Beta 4 Elite Specialization Specter Balance Notes (11/24/21)

[Lucinellia.9247]

12 hours ago, Zacchary.6183 said:

So nobody is going to touch on my math? I see how it is.

One of the problems with what you are suggesting is that it seems to presume that Guild Wars 2 is an environment that supports triage healing where you are given sufficient time of people being at depleted health to gradually heal with abilities like the new Consume Shadows and repeated Infiltrator Strikes. Current damage patterns, and healing output from other healers, means the environment is one that favours quick spam healing (further compounded by the lack of resource-based healing, we don't need to manage mana, compounded yet again by health-based DPS modifiers and then yet again by passive / strike-triggered health restoration traits, signets, ascended food etc). 

 

In such an environment, effective HP heals (absorbs, barrier, however you want to call it) are better if you require healing as the "healing" applied via barrier will always be consumed prior to the healing provided by topping up health pools and it is possible to heal entirely with eHP increases before any direct healing can take action. With such scenarios of damage patterns, a heal of 22k is incredibly niche. Instead, Endless Night is still your most valuable tool due to the barrier, which means you are still going to be targeting through different allies to apply it.

For examples of this see WoW 3.3.5, the majority of WoW Mists of Pandaria due to the smart heal and absorb design of Discipline Priest and training / progress groups in GW2 preferring Scourge over Tempest for support. Effective HP is always more valuable in scenarios that are not triage based, and I don't see GW2 ever becoming triage based without a limited resource for accessing heals on all professions that can heal.

[Zacchary.6183]

10 hours ago, TwiceDead.1963 said:

Because it's just not impressive. You are proud of jumping hoops to do what? Provide alacrity and do terrible HP/s? You are certainly not doing damage, who are you trying to convince here? Drop a healer for your heals? Drop a DPS for your damage?

Not even touching Quickbrand, in the time you have unleashed your burst-heal, Scrapper has done equal or likely more of your healing numbers and provided protection, quickness, regen, might, swiftness, vigor, stability, varying combo-fields, superspeed, projectile denial and can even revive multiple people simultaneously should that be necessary. You are certainly not taking THAT spot. They can even reliably tank with high toughness.

In the time you have spent providing Alacrity, and building Life-force, Renegade has pressed one button and outshined you massively in both the Alacrity Generation and the Damage department.

Meanwhile you are busy spamming wells that do nothing for damage, nothing for boon-refreshing (everybody already has every boon so there's nothing to add), do nothing for condition application (the boss already has every condition conceivable), ALL to keep up Alacrity. When you are not doing that you are spamming infiltrators strike for healing and life-force to eventually release a burst-heal that is your ONE saving grace in this circumstance.

With a power build? Because healbrand needs to stack healing power and concentration, while specter just needs vitality.

Edit: Lol, according the build on metabattle the most healing I see on Healbrand is from Tome of Resolve, staff skills and regen spam. The tiny bit of healing it puts out with symbols is purely there for minmaxing. Best healing output it can do with a single action is with radiant recovery and that is only IF it removes conditions with its max potential at 33k hp on a 50s CD. Hitting Eternal Oasis may cleanse conditions, but then your healing potential with Radiant Recovery drops sharply unless there is a steady supply of conditions being applied for the duration of the healing, so that means you'll need to use Desert Bloom which heals for less than half of RR but without conditionals. Staff has chunk healing with Empower and Holy Strike and shield has Shield of Absorption. And while this is all great, Guard has to equip full Harrier, maximize outgoing healing potency, maximize boon duration and rely on stat boosts for damage to get the numbers it gets. The amount of commitment needed to having an effective healbrand is harsh compared to specter imho.

