Affliction Cooldown and Seed Design Issues

Classes Warlock
1

Hello warlock community and hopefully some class developer readers,

I’d like to address a few issues with the design of affliction warlock at this point in The War Within. I think that the rework of affliction has been a pretty good success. The class is easier to play with plenty of room for optimization. We’ve maintained our dot based spec identity as well as our strong funnel profile, all of which I enjoy. Still, there are always ways to improve!

This post is going to focus on two of the affliction cooldowns that pose problems for gameplay in a PvE setting: vile taint and phantom singularity. In addition, I will discuss a notable ramification of the removal of Sow the Seeds from the affliction spec tree. I will discuss each ability in turn.

Vile Taint: Vile taint is the go to choice for mythic plus content. The ability to spread agony to up to 8 mobs is strong and it provides an intermediate cooldown to our stronger 1min profile with soul rot. In certain situations, the application of curse of exhaustion by vile taint is also useful. However, vile taint has significant design flaws.

  • Because Vile taint is target capped to 8, large pulls or enemies that produce large amounts of add spawns often results in the ability missing the primary target. As a 30s CD, there is no way to apply the vile taint dot individually to the primary target and considering malefic rapture scales based on dot number, this results in a large reduction in priority target damage. In addition, vile taint PREFERS targets without agony. This exacerbates the issue because your prio target, which will already have agony, will NEVER get vile taint if there are 8 or more additional mobs without agony that are in the area of effect of vile taint. Examples in the recent past of this behavior: brackenhide hollow Oak mobs would spawn lashers in large amounts and vile taint wouldn’t hit the oaks, Tindral sageswift roots were grouped on top of the boss for efficient cleave meaning it would be effectively impossible to hit the boss with vile taint and also hit the roots, the roots on Queen Ansurek pose the same issue as Tindral.
  • While the application of curse of exhaustion can be useful in certain scenarios, it is more commonly a hindrance to gameplay. Curse of tongues, particularly when using amplify curse, is a strong utility ability for warlocks. Typically, in an m+ environment, a tank will pull multiple mobs and gather them for effective cleave. This provides a good opportunity to apply curses to dangerous mobs with strong physical attacks (weakness) or kickable casts (tongues). However, when playing Vile Taint, these are wasted casts. If these mobs are slowable, which the vast majority in dungeon content are, then casting vile taint when the pack is gathered will replace ALL of the curses with exhaustion, even curses that were applied using amplify curse. Instead, this pushes application of the curses to AFTER vile taint application. Again, considering our spender as malefic rapture scales off of dot damage, every gcd we are spending after vile taint on curse application is a significant DPS loss during our cooldown window.

Both of the above issues have been present since the start of dragonflight when agony was added to vile taint and the slow from vile taint was converted to curse of exhaustion. I’d implore any developers to reconsider this design for vile taint because of how it negatively impacts affliction priority damage profile in large pull scenarios as well as curse utility in 5 man settings.

Phantom Singularity: Phantom singularity is the de facto single target choice for affliction since the addition of the Infirmity talent. This ability combined with infirmity provides a 10% damage increase for a 16s duration on a 45s cooldown. In addition, the dot does decent ST damage, does uncapped aoe damage and applies infirmity to surrounding enemies. It also has a healing component. All of these aspects of phantom singularity on the surface seems like good things. However, there are also issues.

  • The duration of phantom singularity continues to be reduced by haste. By and large, with the exception of a small number of damage amp encounters, this design is anti-synergistic with the affliction warlock toolkit. Infirmity duration is shorter, the time during which more dots are on the target is reduced meaning that malefic rapture damage is reduced.
  • In most settings, the 45s cooldown of phantom singularity is functionally a 60s cooldown because of soul rot and how malefic rapture works. Dot/debuff overlap provides more damage benefit than the potential for additional phantom singularity casts over the course of a fight.

Again, I implore developers to reconsider the design of Phantom Singularity to address these issues.

