Regional FlagBlood - What is the developer's design goal?Source
Target Source
#1 - 2011/03/02 08:07:00 PM
So I see that other tank classes have been getting some insight and developer time about the design goals of all their specs with 4.1 around the corner. Maybe we can too. I have been wondering for a long time what the design intent for some of our abilities have been, and how the developers expect us to be played, so here are some of my observations and thoughts on Blood - maybe we can hear what the developers think of the DK tank spec too.

TL,DR - check out the last sentence of this post.

Survivability resources and the undervalue of everything else
Death Strike and our diseases are our survivability abilities and they compete for the same FU resources. Blood runes are also a source of survivability, although indirectly through Blade Barrier. This sets up a situation where FU runes are highly coveted and where we don't necessarily care how we spend our B runes as long as we can get them on CD for Blade Barrier.

The Pawning Off of Responsibility: the survivability vs survivability game
We can either put up the standard tank debuffs (STDs) via Icy Touch and Plague Strike or use those two runes for a Death Strike. I believe we are the only tank that has to make such a choice and its usually either a non-choice or a choice that we try and pawn off. In the first situation we can look at the math between the choices and choose the path that leads to better survivability - this is similar to how dps would choose between two dps abilities or talents and one would always loose out. In the second situation, we can try and get other classes to bring these abilities for us. I do this all the time, but I have the luxury of 3 hunter pets and a dps dk in my 25 mans. This principle really started to annoy me last night during Heroic Halfus though. I was tank swapping Halfus with another Blood DK and we had it so that the OT would be in charge of maintaining diseases on Halfus while the MT would just focus on Death Strike. This dramatically made it look like we are designed in such a way that not supplying our own STDs is by far the best survivability 'choice' we could make. If a DK wants to maximize his survivability he gets other people to supply his STDs. If say a Warrior wants to maximize his survivability, he can apply his STDs and not take a hit in survivability, he just takes a mild hit in threat. I'm all for not having classes being the same, but is the DK design a good design?

FU rune abilities are loaded with meaning, B rune abilities are whatever we can use to get them on CD.
This lopsidedness isn't my favourite design as I'd rather have all my abilities to have meaning. The most obvious way to shift meaning onto Blood runes would be to get them involved in disease management.

Old Death Knight Abilities - lessons from the past
My suggestion, at the bottom of this post, has been in the game previously, so lets take a look at its origins.

Pestilence - pre 3.1.0
When DKs first came out, Pestilence was designed to spread disease AND do damage. This was changed to its current form in patch 3.1.0 as it was ruled that DKs do too much AoE damage as it was (full strength diseases anyone?) and no one was really using Blood Boil because Pestilence would spread diseases and hit for a decent chunk of damage. The logical AoE application was then to use one Blood rune on Pestilence and then with the fast refresh rate of Blood runes in the old system, spam the rest of them with Blood Boil. This doesn't work as well with the new rune system where one Blood rune will be on CD already, and both are on a relatively longer CD than compared to the old rune system. I don't know the answer, but ask yourself this question, thinking back to WotLK and then thinking about Cata: After spreading diseases with Pestilence, how many times could you spam Blood Boil (using just Blood runes) before the diseases fell off? ...Is Pestilence really a good ability for Blood?

Disease Management 3.1.0 - the glory of Glyph of Disease
3.1.0 blew open the door for easy disease management for tanks with the inclusion of the much coveted Glyph of Disease, and all it did was allow Pestilence to refresh disease on one more target, your primary target. One GCD and diseases were managed easily, which made it a favourite glyph of tanks that had other things to manage other than the burden of diseases.

Community Manager
Target Source
#20 - 2011/03/03 03:28:01 AM
Are you content with the power contained by DS? Right now, that ability accounts for our block and a portion of our armor. Is the design goal of blood to use DS over anything else?

Death Strike is a centerpiece of the Blood Deathknight arsenal. It’s your core strike, and we expect that’s where most of your runes go. Still, it's easy to over-simplify the mitigation picture. Mitigation is always important for a tank, but its value also varies according to the scenario you're facing. If you’re tanking a hard-hitting boss, or a pack of heavy-hitting lieutenant elites in a dungeon, then survival and minimizing damage are definitely top concerns, and Death Strike is crucial there. But there are plenty of tanking situations where you're not in any danger of suddenly dropping dead from a huge damage spike. You might be rounding up loose mobs, for example, or trying to get very solid aggro on something that needs to be DPS’ed hard. In those cases, using other abilities might help you meet the objective more efficiently, and doing so involves making effective choices as the situation demands.