The placeholder build I posted for Specter has very minimal commitment to get the numbers I posted. For starters, minimum number of shadowsteps to fill force is 10 (in case you miss), but you also have siphon. You get max 50% force from using 5 wells on the stack, 25% from Siphon on enemy (and a well going off CD), another 25% from using Infiltrator's Strike/Shadow return on top of the stack to fill up force in ~5 seconds, having already healed 5.8k outgoing. Get max Consume Shadow stacks in 5-6 seconds and then drop for ~20k outgoing, ~26K total in 11 seconds. My remaining resources afterwards are: Full Initiative, One Well off CD, Siphon with a second or two of CD left (if not off CD entirely). Continuing on, I could use the one well and IS/SR until initiative is out, siphon and then another IS/SR or two to fill force again. By this time Shadow Shroud should be off CD.

So lets say for argumentative sake that it takes me ~10 seconds to get through the second fill and each subsequent fill would take on average 12-13s on a perfect day. One round of healing for the average specter could probably take a comfortable 15-20 seconds assuming minimum issues. And with Marauder/Diviner with Vampirism runes, your total heals per 20 second round will be around 26k. Let's also add Shadestep, because we can, allowing you to grant yourself 2k barrier every 10 through Siphon and then give it to allies when you shadowstep.

[TL;DR] So the final damage is:

Healbrand (from MetaBattle)

• Full harrier required
  • Best stats to use for healing and boons without sacrificing too much damage.
  • Overall damage is reduced due to almost no precision and ferocity in the build and relies on boons and other boosts to compensate.
• Staff required
• Heavy commitment to healing output required
  • Max additional healing output: 88% outgoing healing effectiveness, 55% without Eternal Oasis.
  • Rotation involves three different skill sets.
• Best heal is a Heal-on-Cleanse at 3,993.
  • Requires three conditions to heal max amount.
  • Only has five uses before put on 50 second CD, healing for ~33k at best.
• Most heals involve many tiny heals with some 2.5k+ chunk heals in between.
• Otherwise, still best boon coverage.

Specter (with placeholder build)

• No specific stat set required
  • Traversing Dusk has a base heal of 645 and Specter has huge access to shadowsteps to take advantage of it.
  • Consume Shadows scales off Shroud HP, which is Max HP x1.5, making vitality only required for better healing bursts.
• Sword and Wells are required for fast force generation via Traversing Dusk to access Consume Shadows
  • Infiltrator's Strike and Shadow Return grant a total of 25% force max, 10% for each shadowstep and 5% total for initiative use.
  • Wells grant 10% force max each due to shadowstepping.
• Minimal commitment required
  • Traversing Dusk and Consume Shadows are the only hard requirements.
  • Concentration is a soft requirement for more alacrity.
  • Healing power is useful but not necessary and scales poorly on thief.
  • 20% outgoing healing is baseline. More can help but vitality can be better.
  • Light on initiative use.
  • Shadow Shroud only necessary for healing bursts.
  • Rotation involves keeping wells and siphon on CD.
• Best heal is Consume Shadows at 20k every 10-20s.
  • Vitality and health boosts are the only thing that will increase it's healing, allowing Specter healing to still function with power builds.
• Most healing is done through shadowsteps and Consume Shadows.
• Otherwise, has excellent barrier application
  • Shadestep let's siphon grant you a 2k barrier and allows you to then transfer it to 5 allies when you shadowstep.
  • Consume Shadows grants barrier if you overheal.

@Kain.2310, @TwiceDead.1963

With just a little bit of math and theorycrafting, I made healbrand into a just-boonsbrand. And Kain, a bit of wisdom: Practice what you preach.

Edited November 28, 2021 by Zacchary.6183

[Shiyo.3578]

On 11/24/2021 at 8:09 PM, Zacchary.6183 said:

Plenty of people said this about Deadeye.

Deadeye was a dead on arrival meme spec and specters looking to be the same. Does Anet not realize that Renegade exists and also gives alacrity?

[Blood Red Arachnid.2493]

4 hours ago, Shiyo.3578 said:

Deadeye was a dead on arrival meme spec and specters looking to be the same. Does Anet not realize that Renegade exists and also gives alacrity?

The point of these new specs isn't necessarily to supplant the old ones, but to give mono-players more viable options in groups.  I.E. though quickness scrapper isn't meta, I have seen a couple get into raid groups who just needed quickness.  The whole game isn't raids, and it isn't a bad thing to give players more options with their favorite professions.  Also, the future isn't set, so we may very well see the day where Renegade's alacrity becomes 5 target for nebulous reasons.  

