Something I talked about in an earlier post about the goal of my blog was being “constructively dissatisfied” with things – knowing why and what makes you unhappy about a given thing and understanding it completely so you can address the true root cause.
One thing I see quite often in the World of Warcraft community is this sort of angst about how the game is supposedly being crafted for “elite players.” The idea being that all content in the game is designed with a trickle-down approach – Blizzard designs a Mythic raid and a top end Mythic Keystone dungeon, and then tunes the easier versions from that. I have two key issues with that – one, trickle-down anything simply doesn’t work, and that includes in difficulty design in a video game, and two, that this analysis fundamentally misunderstands Blizzard’s real balancing problem.
Because the truth of the current state of WoW is that there is a balancing problem for content, and a fairly wide swath of it at that – which is that Blizzard designs content to be scalable, but fails to properly account for the impact of said scaling. This is compounded by class design, as Blizzard currently has pushed themselves into the unenviable role of designing 36 unique specifications, which supposedly only belong to 12 classes but so little of the toolkit is shared that they are in effect balancing 36 unique classes, with a class design team that operates separately from the team that designs and balances the encounters these players face.
The Challenge of a Four-Tiered Raid Difficulty Design
The biggest misconception I read rather frequently is that Blizzard is catering to the elite players by designing fights on Mythic first and then scaling back from there, screwing smaller guilds and groups. Now, we don’t really know how they actually set about designing a raid, as our insight into that is relatively minimal, but we can look at the fights and see where we think the breakpoint is.
For those who have been raiding under the current 4-difficulty setup, it is clear that neither LFR nor Normal are the starting point difficulties in design, so where does Blizzard seemingly start? I would contend that a simple glance at most fights over the 5 years of our current difficulty split (and even before when Heroic was the top of the mountain) would suggest that it is the current Heroic difficulty, formerly Normal, that is the primary target of design energy. Why is that? Well, let’s look at a reasonably well-trafficked fight from Legion, Argus the Unmaker in Antorus.
Argus’ biggest mechanical hurdle on most difficulties is positioning to avoid the Cone of Death puddles on the ground and to bait them to minimize the amount of coverage in the room. LFR, Normal, and Heroic all do this easily – you stack as a group and move together for the cast that will leave puddles on the ground. There is also the single target puddle, which is easily managed by having that player backstep to keep everything tightly managed. Mythic adds debuffs that force players to stack on each other or spread out, making the phase 1 mechanics far less manageable. Over time, these mechanics spread to more players, causing a greater amount of positional control to be lost as eventually players must spread out with the remaining players stacked tightly. Coupled with the need to manage the Gift buffs from the Titans during this phase, it becomes quite hectic. I’ll ask from this example to begin with – which of these do you think was made first? One of them has a clear strategic ideal to execute against, one that requires learning (in a bubble where Fatboss isn’t smashing face against the boss on PTR and releasing week 1 videos) and then rewards that learning through a simpler phase, and the other clearly uses that prior understanding against the player – weaponizing it to complicate the clean, simple strategy the other modes instilled.
The relatively short history of Mythic, and the Heroic difficulty that preceded it, is full of stories like this one. Most bosses on the highest difficulty simply add to the Heroic version, taking the prior understanding of the fight and using it to subvert player expectations. Yogg-Saron in Ulduar is designed around using the Watcher buffs, and as you increase the difficulty, those fade away. Kil’Jaeden in Tomb of Sargeras is designed around tight positioning in Heroic, only for Mythic to introduce a literal bullet-hell minigame that makes positioning far more important, and reduces the tightness and compactness with which players can stack. Garrosh on Mythic is a similar fight but makes everything start to overlap and stack, and then uses an added phase that isn’t on any other difficulty in order to push players harder, with brand new mechanics you don’t get to test or learn in any other mode.
I use these examples because I feel like the idea that Blizzard designs for Mythic players is quite provably wrong, at least if we read and interpret the fights as seen on live. Nothing about the Mythic modes screams of having been designed first – they usually feel like iterative or additive changes to the Heroic fights. Blizzard simply adds mechanics for Mythic and subtracts them or dials down the severity for Normal and LFR. As a last example, let’s look to the current raid tier and an interesting case study – Radiance of Azshara. This fight is mechanically near-identical on every difficulty – numbers simply go up or down depending on what word describes the difficulty being played. The only difference is at the top – Mythic gets two adds during the add phase instead of one, which forces the group to split into two halves. This one would actually be an interesting case to make if you wanted to argue that Mythic was designed first – you could reasonably make that case. However, having successfully farmed both of the middle difficulties of this fight, I can say that it doesn’t feel like I’m playing an easier version of the core design when I play Heroic. It feels like I’m playing the core designed difficulty, because nothing really feels out of place.