Are you content with the timing constraints on a DK? Premature DS can mean a significant block loss, does the fact that DKs are using a mod to try to guess their shields match your goals? Do you feel that Blood has to pay an overly large amount of attention to their resources to maintain their mitigation?

We changed Blood Shield to stack cumulatively in 4.0.6 to help ease these constraints.

Are you content with the punishment for errors in the DK rotation? DKs that mis-spend a F/U rune are out a descent amount of mitigation. Do you envision mistakes costing DKs that much?

We’re okay with the way this works out. The deathknight class isn't innately easy to play. We start them at high level and require players to have some familiarity with World of Warcraft (in the form of a higher level character) before they can make one. It's not intended as a class for beginners. That said, while the best DK tank in the world can do some impressive things, we don't expect every player to perform at that level to be a good tank. While it might be challenging to execute perfectly, the rotation is also reasonably forgiving, especially for heroic dungeons and normal raids.

Applying diseases presents a clear danger when you do it by taking runes away from Death Strike, are you content with DKs actively striving not to put up diseases or striving to time them when there is a 0 second grace period?

Outbreak helps with this, and having diseases up helps with threat in general. On the other hand, if you’re specifically in a situation where you've got solid aggro on a mob and you only care about your survival, then it's probably a good idea to prioritize Death Strike over disease reapplication. If you need to handle more than one creature, Pestilence can help by letting you use Blood runes to spread diseases.

Are you satisfied with the healing done by DKs? As it stands, DK overhealing increases with boss damage. DKs on hard modes are healing for about the same as DKs just starting in raids, is that satisfactory? Do you think this healing design will scale correctly into future tiers?

DK healing is intended to scale with incoming damage. A heroic raid boss hits much harder than a normal raid boss and the DK will accordingly self-heal more.

Community Manager
Target Source
#21 - 2011/03/03 03:28:29 AM
Are you content with the current clarity of DK design? Many Death Knights are literally playing the class incorrectly and have to use outside sources to learn how to get their mitigation up to par with other tanks.

As I mentioned earlier, playing a DK can ask a lot of the player. One thing that players new to DK tanking should watch out for is that you don't always need to use Death Strike the instant that it cools down. Death Strike’s healing depends on damage recently taken, so it is more powerful when you’ve just taken a chunk of damage and not when you’re sitting at 99% health.

Are you satisfied with the signature strike (Heart Strike) being on such a low use priority?

Heart Strike a good use of Blood Runes in some situations, so it looks okay for now.

Are you content that DKs are willing to dump 5 talent points for lichborne? Are you content knowing that there is very little to do with the remaining 10 talent points after a DK finishes their talent tree?

There are plenty of utility choices available to DKs in all three trees. If DKs want to invest the talent points in Lichborne to help improve their survivability a bit, then that’s fine.

Are you content with RE and other quirky mechanics that DKs are working with? DKs are able to game RE procs by sitting on their blood runes, forcing additional FU runes for Death Strikes. This goes against the design of Blade Barrier, but more importantly is still more powerful.

We anticipated players sitting on some runes in order to make it more likely that runes they want get refreshed by Runic Empowerment. We don’t mind this behavior as a more advanced strategy.

Above all, are you content with the current DK QoL? DKs have been moaning and groaning about a number of things. Many are trivial. Many just want to be heard. Many don't understand what they should be doing. Does the DK tank class feel polished and smooth to you?

We’re pretty happy with Blood tanking overall. Many of the recent changes in 4.0.6 were really just quality of life issues, not overhauls to the core design. The core of Cataclysm Blood design was to take more damage but offset that with more self-healing. That’s not an easy thing to balance, but we think the results have been good overall. We're pleased to see Blood DKs out there successfully tanking all the content.

Community Manager
Target Source
#338 - 2011/03/03 09:41:00 PM
If people are trying for 100% shield uptime, youll always feel like you're lacking. In the end, healers are expected to heal something. If they think you need to be at 100% all the time, then they force you into overhealing. If a healer lets you dip down 25%, they would see you're doing a good deal of healing yourself.