[Zacchary.6183]

14 hours ago, Lucinellia.9247 said:

One of the problems with what you are suggesting is that it seems to presume that Guild Wars 2 is an environment that supports triage healing where you are given sufficient time of people being at depleted health to gradually heal with abilities like the new Consume Shadows and repeated Infiltrator Strikes. Current damage patterns, and healing output from other healers, means the environment is one that favours quick spam healing (further compounded by the lack of resource-based healing, we don't need to manage mana, compounded yet again by health-based DPS modifiers and then yet again by passive / strike-triggered health restoration traits, signets, ascended food etc). 

 

In such an environment, effective HP heals (absorbs, barrier, however you want to call it) are better if you require healing as the "healing" applied via barrier will always be consumed prior to the healing provided by topping up health pools and it is possible to heal entirely with eHP increases before any direct healing can take action. With such scenarios of damage patterns, a heal of 22k is incredibly niche. Instead, Endless Night is still your most valuable tool due to the barrier, which means you are still going to be targeting through different allies to apply it.

For examples of this see WoW 3.3.5, the majority of WoW Mists of Pandaria due to the smart heal and absorb design of Discipline Priest and training / progress groups in GW2 preferring Scourge over Tempest for support. Effective HP is always more valuable in scenarios that are not triage based, and I don't see GW2 ever becoming triage based without a limited resource for accessing heals on all professions that can heal.

I get what you are saying, but unless the group has a lot of HP, that 22k burst is going to provide full barrier anyway (anet really needs to make Overheal a mechanic lol). There is also Shadestep's barrier granting ability, which should provide enough cover between Siphon barrier and CS barrier. As for healing, using IS/SR might not provide as much individual healing and barrier as Endless Night, but it provides much more force and total healing for the group. Since almost all of your heals are going to be AoE, it would be more cost-effective to heal 7,740 group HP + Shadestep Barrier than hit multiple single persons with a 2,581 heal and a 4,515 barrier.

[Lucinellia.9247]

2 hours ago, Zacchary.6183 said:

I get what you are saying, but unless the group has a lot of HP, that 22k burst is going to provide full barrier anyway (anet really needs to make Overheal a mechanic lol). There is also Shadestep's barrier granting ability, which should provide enough cover between Siphon barrier and CS barrier. As for healing, using IS/SR might not provide as much individual healing and barrier as Endless Night, but it provides much more force and total healing for the group. Since almost all of your heals are going to be AoE, it would be more cost-effective to heal 7,740 group HP + Shadestep Barrier than hit multiple single persons with a 2,581 heal and a 4,515 barrier.

Yes, but you are going to be able to access more useful healing via Endless Night.

I don't really understand where you get the idea of "force" from - mind explaining that? Is this Shadow Force? If so, we don't even need to worry about that because Consume Shadow in its beta 4 form is not useful (see below).

The other thing to keep in mind is that Measured Shot / Endless Night taken together has a targeted heal from MS, splash heal from TD and SS on MS shadowstep, splash barrier from Shadestep and then you have a very high amount of targeted barrier, quickness and regen from Endless Night. Endless Night also does a lot more barrier than 4,515. When I was testing it with a Celestial build, it was closer to 8k. That is such a strong mix of healing and eHP that IS just isn't going to compete.

 

My biggest issue with Consume Shadows as a heal is that the time you spend in it is dead time due to Shadow Shroud skills being pretty awful. This means that the time you spend building the heal, the party is likely recovering health since this isn't a triage environment.

The only real situation I could see where it would be useful in PvE is going to be situations like CA 25% immune phase. You could sit in Shroud and build for the barrier and recovery heal but that is a pretty limited use case. In moment to moment combat, we are still going to be switching targets for MS and EN.

 

It will be interesting to see how Anet responds. Endless Night was nerfed to try and get us to stop clicking different targets in a PvE environment but the designers don't seem to understand that if you make very strong single target healing, the healer will want to click through different targets and use that heal. This is really basic MMO design and I am kind of shocked that Anet is falling for this obvious design trap.

