Continuing our GDC 2022 coverage of Bungie Principal Technical Game Designer Alan Blaine’s talk on building challenges and reusable content in Destiny 2, we’re going to take the stressors he defined, modified for MMOs, and consider how developers can use them for tuning content and making repeatable content.
While most of our readers are not devs, having this vocabulary and development insight can better help us understand why some MMO content may feel like it does or does not work – and why we may see similar development tactics across the genre.
Balancing stressors
It’s important to note that while all seven stressors (time is power, thumbskill, battlefield awareness, communication, buildcraft, planning, and social, all of which we outlined in detail in our earlier coverage) can have a kind of hard gate, Blaine said he doesn’t personally think most endgame content should be hard gated. A few are OK, but there should be some kind of tradeoff for players. For example, last time we mentioned how one player may prefer to tackle content that’s high on thumbskill/fingerskill while another player prefers an encounter that requires a lot of buildcrafting. How can you make an encounter appeal to both players of, say, a traditional fantasy-based action-combat MMO?
Well, theoretically, the fingerskill player wants to show off those twitch skills, which is why we’re looking action-combat rather than the traditional tab-targeting, as that makes it easier for us to find a middle ground with the buildcraft player, who often wants to show off an ability to have just the right skill set/loadout for the job – such as in classic Guild Wars. So, a dev would use a little battlefield awareness to help tune a theoretical encounter; there’s no reason not to be flexible when we have seven stressors to play with.
So for this theoretical encounter, the boss announces when it will use its ultimate attack, summoning high-damaging ghosts that will instantly kill a player. However, the boss also gives a random assortment of debuffs corresponding to various ability categories: damage-over-time, crowd control, heal, and dispel. For the sake of simplicity here, we’ll assume that all of these abilities are available to all players. The players’ debuffs must be removed in the correct order, and doing so will break the boss’ concentration and thus cause the boss to “miss” the player, partially represented by an invulnerability buff gained upon completing the task. The amount of time the player has to use these moves should be generous because we’re encouraging battle prep, not reaction time. Thus, the buildcrafter has to figure out which moves to bring to the battle, not only in terms of DPS but in terms of the categories needed to survive the boss’ ultimate.
Now, one might be tempted to just insert the ghosts as telegraphed hazards for the fingerskill folks, but that leans too much on battlefield awareness. Instead, during the debuff countdown, ghosts will be mobs approaching the players that explode across the entire room at the countdown’s end. They should be numerous enough that they can’t all be taken out, but not so numerous that they cause massive performance issues. Attacks from the front will be negated, but from any other angle will be fine. Each defeated ghost builds an absorption shield and minor attack buff.
Through this, the fingerskill player needs to jump into a dangerous situation and weave through the crowd not only to survive the ultimate attack (it should be extremely difficult for 95% of the playerbase to take no damage from the attack) but to gain the highest attack buff possible. The buildcrafter can potentially take advantage of this as well but will be at a slight disadvantage compared to the all-in fingerskill player. That is, the buildcrafter focuses first on achieving survival via the toolkit, the figuring out a skill set that allows them to take advantage of a fingerskill challenge to get the best of both worlds. The fingerskill player, on the other hand, can focus on reaction time and battlefield awareness in a way to challenge reflexes and show off some big numbers without having to change up a build too much.
Hard numbers, of course, would be tuned through playtesting. But the overall idea is that by understanding the stressors, their strengths, their pitfalls, and how you’re utilizing them, a developer can tune an encounter to appeal to a different or even broader audience, often using tools they already may have. Yes, fingerskill’s twitchiness may require tech often used to make games more accessible, but that should actually be a good thing. Accessibility shouldn’t be a dirty word; it means innovation for everyone.
Ultimately, devs shouldn’t feel locked into designing the same traditional content that every online game does, but they should be aware of what they’re doing and see how it goes over with playtesters. Blaine argues that developers shouldn’t focus on whether something is too easy or too hard as live players will almost always figure something out; instead, they should listen to how the testers tackled the stressors. In our example, did they figure out they can bypass the fingerskill stressors with proper buildcrafting, or that they could merge the two? Did they perhaps find an unintended strategy, like maybe a rock they can hide behind to avoid the damage entirely?
Remember, as much as Bungie itself may make use of fingerskill, Blaine noted that developers should understand that that particular stressor has limitations, so devs not only may make something they themselves can’t beat but also shouldn’t be afraid of doing that either. He says that it’s better for developers to err on making things too hard, as it’s easier to make things easier than harder once it’s out, and players are often happier to hear about difficulty being toned down than up, except of course if you add harder modes with new rewards. And that’s where reusable content comes in.