With regard to Sow the Seeds, I am personally a huge fan of the current rapture based design for ST and aoe. I advocated for this in a previous forum post as well as have talked about it frequently on the warlock discord. However, I believe there is one major growing pain to the removal of sow the seeds - the application radius of corruption from a single seed of corruption is too small. This is very noticeable in m+ scenarios as well as when there are enemies with particularly large hit boxes. Specifically, a single seed of corruption, even when cast on mobs that appear fairly close to one another, will not apply to all of the enemies. And on enemies with very large hit boxes, it appears that the blast radius for seed of corruption application of corruption is from the center of the hit box. So any enemies that are still within cleave range for any other ability but are positioned far enough away from the center of the larger mobs hit box will not have corruption applied to them because the hit box is bigger than the seed of corruption blast radius. This provides significant frustration considering rapture requires mobs to have dots on them for effectiveness and seed is meant to provide a convenient way to apply corruption. This is even more noticeable playing hellcaller with wither application. Considering affliction is already a ramp class, having to spend several GCDs applying corruption because seed didn’t apply to everything is frustrating to say the least.

To reiterate, I like the design of this version of malefic rapture based affliction. I think that the developing team did a good job in the rework of the spec. However, the issues I have mentioned above deserve a second look and some adjustments.

Thanks for your time,
your friendly affliction one trick

12 Likes
2

With exhaustion from VT only lasting 6 seconds, it’s not even particularly good in situations where it should be. You could just completely trip exhaustion from VT with nearly zero downside. Better yet, slap the old Venthyr smart curse onto VT.

I think removing Sow was a fine change, and I also think they could just go a step further and be done with Seed entirely. Just about any other method of spreading Corruption/Wither would be faster with a better ease-of-use, so long as it maintains the 1-shard cost with no cooldown. But we’re stuck with Seed because we’ve always had Seed so we can’t get rid of Seed. Sometimes it’s ok to just trash things that are bad. I promise.

I’m also not convinced PS even needs to be a thing, either. Infirmity could just as easily work the same with VT (it even used to when it was a DF set bonus), and as long as the problem of VT whiffing your prio target was fixed, the only downside would be additional shard cost, which is not a particularly big deal.

While I doubt they’d remove either Seed or PS at this point, having a range increase on Seed’s explosion, and fixing PS’s weird haste issues would be decent band-aid fixes for their current problems.

And please just uncap VT. It can stay as only applying 8 agony and 8 curses or whatever, but the VT dot itself is not strong enough that it needs to be limited to 8 targets too.

1 Like
3

I’m not opposed to any of those changes really. I just wanted to leave it open ended in the original post as to not be too prescriptive.

1 Like
4

Normally I wouldn’t want to give such direct ideas, but some of the designs in Affliction do feel like they’re just there to troll the player.

Haste reducing PS duration, Seed being so small of an explosion, and worst of all, overwriting an amplified satyr/tongues curse with 6 seconds of exhaustion as part of your aoe opener. Any one of these is not good, but having all 3 in the same kit feels a bit like they’re playing a prank on us.

Any fix would be fine, really. I would just like to see something be done about them.

5

I’m not a fan of the funnel niche they want Affliction to be, simply because it results in very poor single-target damage for the sake of balance, and this has been the case since Dragonflight. The problem with this design for a spec is that we don’t always have relevant funnel fights in raids, and even in dungeons bosses since not everyone plays on a MDI group that can pull packs into bosses. In the current raid, for example, Affliction is literally only played in one boss and is mid/bad for everything else that is pure single-target. It’s fine if they want to keep the funnel aspect for the spec, but at least make a strong choice node against Dark Harvest for pure single-target, that would be great.

1 Like
6

I consider this a tuning choice and not a design flaw. i wouldn’t be opposed to a competitive choice for DH or cull the weak. but i genuinely do enjoy the funnel aspect of the spec. i just wish they wouldn’t as a result tune us into a funnel fight niche spec.

1 Like
7

It’s 100% a tuning choice because blizzard is renowned for leaving some talents literally on the floor because of strange design reasons. It’s also because, as alluded to in the thread, affliction is one of the very few specs in the game that consolidated both it’s aoe and single target spender – a good comparison would be something like ignite fire mage currently (even then, flamestrike has it’s mass aoe niche). Something like oblivion not scaling with mastery, as a single target spender, is absolutely criminal in our capstone talent section. I would assume there might be some tweaks coming for underplayed talents in the future for this. Literally adding mastery scaling to oblivion would solve that section of the tree.

I’m also not convinced that blizz cares about giving affliction quality of life changes and continues to double down on wanting dots to mean something when you press them. I think it’s you either make dots hard to put up (unlikely in modern wow, or you borrow the legion playstyle) or you make them easy. My biggest wish for the past few years has been that affliction ends up looking more like unholy dk in the way that dots are applied, and that our spenders either enhance or are scaled by dots. If we could merge shadow priest and unholy into one spec, it should be aff (easily applied dots, tick rate increases, lots of chip damage).