That’s a good point. I think what we were really hoping to address is the perception that a death knight tank must use every single Unholy and Frost rune on Death Strike, and game Runic Empowerment procs to generate even more Death Strikes. If you had to do all that to be successful at all, then that would be a problem. The spec isn't designed with that necessity in mind, though. Blood is expected to use (as an estimate, and assuming effective usage) around 6-7 Death Strikes a minute; any additional Death Strikes are helpful of course, but they shouldn't be mandatory. Blood is balanced against the other tanking specs with this assumption in mind. To be fair, one of the drawbacks of the active mitigation that Blood has, contrasted with the passive mitigation of other specs, is that it does require more work - but the trade-off is that death knights have an enviable array of tools to manage incoming damage and promote survivability. That's not to say that those tools must be used flawlessly - the spec is designed around using those tools at an adequate level. We don't generally balance specs around a theoretical maximum; in many cases, those levels aren't achievable by most human players in an actual encounter and only show up in simulations. Instead, we design around what players can reasonably accomplish.

We also recognize that some of you have concerns about Death Strike overhealing, and that it represents a wasted mastery. We didn't go the route of Death Strike over heals creating a 'bubble', because it seemed that it would create even more pressure to use Death Strike at every conceivable opportunity. Nonetheless, it's not off the table completely, and it's something we might consider for the future.

On that note, Death Runes are tied to Death Strike to promote flexibility. It’s too easy to allow the discussion to center around a hypothetical Patchwerk style fight without external support debuffs, where threat doesn’t matter, and the only relevant issue is absorbing every single point of damage possible. Tanking is dynamic though, most fights are more complex than that and the situation can change suddenly and require immediate reaction. Adds run in, things get loose, you may need to move out of fire and then silence a ranged caster that didn’t move with you, and so forth. Death Strike is a core ability, as noted, but keeping the runes it generates as Death Runes helps make it more likely that when a complication arises that the tank has the right runes for the job.

We're not being forced to choose between threat and survivability. I really don't know of many DK's having threat problems right now (Rune Strike is kinda insane...). What we're being forced to do is choose between surviving the next blow and reducing the bosses damage output for the next x seconds.

When I mentioned diseases earlier, I was addressing the question of trading threat for mitigation. We understand that you need to keep your diseases up for the debuffs they provide. Outbreak does help, but there will also be situations where you might end up skipping a Death Strike to keep diseases up too. In a multi-mob scenario, you can use Blood Runes and accomplish that with Pestilence. In a large raid setting, it if's an issue, then it's pretty likely that the debuffs provided by diseases might be available from other sources, such a Warlock's Curse of Weakness, other DKs, Demoralizing Shout and Thunder Clap, hunter pets, etc. Still, we understand your concerns and that you want to be self sufficient when it comes to applying the debuffs that you need to apply to do the tanking job well, and we'll keep that in mind as we continue to refine the Blood tree.

Community Manager
Target Source
#339 - 2011/03/03 09:41:00 PM
Good feedback does indeed effect change. The cause and effect can be hard to see, but it happens.

That’s absolutely correct. We are sensitive to your concerns and we're certainly open to improving quality of life for death knight tanks. We recognize and appreciate that you aren’t asking just asking for buffs, but rather playstyle changes. Conversely, some of the feedback being offered (e.g. Blood Runes as Death Runes) would just push DKs towards a state where they spam Death Strike and absolutely nothing else, which we don’t see as an improvement.

We also recognize that some players are trying to make DK tanking more approachable for new players, which is a noble goal and much appreciated, but we also think that it’s okay for the class to be slightly more complex and challenging to play perfectly given its high-level nature and requirements. Fortunately, you don’t need to play perfectly to be effective. Like I mentioned in the previous set of posts, we see DKs tanking everything from 5-player dungeons to heroic raid content without needing to do so.

Talent design has been an issue that we've seen expressed as well. Admittedly, Blood-Caked Blade is not a shining exemplar of talent design. Tank trees are quite challenging to design, and all the trees have similar design issues. Talents directly related to survival are mandatory. Talents related to damage output are either mandatory if threat matters, or worthless if it doesn't. Utility talents are either mandatory if they help with control or survivability, unless a bosses are immune to the effect, then they become useless. It's too easy to get a point where when the survivability talents run out and frustration results. We want there to be compelling decisions between talents, but finding that balance is exceptionally difficult, perhaps more-so than for any other role.

On that note, I'm curious: what kind of trade-offs do you guys find interesting when you're selecting talents in the Blood tree? If you had the opportunity to alter or devise talents that appeal to different styles of play, without being mandatory to both (and without cherry-picking talents from other tanking trees, since tanking styles differ so greatly) where would you start?