A 2014 Manifesto: Does The Stealth Genre Even Exist?

Sneaky Bastards takes no prisoners as it presents its 2014 manifesto by asking whether or not the stealth genre as a standalone concept can even be identified.

[one_third]Sneaky Bastards takes no prisoners as it presents its 2014 manifesto by asking whether or not the stealth genre as a standalone concept can even be identified.[/one_third]

Why do we love stealth games? It’s the power trip. The subversion. Outsmarting and outthinking the threat. The taboo of voyeurism. It’s knowing that we are in control of the situation, even when we are right under their nose and they don’t even know it. The stealth genre tempts us with such things through high level of interactivity, immersive environments, and demanding situational awareness. Just as a stealth game protagonist would subvert security measures in-game, players would subvert the typical action approach. For the best stealth run is one in which nobody even knew you existed.

This is the basis of the stealth genre.

Pure Stealth?

It follows that a “pure” stealth game would be about this kind of stealth run: remaining unseen, leaving no trace, being no more than a ghost.

Our Editor, Dan Hindes, has his own definition:

“Pure stealth for me is when there is no way to permanently eliminate (kill or knock-out) the threat (the reason for being stealthy) – in a way that allows non-stealthy behaviour to be adopted”

Rob Storm, developer of indie stealth-shooter Intruder, takes a further step back:

“Stealth, as an act, is merely having more information than your opponents – simultaneously gathering information about your opponents while avoiding them gaining information about you. To do any action in a way where someone/your obstacles do not know that you are doing it, or the extent to which you are. It’s a war over information.”

The Good Stuff

So let’s go right back to basics. Here are some key words, the implications of which often result in great stealth games.
[list type=”circle”]

  • Camouflage
  • Hiding
  • Watching
  • Darkness
  • Deception
  • Distraction
  • Voyeurism
  • Subversion
  • Planning
  • Execution
  • Reaction

Storm brings up more specific scenarios: AI that behaves more like humans would; a high degree of interactivity in the environment; being tricked or tricking the AI; making plans and carrying them out. Great stealth games are also lauded for offering multiple paths to the same goal.



How does the good stuff manifest, mechanically? See, there is another thing that almost every single stealth game in existence carries with it: action mechanics. It’s a vague descriptor, but think of it this way – the granddaddies of the stealth genre were action games with stealth mechanics. Subsequently, nearly every stealth game since has been as well.

Which brings us to the following point:

There are almost no “pure” stealth games, but there are many many action games with stealth mechanics. Thus, the stealth genre does not exist.

We see a significant difference between mechanics and genre. Even though the mechanics often determine the genre, what genres exist to do is group things by commonality. But this is not a discussion of the semantics of genre as a word. This is a discussion of stealth games and their habitual creation as an off-shoot of action games. Remember – action is okay! This is not a tirade against violence. It is about unexplored territory for stealth as a standalone genre.

Stealth games were born in action, from which they grow – and still remain.

These games usually tell us that the action is optional and a stealth approach is more challenging and more rewarding. However, it’s almost always more optimal to remove an enemy from watching you than it is to sneak around them. Objectively – mechanically – it is easier and far less trouble. This interaction is usually irreversible; a kill or takedown. It is an action mechanic fallback. The choices presented are narrow, tired. They are constantly the same old.

That isn’t to suggest that “removal” as a mechanic has not been done in more interesting ways, such as traps, obscuring vision, distractions, and deception. It’s that all these juicy, interesting and more interactive choices of stealth are undercut by the very existence of a more definite and final act of violence. Where is the exploration of the muddier, analogue stealth mechanics? These options are sometimes there, sometimes better, but still often shadowed by the inclusion of action mechanic fallbacks.

Let’s think about the stealth-action hybrid’s target audience. There are often two main camps:
[list type=”circle”]

  • Damn, I wish this stealth game had more stealth and less action!
  • Damn, this stealth section is the WORST part of this action game!

Few games strike a delicate and effective balance between the two conflicting audience types, yet these are the ones applauded as the best “pure” stealth games ever. This is a indicative of the inherent co-existence with action mechanics, and it bothers us. These games are often incredibly well-designed – but what does it say about stealth as a genre?

Is the best that designers of the stealth genre can do are these action games that do stealth pretty well too?


So what is the core problem with this? The action mechanics are included when they don’t have to be. Knowing that this fallback is available, but also the most final option, gnaws at the fringes of the “pure” stealth fantasy.