Edited November 28, 2021 by Lucinellia.9247

[Lucinellia.9247]

11 hours ago, Zacchary.6183 said:

With a power build? Because healbrand needs to stack healing power and concentration, while specter just needs vitality.

Edit: Lol, according the build on metabattle the most healing I see on Healbrand is from Tome of Resolve, staff skills and regen spam. The tiny bit of healing it puts out with symbols is purely there for minmaxing. Best healing output it can do with a single action is with radiant recovery and that is only IF it removes conditions with its max potential at 33k hp on a 50s CD. Hitting Eternal Oasis may cleanse conditions, but then your healing potential with Radiant Recovery drops sharply unless there is a steady supply of conditions being applied for the duration of the healing, so that means you'll need to use Desert Bloom which heals for less than half of RR but without conditionals. Staff has chunk healing with Empower and Holy Strike and shield has Shield of Absorption. And while this is all great, Guard has to equip full Harrier, maximize outgoing healing potency, maximize boon duration and rely on stat boosts for damage to get the numbers it gets. The amount of commitment needed to having an effective healbrand is harsh compared to specter imho.

The placeholder build I posted for Specter has very minimal commitment to get the numbers I posted. For starters, minimum number of shadowsteps to fill force is 10 (in case you miss), but you also have siphon. You get max 50% force from using 5 wells on the stack, 25% from Siphon on enemy (and a well going off CD), another 25% from using Infiltrator's Strike/Shadow return on top of the stack to fill up force in ~5 seconds, having already healed 5.8k outgoing. Get max Consume Shadow stacks in 5-6 seconds and then drop for ~20k outgoing, ~26K total in 11 seconds. My remaining resources afterwards are: Full Initiative, One Well off CD, Siphon with a second or two of CD left (if not off CD entirely). Continuing on, I could use the one well and IS/SR until initiative is out, siphon and then another IS/SR or two to fill force again. By this time Shadow Shroud should be off CD.

So lets say for argumentative sake that it takes me ~10 seconds to get through the second fill and each subsequent fill would take on average 12-13s on a perfect day. One round of healing for the average specter could probably take a comfortable 15-20 seconds assuming minimum issues. And with Marauder/Diviner with Vampirism runes, your total heals per 20 second round will be around 26k. Let's also add Shadestep, because we can, allowing you to grant yourself 2k barrier every 10 through Siphon and then give it to allies when you shadowstep.

[TL;DR] So the final damage is:

Healbrand (from MetaBattle)

• Full harrier required
  • Best stats to use for healing and boons without sacrificing too much damage.
  • Overall damage is reduced due to almost no precision and ferocity in the build and relies on boons and other boosts to compensate.
• Staff required
• Heavy commitment to healing output required
  • Max additional healing output: 88% outgoing healing effectiveness, 55% without Eternal Oasis.
  • Rotation involves three different skill sets.
• Best heal is a Heal-on-Cleanse at 3,993.
  • Requires three conditions to heal max amount.
  • Only has five uses before put on 50 second CD, healing for ~33k at best.
• Most heals involve many tiny heals with some 2.5k+ chunk heals in between.
• Otherwise, still best boon coverage.

 

I think you are misunderstanding what makes Healbrand so strong. Healbrand certainly doesn't need full Minstrels to succeed - you can run Celestial and Seraph variants for more offensive builds without issue and a Celestial Firebrand definitely does more DPS than a Celestial support orientated Specter (as of the previous beta, we need to see how things are on Tuesday but with Endless Night being single target now for Rotwallow Venom, I can't see how the offense of support Specter goes up).

 

It doesn't matter that Firebrand cannot do such a large singular instance heal as Specter because, as I noted in my other post, GW2 healing is not a triage environment. The damage patterns do not reward, or even allow, for large singular healing to be important. And despite GW2 not being a triage environment we still have barrier and Aegis which really minimises the usefulness of raw healing.

