The Great Thing About Helldivers 2's Design is That It Exists
On vision and commitment to it
It’s rare for AAA games to have noteworthy high-level design.
They have low and mid-level design. Level designers lay out spaces, combat designers come up with new systems and movesets. Someone designs enemies and bosses and guns and swords. Someone makes a skill tree. But at a high level many AAA games play like other AAA games, mixing the same design elements in slightly different ways.
I’ll be comparing Helldivers 2 to Suicide Squad: Kill the Justice League a few times throughout this. Suicide Squad is a good example of a game I’d say has no real high-level design. Combat has some wrinkles like shield harvesting (a clunky version of Doom’s Glory Kills) and melee attacks that transition into gun juggles - someone had to design those, sure. Someone worked hard on elaborate areas like the Batman Museum and the individual traversal mechanics. But the high level vision for the game is “it’s an open-world mission-based GAAS looter-shooter - get off my back.”
Suicide Squad had 4 different director positions. (Maybe that was part of the problem!) I’m sure the directors had plenty to do in terms of managing production, but I suspect if you polled a group of open world looter-shooter players and averaged the results you’d get the design of Suicide Squad, more or less. It’s “travel to a location and defend some points for 30 seconds” missions with superhero themes on top. The toughest high-level design challenge it poses is how do you reconcile a looter-shooter with superheroes? That’s a nut they didn’t even try to crack - it’s really not a superhero game at all, it just uses superhero skins and names. In that sense it’s well behind the much-maligned Avengers.
I don’t think you could poll a group of players and end up with the design of Helldivers 2. It makes particular choices rather than a series of default choices - which isn’t that common in AAA or even AA games.
Someone on Batman: Arkham Asylum invented Detective Vision. As a AAA designer working today your job isn’t to invent your own unique element similar in spirit to Detective Vision that will become an industry-standard feature moving forward; instead your job is to copy Detective Vision, give it a new name, and invent some in-fiction justification for why it appears in your game. Lara Croft has strong instincts so she has “instinct vision.” In the upcoming Wolverine game Wolverine, with his strong animalistic senses, has Smell Vision. (Probably not the name they’ll go with) In the comics Wolverine uses his senses to investigate, sure. But let’s not kid ourselves - if Wolverine didn’t have super-senses they’d come up with another rationalization for including a vision mode. Spider-Man also has Detective Vision in his games, because reasons.
That’s how AAA design has trended for some time: borrow “tried and true”, “industry standard” ideas and combine them in not-particularly-interesting ways. And often water down the borrowed ideas. Ubisoft’s take on Breath of the Wild, Immortals Fenyx Rising, is the game dialed back from 10 to 7. Wherever Breath of the Wild made an interesting choice Immortals replaced it with a more conservative one.
The big choices in AAA games tend to be in narrative, technology, art and budget, not in design. I suspect that’s one reason big publishers are slow to jump on trends like Minecraft and survival-style games - those are genres that trade in game design, not face-scanned models and motion capture.
I don’t demand innovation - it’s fine that some games are comfort food, just more of the same but executed well. But it’s refreshing to play a game with genuinely new ideas, or at least new combinations of old ones. It’s nice to play a game where people with job titles like “creative lead” did something creative design-wise.
Which brings us to Helldivers 2.
The Design of Helldivers 2
The other day I read an interview about Dragon’s Dogma 2 and the design of fast-travel options. It struck me how rare it is to hear game creators talk about game design specifics. Not in a hand-wavy way like “we want the player to feel powerful emotions and a sense of purpose”, and not about low-level mechanics like “we want players to move constantly in combat which is why enemies throw grenades.” But more “this is the specific design choice we’re making to support our overall philosophy.”
Helldivers 2 is a game brimming with design choices that support the overall thrust of the game. I keep returning to one thought: if this were a game with a higher budget and higher sales expectations it would look very different; the interesting decisions and rough edges would be sanded off.
Friendly Fire
Helldivers 2 has friendly fire damage and that damage is high. It adds unpredictability but also frustration and “unfairness”, two things AAA games are keen to avoid. A mid-budget game can get away with “if you don’t like the friendly fire play another game, we find it funny”, but games with higher budgets eschew that approach.
There are “rational” arguments for toning down friendly fire. I’m sure someone could engineer a playtest session that “proved” that friendly fire is a bad idea, or have a “player experience engineer” well versed in “player psychology” make a presentation about how friendly fire “interrupts flow states”, “penalizes players for factors outside their control” and introduces “friction-inducing stressors.” And the higher the budget the more likely this is to happen, in the same way that increasing budget increases the likelihood of a side character saying “watch out! Those bugs spit acid!”