There are many action games which have been polished, tweaked, balanced and perfected over years as a genre. The build upon decades of good and bad mechanics and presentations to bring the core audience something amazing. Where are these polished, focused and targeted games for the stealth genre?

If you take another genre, and ask a designer, “I have a superflous mechanic that can (and often does) rail against the core gameplay vision and essence of base mechanics. Yay or nay?”

The chances are their response will be simple – to cut it.

Action Within Stealth

We are not demanding that stealth games exhibit no action gameplay whatsoever. Action moments when played within the wider context of stealth, likely as a part of failure design (fleeing from pursuers or switching to reactive play to escape a situation) still play to the core tenets of immersion, tension and systemic interplay.

What we need is variety. Pure stealth as a genre is largely unexplored.

Let’s take a quick look at what we could consider to be pure stealth games.
[list type=”circle”]

  • Uplink?
  • The Novelist?
  • Stealth Bastard Deluxe?
  • Vyde
  • Master Spy?
  • Not The Robots?
  •  …??

Okay, we’re a little limited. If you have any further examples, leave a comment below and we’ll add them to the list. But it’s hard not to feel like the stealth genre doesn’t actually exist outside of a subset of action.

Design Considerations

Now, any stealth designer knows that there are challenges with stealth – such as a lack of direct and indirect feedback when your point is to remain invisible to the simulation. Both Doug Church and Warren Spector spoke at length about this and how they made their games great with these challenges at hand. Rock, Paper, Shotgun’s excellent series called The Stealth Letters features interviews with other stealth designers covering similar ground.

Great designers solve these challenges in meaningful and interesting ways, by inventing and (hey!) designing solutions to problems.Their solutions are often innovative and imaginative and very effective in the context of stealth games. But they were not approaching pure stealth games, either; they were designing for action games with stealth elements.

We must reiterate – they are not bad designers. Stealth-action games are not bad games. No, we are here to ask questions that the stealth genre has yet to answer:
[list type=”circle”]

  • Why are there no mechanical fallbacks outside of action?
  • Why is the fail state for “pure stealth” always fighting, or this action fallback?
  • Where are the pure stealth games that attempt to address this and fail – ones we can learn from, polish, and build upon?

We want to know what gameplay options this deep and interesting genre COULD hold. In a genre that values high-level choice and interactivity, the choices for the player when not succeeding at the core mechanics are borrowed from the parent genre.

More “Realistic”

There seems to be a devolution here as soon as it comes to the player vs world when the player fails to be stealthy. Suddenly, as the player, we are forced into situations well outside of the zone of stealth gameplay like being a deadly ninja or shooting a ranged weapon to remove the oppressive force (an enemy, a camera, lights). Conversely, we’re put into a more cowardly position where we are being forced to wait and hope the AI doesn’t catch us. Hiding is part of the thrill of stealth, sure – but are there more mechanically interesting options?

The stealth protagonist is usually contextualised in a way that makes these mechanics relatively realistic, but not necessarily in line with the core essence of a pure stealth experience. What choices outside of “bad-arse action man” are there to play as? Out of how many games with stealth, how many options do we really have? Uplink gives us the “cool cyber hacker,” and… Hmm.

Simulation vs You

The reason these stealth hybrids outshine straight-up action games is the high degree of simulation. Enemies hear and respond to you, they communicate with each other. We have seen unprecedented strides in the complexity of core stealth mechanics. But “simulation” doesn’t mean infinite possibility or hyper-realism. There can be hand-waving and approximations and simplistic modellings and it can still be a whole lot better than the existing limited choices.

Intruder, Storm’s stealth-shooter, offers incredible detail in its geography and array of gadgets to make the simulation aspect of sneaking into a building very deep. But it is still an action game at heart.


Aim Higher

Imagine you were tasked with sneaking into an office building, and stealing a file from a computer with a flash drive. Let’s set this in a futurism fantasy, like Deus Ex: Human Revolution.

You’d have multiple paths, windows, vents, passages, cameras to hack so you can watch out for people, all the usual super fun stealthy goodness. Gadgets galore, all the options! We’re fired up, there is the threat of being caught and it’s all a huge thrill.

Consider: Is the average office worker most likely to catch you carrying guns/tasers/crossbows? No.

Now, let’s remove the action fallback crutch for a moment, for the sake of exploration.

Someone walks in and catches us. (Our bad, we should have locked the door using the door lock simulation!) What are the possible outcomes?
[list type=”circle”]

  • Deceive your way out of it
  • Run past them or shove them out the way
  • Hope you have some kind of disguise
  • Walk out silently
  • Escape and hide
  • Negotiate
  • Distract

This is being conservative, with regards to existing (or even really old) capabilities of simulations. But it illustrates our point. In stealth games, we aren’t often presented with verbs from the heart of the genre.