 

Firebrand excels because of the sheer amount of Aegis that is being put out. Aegis is stronger than barrier (which, in turn, is better than raw healing for healing priority) and Firebrand can apply so much of it that the raw healing doesn't really matter. Add to the Aegis the ability to cover a huge number of damage reducing boons when needed (Tome 2, Tome 3- and this is without a 5 second windup where you cannot do any substantial healing or sustain) and it is just far better than what Specter offers. Why even worry about maintaining barrier or topping up health pools when you can just block things?

 

Theorycrafting is fine, and it definitely has its place, but it doesn't reflect how things work in practice. Take Rifle Deadeye for example - it looks like it should be a very strong performer in raids on the simplified situation of a golem, yet due to how fights work and the limitations of how damage is dealt it doesn't compete.

Edited November 28, 2021 by Lucinellia.9247

[Lucinellia.9247]

9 hours ago, Shiyo.3578 said:

Deadeye was a dead on arrival meme spec and specters looking to be the same. Does Anet not realize that Renegade exists and also gives alacrity?

 

Renegade and Mirage definitely need their alacrity share reduced to five targets otherwise all of the EoD alacrity share specs are dead on arrival. Renegade then seriously needs its DPS reduced, condi RR basically removed and trade-offs for the high amount of CC it can offer.

[Graeaw.6329]

Okay, let me put it like this: Does firebrand need to plan 5 seconds in advance and change their gameplay accordingly, when they anticipate a big incoming hit? No. 

[Zacchary.6183]

11 hours ago, Lucinellia.9247 said:

I don't really understand where you get the idea of "force" from - mind explaining that?

Yes. Traversing Dusk makes you heal in an area whenever you shadowstep. It also gives you shadow force for every person you affect, minimum 5% force for use and then 1% per affected target. You also get 1% force for every point of initiative use. Using IS/SR on the stack gives you 25%. That is much more force than you will get on Specter beyond Siphon.

Also you keep saying triage environment in regards to the build when aoe healing does anything but. So long as people stack, they'll get healed. There is barely any selection involved. And even then, we don't even know what kind of combat mechanics anet will introduce. It is entirely possible that there will be enemies and bosses that will take huge, possibly unavoidable/unblockable, chunks out of the party. Healing checks are a thing and are easily implementable. Considering anet has given a lot of support and mobility to EoD elites, it isn't a stretch to say they have this planned for EoD. Something like Consume Shadows would be great for heal checks, either to cushion massive hits or to counteract them when they happen. You could also use it to instantly soft-revive downed players.

So I am having a hard time trying to understand how such a large heal after some smaller heals has any downsides. Specter can afford to waste the large heal because it is easily and quickly attained, entirely saveable and it overheals. Firebrand does not grant barrier so having something like this on them would be a waste, they are better off with chunk heals anyway. Regardless, Specter isn't Firebrand.

11 hours ago, Lucinellia.9247 said:

Theorycrafting is fine, and it definitely has its place, but it doesn't reflect how things work in practice. Take Rifle Deadeye for example - it looks like it should be a very strong performer in raids on the simplified situation of a golem, yet due to how fights work and the limitations of how damage is dealt it doesn't compete.

Part of theorycrafting involves testing and part of testing is getting data (numbers or benchmarks). You can safely separate the numbers from circumstance to quantify what which build can do. Afterwards, the issue becomes "Skill In to Power Out" which is entirely relative to the player and the situation they put themselves in. Rifle DE has many built in damage boosts and M7 which covers them in lasting boons. On trash, you can always max them all out and you do not need to commit to them to get to that point, but that boon application is selfish. This means that the DE doesn't need to stay on stack, which is great for mechanics that require a split and allows it to find better positions. But it isn't raid-worthy because it isn't max damage and it separates itself from the meta, even though it could still be useful in a raid granted the Deadeye knows what they're doing.

This goes back to my point of the community's role in the success of elite specs.

11 hours ago, Graeaw.6329 said:

Okay, let me put it like this: Does firebrand need to plan 5 seconds in advance and change their gameplay accordingly, when they anticipate a big incoming hit? No. 