Stratagems
Stratagems are special abilities in your loadout - to use one (on controller by default) you hold left bumper and enter a code via the directional pad.
To enter the code you either have to take your thumb off the left analog stick, rendering you unable to move, or you have to use your right hand to enter the dpad code (awkward!), or do something even more exotic like take over the left stick with your right thumb.
This is non-ergonomic and “inaccessible” by design. It’s error prone, and if you enter the code wrong you have to start over. Sometimes in the heat of battle you enter the last bit of the code wrong, don’t realize it, and throw a grenade at your feet thinking it’s a stratagem.
Some designers believe that control difficulty is an invalid avenue to explore - that any control mechanism that takes effort isn’t part of the game but rather gets in the way of the game. They’d resent that the complexity of the codes is a balancing mechanism; stratagems that are either more powerful or intended for use outside of combat have more complex codes.
Then there’s the crowd that believes that controls have been solved per-genre and should be “standardized”, so Helldivers 2 should just do whatever Call of Duty does. And if that’s not possible because the game has different features those features should be changed to be compatible with an “industry standard” control scheme.
There’s little chance this system would survive user testing at a AAA studio. Most testers would frequently mess up the stratagem inputs, which would be flagged as a problem, and when the game shipped you’d hold a button to open a radial menu then select one of 8 possible stratagems from a wheel.
General Obtuseness and Reliance on Diegetic Elements
I died multiple times in the tutorial when I was supposed to crawl prone past turrets. One time I went prone too late, another time I dove then immediately stood back up. That’s not a problem but it is an early indication of a design tenant throughout: you’re allowed to fail.
The game is unfriendly, or at least seemingly ambivalent.
As I wrote in On Unfriendly Games
The designers of friendly games are helicopter parents. They show their love and affection by constantly intervening on your behalf. They are benevolent gods, and while they aren’t standing over your shoulder they help you through their angels of algorithms and floating arrows.
The designer of an unfriendly game is a Deist god - they set the world in motion then go drink a margarita. They show their love and affection by saying “I have faith that you can do it on your own little buddy.”
Helldivers 2 expects you to muddle through things. How useful equipment and stratagems are, or even what they do, is determined mostly through experimentation. Map objectives are often confusing on first encounter - the terminal tells you to “adjust radar dish” but it’s not clear whether to turn if left or right or by how much, or even where the radar dish is exactly. Some objectives have you turn multiple valves - one player will turn a valve to the on position, then another player will come by and mistakenly turn it back to off.
The other day I played a mission where 2 members of our team died a bunch then quit in the first couple minutes. At one point I also died and spent time observing my remaining teammate - he crawled on his stomach and crouch walked around, killing robot enemies one at a time with a pistol. The game explains nothing about aggro mechanics, lines of sight, noise, whether you’re less visible prone or crouched, etc. You could play the game for dozens of hours and never know that stealth is even an option.
Enemies have weak points but they aren’t glowing and obvious (well, ok the robot weak points do glow somewhat). How weak those weak points are exists on a spectrum - a thick frontal shield takes no damage, legs take some damage, back takes a lot of damage. The interaction of weapons and armor is left unexplained - again you just have to try things out in real-world conditions and see for yourself.
To destroy a robot factory you can throw a grenade through a vent, but you can also throw a grenade through the door as a robot walks out. That’s just something you might try or luck into. Some non-explosive weapons can destroy robot factories if you bank a shot into a vent by bouncing it off the metal covering - again something you might not realize even after dozens of hours.
I often say that in unfriendly games the hand of the designer is less evident, but what I mean is that the guiding hand is less evident. There’s no shortage of evident design in Helldivers 2 but there’s little sense the designers are helping you along. There’s not much indication that they set fresh play-testers down in front of the game then freaked out when they struggled.
Suicide Squad: Kill the Justice League constantly tells you what to do. Here’s an example of the Green Lantern fight. The top-right says “destroy construct defenses” - during this part of the fight Green Lantern summons “adds” that you have to destroy. A similar message sits in your face at the center of the screen: “destroy Lantern’s constructs to remove his shield.” The game designer is telling you exactly what to do. Green Lantern also has dialogue hinting at what to do - that’s three different pieces of onscreen text telling you how to beat the encounter. Character barks in this fight also function as hints - members of the Squad shout out (roughly paraphrased) “his shields are down, blast him” or “destroy his summons!”
This is a game deathly afraid of players not getting it. Many AAA games are designed with fear rather than confidence - what if a player gets stuck and then they hate the game and then they return it or don’t buy the sequel?