Another interesting talent they could learn from is something like Necrotic plague in the past, and have a talent that when agony ticks it spreads to the nearest target in 40 yards every tick (so you it would exponentially scale) but it wouldn’t refresh your agonies. This could even find a place in the hellcaller talent tree tbh, and just have a talent node there to include a clause that immolate/agony spread to nearby targets when it deals damage. Finding ways to make important conditional damage factors (which are every dot for affliction) easier to be passive in dungeons would massively solve the skill floor for affliction.

I’m honestly just not even sure what they can do to affliction aside from small changes like the ones you mentioned to make it feel better, but it’s going to be languishing in tuning purgatory until they can decide on a way to make dots less relevant in mass cleave.

5 Likes
8

NGL, hellcaller destro feels like a better dot spec than the supposed dot spec aff was meant to be (HC destro’s fun though, for what it’s worth)

The way that MR interacts with the spec is a tad off putting, to say the least. Not to mentin the horrible lack of movement & long set up for burst windows

1 Like
9

Agony’s application methods need a complete overhaul. I don’t think Vile Taint needs to exist. Rather than PS/VT being a choice, PS can become the only talent and a choice can be added with Infirmity—one to focus on ST and one to focus on AOE. Reduce PS for 30 seconds and cut a bit of its power to account for that.

I do agree the problems with PS need to be fixed, and there needs to be another method of AOE and/or automatic cursing.

1 Like
10

Ive been calling for VT cd and application be reworked for along time. Considering agony is our primary shard generator, applying agony should not be limited like it is.
Blizz never really thinks things through. They are more about limiting everything than producing a smooth and enjoyable playstyle.
VT should not have a CD. It should not have a target cap. If this is unacceptable, than just have seed of corruption apply agony along with corruption. I really dont get why it doesnt already. Seed does almost no dmg and all you use it for is to apply corruption. It already has a huge cast time plus a shard cost, so its double taxed. Thats enough of a cost to apply agony. Have VT be baked into something else. Idc. Having to only use seed to apply dots instead of a bunch of different things is totally worth it.

1 Like
11

Excellent post.

I say go even further with VT and PS and use them as means of bringing more valuable group utility and mob control to affliction.

Add a knockdown to VT and a mass grip to PS.

1 Like
12

Soulburn-Corruption/Wither.

Hell, maybe this can be extended to Agony as well? Soulburn-Agony spreads the dot to 3-8 targets? Something like that?

2 Likes
13

Here’s a crazy idea. Get rid of rng shard generation, every other class has consistent ways to build their resources to spend but there have been so many instances of me sitting around waiting for a shard to generate. It’s ridiculously bad game design.

3 Likes
14

Yep.

  1. Rng shard gen needs to go away
  2. Make seed as primary dot spreader for all dots
  3. Make MR passive.

There. 3 ways to make aff stomachable

3 Likes
15

You know what makes this whole discussion on agony shard generation even more halarious, is that agony is actually not a pure rng shard generator, nor is it actually random in any sense. The issue is that blizzard has made the formula for agony so vague and impossible to find that people assume it’s random because the tooltip says “sometimes generates 1 soul shard”.

I’ll preface this by saying this is how it worked, and was explained in the past, the current assumption is that it works the same way, but the numbers may have changed.

Behind the scenes, agony is an accumulated buff that uses a specific number range (say 0.1 - 0.2 for example) that every tick counts upwards to the whole number 1 (with the assumption that it overflows as well) within that number range. So on average you should see a shard every 6 agony ticks in single target. Each tick scales exponentially less with more targets active (therefore each agony is not a 1:1 ratio for shard generation in multi-target situations). It also has built-in bad luck protection so that if you low roll the number range (say 0.1 out of the potential of 0.2/1 total) it will bump you up on subsequent ticks.

Anyway, long story short, either make the formula available for us to see instead of purely speculation, or remove the rng (which I agree with). How they remove the rng would be extremely interesting though, as if we add more raptures to our rotation, it will crowd out our potential gcds for dot maintanance (unless they fix that too). I guess one possibility is to lock agony at giving us 2 shard bits per tick instead of a range of bits (the rng element, and why it isn’t reflected on the ui) and update the ui/ux to represent this change. Would go a long way to making the perception better. No idea how to make it function in multi-target settings – maybe flooding us with shards would be too much (or maybe cut any agony after the first applied to 0.1 bits per tick)?

more math stuff

For more math nerds as well, this is also why you will most likely get linearly more shards from haste instead of significantly more chances through higher tick rates (through exponential haste scaling). This also means that the designer was intent on having 1m shard dump windows, as the assumption is that at 0 haste you are likely to have had at least 1 shard per 12s (6 ticks) based on how high you roll your shard bits. Ideally with haste this should allow you to dump 1-2 extra shards per minute before your 1m shard dump window in single target.