Who said anything about planning? You build it up and then use it sometime before someone drops. That isn't hard. ;P

[nopoet.2960]

My biggest criticism of the spec is not mechanical or interest.  I'm just not convinced its needed.  Specter seems to be a solution looking for a problem.   

[Vidit.7108]

The value consume shadows had was not in how much it could heal. It was good that it was a lot of healing, but what was more important was how fast it could release that healing to your allies. 

In this game, in the difficult content anyway, if a healer sees someone is at half hp and waits longer than about 3-5 seconds that player is probably dead or will be dead as you heal them. For them to avoid being dead in that time they have to dodge or have barrier/aegis or some higher damage mitigation beyond protection (though sometimes protection/resolution is enough). Sometimes passive heals can bring them back up enough that they survive, something Specter doesn't really have. Some encounters are really crazy with the amount of damaging attacks, I've seen people go down so fast in fractals it looks like one hit killed them.

Guardian has an elite signet that's a full 100 % heal and takes less time to cast than entering shroud and waiting. It's a neat skill, I affectionately call it training wheels because it's usage is really narrow and I take it out when i can't rely on new players to stand still. Less often because I'm struggling to keep the group alive, but sometimes if i need a revive skill. 

I'm not mentioning Guardian because it's a good healer that's busted please nerf, it has that skill and no one is saying it's the best healer in the game. You take Guardian out of this game entirely and there are at least three other healers I'd rather have in fractals before I prefer the Specter, because the field of healers in this game is really really competitive. 

Changing consume shadows took a great option off the table. The endless night target reduction was another great option that has to be replaced with much slower barrier producing skills and combinations. Healing in this game is about having a lot of options and ways and means of healing. If your options are slow or require set up or caveats they are less reliable. There could be a build that only cleanses condis by blasting light fields, it doesn't mean it's better than another option, and having it is better than not, but needing to rely on it over something simpler is not a good option.

Specter is severely lacking in utility because of the 3 wells for alac uptime, anet seemed to think the novelty of single target healing would be great enough to justify this. Maybe it would be if changing target were easier, I highly doubt at this stage that it will be a reality any time soon even if they are working on a way to make it easier. 

[RyuDragnier.9476]

2 hours ago, nopoet.2960 said:

My biggest criticism of the spec is not mechanical or interest.  I'm just not convinced its needed.  Specter seems to be a solution looking for a problem.   

Because we DO have a problem. Thief can only do damage, not support.

Ele? Multiple attunements for support, multiple builds.

Ranger? Druid and spirits.

Warrior? Banners.

Guardian? It is living support.

Mesmer? Mirage and Chronomancer.

Necromancer? Wells and Scourge.

Engineer? Scrapper.

Revenant? Ventari, Herald, and Renegade.

Now compare Thief to all of them. What does it have for viable support? Outside of lolVenoms (only Basilisk Venom is useful, and that's only for Defiance bars), there is Signet of Agility to remove condis  from allies and restore endurance, Shadow Refuge, and Shadow Portal. That is it, there's no real support to be had for Raids, Strikes, Fractals, or WvW. Thief is the ONLY profession that suffers from a lack of support. So yes, there is a problem for which Specter is the solution to...if Anet would let it excel in support instead of nerfing it.

Edited November 29, 2021 by RyuDragnier.9476

[nopoet.2960]

34 minutes ago, RyuDragnier.9476 said:

Because we DO have a problem. Thief can only do damage, not support.

E..

Yeh, I get that.  It's the single target healing/buffing focus that I'm not sure there is a need for.  There hasn't been a need for it since launch. Why now?  I kinda think we are in for a wild ride of nerfs/buffs and a reworks until Specter is "right" because of this.

[Tails.9372]

1 hour ago, RyuDragnier.9476 said:

Because we DO have a problem. Thief can only do damage, not support.

That wouldn't even be a bad thing if there was a meaningful difference in DPS between damage and support which there is not. In fact several full glass builds on thief are not just offering little to no support but also have considerably less DPS than most other proffesions who do bring good group support to the table.

Edited November 29, 2021 by Tails.9372