In Helldivers 2 you can defeat a tank by throwing a few grenades on top of it but there’s no immediate way to know that - maybe the top of the tank looks vulnerable but it’s pretty hard to tell through the constant stream of gunfire melting your face off. Meanwhile Suicide Squad gives you 3 or 4 layers of redundant help at once - text on the right that says the same thing as text in the middle that says the same thing as a character bark.
Presumably in Suicide Squad players weren’t picking up on the natural and diegetic elements - it wasn’t clear when or why Green Lantern’s shield went down, which made the fight dull and frustrating. So they added more onscreen text and barks. But then the text covered the shield graphics and the barks masked the sound of the shield taking damage and going down, so the game became less readable, so they added even more in-your-face hints to compensate.
The result is that despite being a faced-paced shooter the game often revolves around following on-screen text prompts like in an MMO. And there’s so much text and bark hints that they amount to noise.
Commitment to the Bit
Helldivers 2 is a game with vision. But vision is nothing without commitment to it.
AAA games have a strong pressure to revert to the norm. Later Dead Rising games eased off on the real-time elements. Similarly Pikmin games eased off on time pressure with future installments, while Resident Evil moved away from limited ink ribbons for saving.
A few weeks ago a story circulated about how 343 Industries had pitched a Halo game similar to Helldivers. But even if the pitch really was like Helldivers it’s hard to imagine the finished product being similar.
Helldivers 2 commits to the premise with the mechanics highlighted earlier, but also in more subtle ways. When your character dies they don’t respawn, they get replaced by a different character. If all players die and fail to extract missions still count as completed as long as the objectives were done. The game has minimal instruction, and the in-game tutorial (which is canon for your character) doesn’t prepare you at all for what’s to come. All of these tie into the same theme of disposable soldiers.
Loading screen hints, in-game orders, informercials that play on your ship’s tv screens - all of them represent commitment to the bit.
An example of strong vision with weak commitment is Watch Dogs: Legion. On paper the premise is intriguing and promises all sorts of possibilities. Maybe your entire squad dies except for one elderly grandma, a humorous near-loss condition. In theory two players could have radically different squads based on who they recruit.
In practice I suspect most players identify a few core team members early who cover the bases then stick with them the rest of the game. While you can play as anyone much of that time your character is controlling a Spider Bot or other gadgets, so who you’re playing is immaterial. The elderly characters aren’t particularly different from other characters - they animate differently but still ride around on motorcycles and capably engage in firefights.
There’s a permadeath mode that forces you into new operatives, but it seems like an afterthought, with players frequently complaining that they lose characters to glitches or control issues.
The pitch is much more interesting than the execution.
Perhaps the all-time example of a strong premise the developers shied away from is Spore.
This is how early Spore is described:
In the extremely early versions that I toyed around with, I was able to make creatures that shifted under their own weight,” explained mflux. “Creatures that exploited the length of their arms or legs for greater reach. Creatures that behave and move true to how they were built. A short bunny-creature would definitely be out-run by the long-legged dragon-giraffe. That was very neat, and it implied several exciting possibilities in gameplay…The strategy that earlier prototypes implied went beyond placement of parts. The length of limbs or spine felt like it mattered. If you had a forward-heavy animal with legs placed in the back, it would run poorly as it tries (and fails) to counteract its own weight.”
And this is how it ended up:
Of course, the editor enables an infinite number of visually distinct creatures, but the gameplay effects of the creature parts are unfortunately quite discrete. The feet components each carry with them a canned set of attributes – for example, Stubbtoe gives “Sprint 2,” “Dance 1,” and “Speed 2” – regardless of the position of the foot, the length of the attached limb, or the shape of the body. Thus, the attributes of each creature is simply a summation of all the named body parts, and although the procedural animation guarantees that a many-limbed creature will walk convincingly, the player’s creativity in designing the creature’s shape has no impact on actual gameplay.
I would go further: calling it “procedural animation” is a bit of a misnomer. The animation in Spore is authored, not procedurally generated, such that creatures animate similarly regardless of limb number, length or shape.
The combat and other interactions are mostly stats-based - fighting an enemy is like MMO combat.
A group of devs working on Spore were concerned that players could create creatures that weren’t effective - that would have difficulty manipulating objects or feeding themselves or navigating terrain. Rather than let the chips fall where they may, shrug and say “that’s nature” (or: “don’t design your creature that way then”), the solution was to make that sort of failure impossible by reducing creatures to their core stats.
Meanwhile in Helldivers 2 sometimes you try to use a terminal and there’s so much glare from an overhead lamp that you can’t read the screen. That’s a fundamentally different, and refreshing, design philosophy.