Making the old new
The specifics of how Destiny 2 retrofits old content for something more reusable may be interesting to D2 players, but we’re going to talk in more general terms to broaden our understanding of not only what Bungie does within the MMO-space but what other developers could do to mimic it. But we’re essentially talking loot hunting and grinding, so if you want unique stories and scenarios, parts of this may work, but other parts, maybe not so much.
First, Bungie picks old content to tweak. Obviously this helps save money on art assets, programming, voice-over, and lots of the usual costs, but the work isn’t done. That’s just a starting point, and you won’t be able to just pump out the content and leave it as is. Blaine notes that content needs to have nearly infinite desirable rewards, akin to Diablo 2’s gems – something random but potentially useful that can motivate players to keep grinding content.
But the content needs to be changed too. Yes, you’re going to have to make more, but to help keep things fresh, you need to rotate that content. Leaving it out for too long makes it easier for players to master and figure out where to get the most bang for their buck, and that’s kind of a problem. Players love to optimize, and if you make 20 different dungeons and players figure out one gives more rewards for less effort, you may have 19 dungeons no one really plays anymore. By making that content less available, people have to contend with what they have and make the best of that.
From there, Bungie developers a kind of ladder approach similar to Super Mario 3D Land director Koichi Hayashida’s use of kishōtenketsu in game development or a teacher’s scaffolding approach to lesson execution. The difference, however, is that while the former two are a bit more about learning mastery, Bungie’s feels a bit more like juggling and goes through five difficulties, rather than four (thought I think readers will understand the concept after four).
Let’s say you have a lowbie dungeon that’s popular, like World of Warcraft’s Deadmines. What Bungie would do is take that basic dungeon and create several new versions of it for high-level players that they can complete, say, once a week.
Version 1 would simply increase the time = power stressor so it’s more appropriate to high levels but not necessarily level-capped players. Obviously things like battlefield awareness are represented in some way but are taking a back seat. Perhaps the spawns are thinned out and the bosses are soloable with an option to team up via a looking-for-group tool.
Version 2 would be for level-capped players, but we’re going to have the mobs deal elemental damage and have elemental weaknesses. Players will see this warning before entering Version 2 so they know that Buildcrafting is an additional stressor. The mobs may also be tuned for multiple people, but if we had an LFG tool, we can focus on just basic communication and largely skip hard social requirements. As in SM3DL, we’re building skills and things are progressively getting harder.
Version 3 is where the juggling comes in. Bungie does something like taking version 2, making everything a bit harder, but now you no longer have your LFG tool and your equipment is locked (which would also be like locking your talents/class/hotbar selections in certain MMOs). There’s also a fingerskill component, such as defeating the boss faster gives better rewards. You can call it a twist if you’d like, but it doesn’t stop here.
Version 4 is going to take everything in Version 3 and turn that up more, but also we’re going to juggle some more balls with a communication stressor. Perhaps a boss has an interruptible ultimate ability, but no one class will have cooldowns low enough to stop it each time. You can think of that also as a fingerskill and battlefield awareness stressor, but oh yeah, now there’s also an enrage component and group wipes reset the dungeon.
And if you don’t quite see where Bungie’s going with this yet, Version 5 is going to do all of that, but now maybe the group as a whole gets only one death per player and rezes work only within a limited time frame, but you at least can earn lives by beating bosses within a certain amount of time. Yes, Bungie does lean a bit on those fingerskill stressors.
While it may sound like a lot, Blaine’s team often started with needing a few weeks of development time, but now it’s less than a day’s worth of time, with smaller teams and (most importantly to the guys upstairs) less money spent.
For some MMORPG readers, the tactics may also seem familiar. Didn’t WoW do this by adding looking-for-raid as a stepping stone into the raid scene, or the original Dead Mines in Cataclysm? Yup, pretty much, but I feel Blaine has given us the vocabulary and structure to help us look at these examples, among others, and better appreciate them at the very least.
Looking at design this way may seem a bit cold, and people who prefer new content to repeatable content obviously will have limited interest in using this to talk about grinding dungeons, but my hope is that by using Blaine’s vocabulary and execution methods, we’ll have an easier time both discussing and assessing the issue of content design in MMOs, not just in terms of grind content but in how such stressors and retuning of content might or might not be applicable to harder-to-dupe content like branching questlines or GM events.