I use all of those examples because I want to lay that idea to rest before moving to where I think Blizzard has always struggled and for much longer – balancing around player counts. Ever since Wrath of the Lich King, Blizzard has fought with a design problem of their own making – the scaling of raids to player count. Part of Wrath’s big feature list was that all raids were both 10 and 25 player for the first time ever. This was perpetually a problem in Wrath – Blizzard treated 10 player as the easy mode, making it give lower item level loot, different achievements and titles, and leading to the perception in the community of smaller raid sizes as lesser. However, it was demonstrably false – if you put a 25 player guild into the 10 player item level and took their 10 best into the “lower” difficulty, it often proved harder. Blizzard had made it easier by offering overpowering loot from 25 player difficulties and Emblem vendors, and the perspective this created became an accepted “fact” within the WoW community, one that only remained even past Cataclysm when the two sizes were treated as the same difficulty with the same rewards. Frequently, even with adjustments to mechanics for the 10 player version, it was often harder than 25.
The problems were many – the ratios of players were very different between them, as both raid sizes had two tanks and usually 1 healer per 5 players, but DPS would flex up and down to different scales, so you could not simply offer a boss that was 2.5x the health and damage output, especially since your players had similar health pools between the two raid sizes. It also offered some flexibility to 25 player raids that 10 player raids didn’t have – it was feasible to think about taking a smaller number of healers on 25 player, since the damage output could not scale proportionally, where on 10 player, taking 2 healers was a challenge and scaling down to 1 was effectively impossible in nearly all fights. Blizzard’s balancing always erred on the side of caution, generally leaving 10 player raids overtuned and 25 player ones undertuned.
At the end of Mists of Pandaria, Blizzard opted to make Siege of Orgrimmar with one Flexible raid difficulty, in addition to the fixed size LFR, Normal, and Heroic. This revealed a lot of interesting flaws in Blizzard’s design – players could math out ideal scaling points, like that most of Siege Flex was ideal at 14 players, with fixed numbers of healers and DPS inside of that crafting a group ratio. With this one difficulty, you could see the points at which Blizzard expected you to bring more healers and tune to just under that to maximize damage output and minimize mechanical complexity.
The problem with flex raiding has always been this – Blizzard does not really change their assumptions about raid composition and so there are always breakpoints in raid comp that can be exploited. Blizzard has tried to re-tune health and damage numbers to reduce the difficulty droop that existed in many of these cases, but that has created a problem – smaller groups and guilds are hit harder. The breakpoints were a god-send for a social raid with a fixed number of dedicated players and some friends who don’t really want to full-time raid, but are willing to jump in to help when needed. If you have a core group of 12 back in SoO, you can get two fairweather friends who will jump in for flex, hit the breakpoint, and smooth over the raid so everyone has a fun, chill time.
The current design paradigm has attempted to nudge these breakpoints up in difficulty, but the challenge with this is that it always hits a smaller group harder. When my raid was around 12-14 players with 2-4 space cadets who hang out in every void zone like it will dispense candy if they wait long enough, that sucked. Those players were burdens, plain and simple – which sucks, because I usually liked these players as people, but their presence in a raid actively made things worse. In the current situation my raid now finds itself in, we have between 19-24 people a raid night, with around 6 space cadets. Those people can still screw up, and it still sucks, but significantly less so, because their numerical contribution to the raid’s difficulty can be made up by a less exemplary performance.
When we were small, we needed game-changing, championship-level performances from our best players to make up for our friends that we shackled to our ankles and dragged through the raid. At the larger raid size, we have both more players capable of that high level of performance, and each of them needs to put up a smaller amount of over-performance because proportionally, it matters less. If we lose the space cadets at 24 players in raid, it’s fine. If we lose them at 14, even with fewer space cadets, it’s probably a wipe.
Smaller guilds have always faced this amplification of individual-level errors. In BfA, the scaling factors at play have made them the worst they’ve ever been, as there are more individual-level mechanics forcing a higher degree of failure, which would cause a small raid to wipe while a larger raid can often battle rez the affected player and shrug it off.
Compounding this is the waning popularity of the live game, as most publicly-available indicators show that the audience is shrinking overall, minus momentary spikes for new content releases. It will end up being a larger problem in a world with WoW Classic, where, from a business and investor perspective, they will be growing the game, but from a player perspective, I fully expect there to be sharp divides that will push live guilds to thinner numbers. I do expect that the Classic rush will be temporary, but it will play a role and some guilds and raid groups can’t survive a major outflow of players without losing their core and being unable to continue.
Lastly, I’ve wound around discussing it so far but we need to touch on class balance. Class and spec balancing plays a substantial role in the raid game and who wants to play what. At a macro level, there is an imbalance of specs in favor of melee, yet melee are often sharply disadvantaged by fight design in the game, something that has remained the case for years. All classes added to the game since launch with DPS specs have had exclusively melee specs, and with the Survival Hunter changes in Legion, the game lost an additional ranged spec to melee on top of the furthering imbalance with each new class added. While smart-targeting mechanics can specifically select ranged players, this is a bandaid solution to a fundamental design issue that is exacerbated by the class choices available.