This also means that the inferred interaction of relinquished (agony shard increase talent in our talent sheet) is that the accumalator is set lower say at 0.9/1 instead of needing to accumulate to the whole number 1. This would mean that shards should more reliably come in quicker on average, instead of backloading shard gen due to rng. It smooths the cadence of shard generation since it doesn’t increase the randomness of the shard generation, but makes shards more likely to come from accumlation more quickly (which might give you a 1 shard leeway every 2m). It’s nearly imperceptible though, which is why nobody takes it in pve. This would also mean that this talent also scales linearly with haste, and has more relevance in high hasted settings as you will be able to get your “10th” shard in your minute cycle more quickly as it accumulated more quickly.

It’s actually well-designed, if only the game played like a spreadsheet.

Would actually be interesting if a class designer could explain the formula here so that there’s a more definitive answer on this front?

@kaivax

(if any mods look at class forums at all)

3 Likes
16

Wait… if Afflictions Soul Shard generation works like this behind the scene, why doesn’t Aff just use Soul Shard Fragments? I.e each tick of Agony generates X fragments. With some sort of cap or negative modifier in AoE.

I must be missing something really obvious here.

3 Likes
17

Agreed.
This sounds like white knighter defending blizz stuff. If this was the case then there would be wa counting agony ticks to keep track for how many shards are being generated and how many ticks it is taking to do so. Also we would see the cutoff to how many targets we need to spread agony to before its completely useless to do so.
Either 1. Blizz doesnt release bc it actually doesnt work that way. 2. That would make aff much more understandable, yet nonetheless still completely illogical regarding shard generation.
I agree with the previous statement but i would take it 1 step farther, maybe 2. 1. Have agony AND corruption/wither ticks contribute shard fragments. Unless blizz is willing to fix seed to spread both, its best we get them from both. 2. Crit provides additional shards. 3. The number of dots on targets increases the amount of fragments provided by ticks.

18

Since the shard generation is partial they can’t reflect it on the ui unless they show partial shard bits.

Here I asked my TC contact and they sent me this:

(https://docs.google.com/spreadsheets/d/1A7Q-QZpiWFabB5TPwkOUBLESHKDrboEy3fsKzc6ZdKE/edit?gid=0#gid=0)

Here is what blizzard said previously:

Agony still has a rising proc chance for each passing tick that doesn’t give a shard. It’s currently 2.5% on the first tick, increasing by 2.5% for each tick that goes by without a proc. This increasing value is stored on the player and is shared among all targets with Agony (the effective average proc rate per each tick of Agony is left as an exercise for the reader).

Every tick of Agony, a hidden accumulator is incremented. When it exceeds 1, it generates a shard and rolls over the remainder to start the next one.

The added value per tick is currently 0.16 on average, with some variance (we’ll try to remember to update this since there’s no way to see it in-game). Each added amount is divided by the square root of the number of Agonies you currently have active.

When you have no Agonies out, the accumulator is cleared. When you next cast one, it’s reset to a random value from 0 to 0.99.

1 Like
19

Blizz has released that it worked that way back in legion beta – They actually made a specific post about it (trying to find it now). This is also how it works in practice if you look at average shards per minute through agony ticks. You can also just try in game btw.

This doesn’t work that way since the scaling is logarithmic negatively. Therefore it’s exponentionally scaling downwards. Refer to this spreadsheet for more data it should never reach 0:

(https://docs.google.com/spreadsheets/d/1A7Q-QZpiWFabB5TPwkOUBLESHKDrboEy3fsKzc6ZdKE/edit?gid=0#gid=0)

HOWEVER,

We have found through sims that after 7 agonies you get no benefit any more to keeping them dps wise, and that after 5 it’s relatively close to a loss.

20

Okay, so the entire system is completely illogical and unworkable. Got it.
Lets move on to shard fragments. Sounds more realistic option to me and the 99% of players who dont have a MA in engineering/math/computer science, etc…