Now imagine they call security on the phone (You have no action mechanics to just remove this person as a threat!) So what are your choices?

Wait for security and comply? Where do they take you? What if they take you where you need to go, where you can eavesdrop? What about using their capture as a way to subvert much stronger more dangerous security to get inside? What if the security arrived and called the police – giving you a new threat to deal with.

Now, what if your protagonist was more athletic, but not to the point of snapping necks. Do you:
[list type=”circle”]

  • Grapple and handcuff the security guard when he tries to cuff you?
  • Deploy a gadget and reposition?
  • Toss something at the cop so he drops the handcuffs?
  • …Dive out the window?

All of your options should fall back on what you know about the world, your awareness of your situation, understanding of the simulation and the possibility space mechanically. And, if it’s a great design, all of these things should cohesively gel with the heart of the pure stealth genre.

By offering fallback mechanics outside the realm of action, this situation presents itself as one infinitely deeper and more engaging. It’s a chance for a more interesting simulation, where the core mechanics, essence and vision of the stealth genre can be expressed by the player.

Sidenote : Most spy fiction inevitably sees the infiltrator captured but ultimately to their own benefit. Even Splinter Cell: Chaos Theory riffed on the trope.

But It’s Hard

Yep! These are design challenges. There are technical challenges. Input mechanic challenges. Feedback challenges. Messaging challenges. But if you’re a stealth game designer, this is your job.

Nobody said innovating was easy. We hope that we can aim higher, fight harder and create more interesting stealth games that don’t rely so heavily on, or even include, action mechanics – because it’s easier, more convenient, more accepted, opens a game up to a wider audience, or simply because it’s what everyone has done before.

We believe that if you wanted pure stealth games to flourish by virtue of the things that make them so fantastic, the actual target audience are the pure stealth fans. Maybe then, we can at least say we have stealth games with action mechanics, instead of the other way around.

Watch_dogs promise


Action games with stealth mechanics aren’t cool? No.

Stealth-action has less value? No.

No more stealth-action games? No.

Explore more than JUST action? Yes!

Explore the possibility space of pure stealth games? Yes!

Explore more in terms of stealth as a genre, not as a sub-genre of action? Hell yes.

Attempt to solve the design issues like possible lack of feedback, what happens in these new fail states, communicating the threat and challenges associated with these new mechanics – without crutches or fallbacks from the action genre, and things that can possibly devalue the core target audience’s experience?

Please. Yes.

Let’s make some stealth games.

This is a cross-post from Underscore Discovery.

9 thoughts on “A 2014 Manifesto: Does The Stealth Genre Even Exist?

  1. I think one of the most \”pure\” stealth games was Velvet Assassin. Very very limited ammo. However, the execution of that game was bad: lack of multiple paths, lack of information to give you the \”power trip\” and so on.

  2. Interesting manifesto – straight to the point and asking the right question.

    I\’m getting the itch to start designing following this \”Dogma 95\” of the stealth genre… Who knows, maybe I\’ll have the time to do a prototype !

  3. Though I agree that pure-stealth is rare I can\’t accept that it\’s because the stealth genre exists primarily as \”action games with stealth mechanics\”. Such a description is historically inaccurate, serves to simplify action games to \”press button to remove threat\”, and is misrepresentative about the nature of the stealth genre. Stealth games have action or other non-stealth elements because, like pure-horror, pure-stealth operates within very narrow simulation boundaries.

    \”Remove the threat\” mechanics are not included because stealth games are really action games underneath but because detection is a simulation boundary for the stealth genre. The mechanics of pure-stealth are those that only really function when you are able to remain undetected, once that state changes those systems begin to break down; hiding in a locker doesn\’t make much sense if an NPC saw you go in there.

    There are a number of ways to deal with such a hard simulation boundary (Clint Hocking discussed these in his GDC presentation \”The Interesting Thing About Bishops\”), you can strictly enforce this boundary (1), redirect the player back into the valid simulation space (2), or attempt to simulate beyond this boundary (3).

    1) Leads to a situation where detection results in instant failure, the boundary has been reached and the simulation ends. This is a method often used in action games that included stealth levels, Assassin\’s Creed being a prime recent example. Though probably the most honest way to handle simulation boundaries it is also the most likely to cause frustration if those boundaries are not clear.

