it either.
On Rune Strike macro'ing. This connects to what I was talking about above with Boeten. The only way in which using RS on every available proc, immediately, will compromise your ability to tank is if you are using your RP dump too liberally and running around with no RP. If you follow the simple directive of conditional RP dumping you will never have this problem, ever. That means, never use DC/FS if you have less than 60 RP and never use glyphed FS if you have less than 52 RP. If that is your method you can use it blindly and never see a problem, all that will happen is you *might* forfeit a RS in favor of IBF, AMS, or Mind Freeze.
The problem will not be with the macro but with the player's use of their dump. That said, I think the macro is a fantastic tool for new DK tanks as it will mean one less thing you have to try and pay attention to, but for more experienced tanks it is not a big deal to use it consciously and it may help you have a better sense of how often you RS.
Did you forget about this in your calculation or is my theory knowledge way off?
If the former, this changes your calculations and actually proves your point even stronger against SoB being a viable deep blood talent.