One thing that is quite obvious from a long time playing WoW is that the team designing classes and specs, ensuring that they meet balancing targets, is not the same team that is designing and implementing PvE bosses and content. I don’t doubt that the WoW team has some degree of balancing they do to ensure that all specs perform up to a given standard per fight, and it is perfectly alright if the positions on the DPS ranking fluctuate from fight to fight. However, the thing that persists is that there are fundamental gameplay feel issues for some classes on given fights. The thing that I notice about world-first guilds is that they often class-stack and optimize to an extreme degree, and while it is easy to simply dismiss this as these classes being better at given fights, the truth is that many of these classes do far better with the common core design mechanisms Blizzard loves – target switches, cleave and AoE fights, and constant movement. Rogues do well because they have lots of speed boosters and gap-closers, plus with their current combo points mechanic, they can ramp up their damage faster. Mages tend to do well since Shimmer allows a higher degree of casting mobility. Warlocks, especially Affliction, can ramp a lot of damage via DoTs and rapidly adjust positioning to resume casting their fillers.
This creates an issue because these classes and specs are frequently stacked, since they have a lot of general strength, and then end up being rewarding because they have utility that matches to a lot of raid mechanics – Rogues can use Cloak of Shadows and Evasion in cheesy ways to manage mechanics, Mages can Ice Block through a lot of things, Warlocks have their Demonic Circles and movement portals, coupled with a battle resurrection and healthstones. Tank-wise, Monks have been a strong choice for most of Battle for Azeroth, as Stagger allows for some incredible reductions to incoming damage that cannot be matched by most of the other tanks at the same played skill level – a bad monk tank is often still better than a slightly-above average DK or Demon Hunter, as an example. An exceptionally played Monk tank is damn near untouchable. These sorts of design issues are highlighted in raid content – when the common Mythic levers are absurd tank damage and soak mechanics that cause magic damage, it turns out these same toolkits will continue to shine!
This happens in Mythic Keystone dungeons too – Rogues are the highest represented class by far because their AoE stealth is one of the strongest and simplest ways to overcome unnecessary trash pulls. Death Knight tanks were dominating in the 2018 Mythic Dungeon Invitational, far and away the most represented tank, but in 2019, that title now goes to the protection warriors, because of the intense synergy between some of their talents, their AoE damage abilities, and certain Azerite traits, allowing them to do ridiculous AoE damage while remaining sturdy and strong. Similarly, restoration Druids are strong in this season because they are able to roll heal-over-time effects on the party while weaving in damage, allowing them to be strong for the timer-focused gameplay in a way that discipline Priests can’t be, because of the ramp-up required for their group healing and the need for larger amounts of targeted healing, including healing after all enemies are dead.
These issues all compound, and the thing is, I can sort of see why people might think that class design is clearly targeted at the high-end. A lot of times, the ridiculous powers of certain specs are only seen at this level of gameplay, are nerfed accordingly, and this unfortunately causes the players at other levels of the game to suffer as their preferred spec no longer has the same oomph, even if they weren’t taking advantage of the specific abilities causing the imbalances. In truth, though, this is a consequence of the way that Blizzard designs the game – still bad, but it isn’t so much that they set these fun easter eggs up so that really outstanding players can optimize into a massive advantage, but rather, the tail wags the dog – the players at the high end discover the synergy, Blizzard decides if the game’s balance can handle the ability as is, and then it is either left alone and unmentioned or nerfed. These changes often take into account the current class balance or class balance across the tier, but not much further than that. In WoD, for example, the first raid tier saw Brewmaster Monks be very undertuned and their already low spec representation dropped further. To adjust, they received a fair number of buffs in 6.1 and further adjustments in 6.2, which saw them become one of the most powerful tanks. This is when I started main tanking, as my monk, because the power could not be denied – getting Guard shields for 2 million damage on top of old Stagger where Purify fully cleansed the delayed damage, made learning how to tank really easy. It’s hard to be anxious when it will take a mob 10 seconds or more to chew through all the damage reduction you have!
With all of those discussed, what do I think Blizzard needs to do for the future of the game’s PvE content?
I think there needs to be a strong, central designer or leader that works on better mating class design to encounter design. I won’t pretend to know that Blizzard is or isn’t doing this, but from what they share publicly about how the team is divided, I would wager they do not. I would like to see a better balancing of melee and ranged DPS roles in raid encounters – movement mechanics should not be excessively punishing on melee, but they can be better-designed to allow strong melee players to stand out. Most importantly, if flex raiding is to remain in the game (and it should!), it needs to be far better balanced at the low end. Losing a player should not feel like a death sentence unless you are still progressing and need every single point of damage and healing you can muster.
All of these would represent the start of something better balanced and more interesting.