    2) Is the method used by a lot of games, instead of enforcing a strict boundary they provide the player with a means to recovery and reset the simulation. This can be as simple as waiting until a timer counts down, as in Metal Gear Solid.

    3) Creates complications as all you are really doing by simulation beyond the boundary is pushing the boundary further out and running the risk of combinatorial explosion as the size of your simulation grows. If players are detected then you simulate beyond that to include evasion and combat mechanics, and then if those fail you either need to resort to option 1 and end the simulation, or option 2 and allow a soft reset or attempt to simulate even further.

    For a pure-stealth game to exist it would have to either avoid options 2 and 3 almost entirely because the moment the \”detection\” boundary is crossed the game is no longer a pure-stealth game but some form of hybrid.

    The suggestions offered above fall largely into the category of option 3, they are means of simulating beyond the detection boundary, eventually another simulation boundary will be reached and it will be necessary to simulate even further or fall back on options 1 or 2. Meaning as useful as option 3 can be all we do by adopting it is shift the goal posts.

    The choice we seem to be left with is to either accept that pure-stealth means we have to strictly enforce the detection boundary with instant failure conditions, or that hybridisation is not a failure and that simulating beyond that boundary is acceptable provided the means by which we do so operate on the same basic principles of the stealth genre itself (deception, distraction, subversion), even if they may not be traditionally considered as such. Examples of the latter could include a conversation systems where we can use deception and misdirection to talk our way out of being captured.

  4. Good read! Since you asked, I would consider Republique and Rain just this past year to be rare examples of the elusive \”pure stealth\” game too.

  5. An alternative version of Justin\’s option 2) is allowing the player to backtrack – fluently Braid style or by discrete checkpoints/(limited) saves – to a point of time/progrrssion safely within the boundary. While giving technical challenges as well as mechanical, it would avert the combinatorial explosion and inevitable feature creep of option 3).

  6. For me, Stealth=No detection. Now that could range from Silently killing enemies, to knocking them out silently, to leaving them untouched, or even spooking them to the point where they know something\’s wrong yet they don\’t see you physically (a la Batman Arkham Predator gameplay).

    The moment you are detected, it breaks all stealth. However I\’m still debating over whether the kind of gameplay in Splinter Cell Conviction falls into the stealth spectrum. Practically, you could be hiding in a safe spot lobbing frags and detonating mines and the enemies don\’t see you but you ARE creating a loud ruckus with your antics. I\’d say that this kind of gameplay falls dangerously on the border between the stealth spectrum and action region.

    \”Pure\” as in the word itself means :\”free from anything of a different, inferior, or contaminating kind.\” (Dictionary.com). Within the stealth spectrum that I spoke of, I don\’t see anything that adulterates or pollutes the concept of stealth. Hence the concept of \”Pure Stealth\” is probably a fallacy that we need to understand and abolish from our mindsets. Or we could end up respecting different opinions of what \”pure\” means to people and just accept them as opinions.

  7. I think there needs to be a move away from a certain type of protagonist (ie: ninja-types, capable fighters, etc) for stealth games at times.

    I\’d like to see (and this may exist, of course, and I\’ve just not realised) a stealth game where you play, for example, a child in a hostile war zone.

    I\’m envisioning a plot where you\’ve got a sandbox mechanic in a wartorn city. You\’re a child, so mainly powerless in a physical capacity (you\’re not going to be able to subdue an enemy soldier, for example) and you\’re trying to find the other members of your family, who you were separated from when the invasion/bombing/whatever began.

    You could use the environment to your advantage (a child can fit through gaps an adult can\’t) for escape, stealth and your wits to cause distractions or (if needed) damage to opponents. You follow a set of clues, unordered, to eventually reunite with the group.

    If the power was available to do it, I\’d have the city layout randomly generated in every new game, too (albeit along a certain set of parameters so that you don\’t end up with all the clues in one room and your family next door – although that, in itself, would be interesting purely for being evidence of how random life in that situation could be!).

    In my game, it\’s not about sneaking through dark tunnels and creeping past stationed baddies. It\’s about evading hostile eyes, escaping or avoiding patrols and even the potential to online co-op band up with others in your \”side\” to request help (\”I need someone to distract a guard so I can get into that building, can you help me?\”) in a team style game.

    I know this is all a bit flighty, but it\’s a game I\’d enjoy a lot more than the latest on-wires MGS style game with virtually fixed guardposts and \”hit this lever to kill that guard via an elaborate sequence you have no part of\” mechanics.

    Right, flame away! 😉

Leave a Reply

Your email address will not be published. Required fields are marked *