I wanted to make Sub Terra scare you. In the first half of this post, I discussed why people enjoy horror experiences, what makes them feel anxious and scared, and what sort of threats we can deploy in a board game setting to push them in this direction.


Is there anything else we can do to turn up the terror?


Our brains can be thought of as having two main modes of thinking.

  • The first is referred to as System 1, and exists to very quickly make subconscious, approximate judgements based on limited data without expending much energy. The idea is to improve your reaction time to complex stimuli, so your brain goes “TIGER!” rather than “Hmm, that pattern of orange and white stripes in this nearby bush looks an awful lot like that large predator I saw yesterday, let’s take some time to make sure”. This is helpful for you, though admittedly less helpful for the tiger.
  • The second is referred to as System 2, and requires a lot more effort to think consciously, logically and deliberately to get much more accurate results. Not only does this allow us to perform more complex tasks, it also allows us to override the approximate snap-judgements of System 1 – after your brain has gone “TIGER!” and you’re sensibly running away, subsequent processing might reveal that what you thought was a tiger was actually a tiger mask being worn by your friend Steve, ever the prankster. (No-one likes you, Steve)


Because it’s fast and cheap, System 1 does most of our thinking, and it’s usually pretty reliable and good at keeping us alive. However, it can make mistakes, which requires the slow and expensive System 2 to step in and override at some later point (if it can, or if we even notice).  It’s the gap between these two systems that causes us to feel fear even in safe settings – our brains are just trying to protect us.

This gap also leads to a number of ‘cognitive biases’, or intuitive mistakes that humans have a strong tendency to make again and again.  A couple of these are particularly useful for our sinister purposes:

  • The Availability Heuristic
    Humans are terrible at intuitively grasping probabilities. This mental shortcut is one of the things responsible. Essentially, events that are easier to vividly imagine or recall (more “available”) are intuitively seen as more likely. This is the driver behind many irrational fears (such as flying, or terrorism). Even when the outcomes are much less extreme, people will instinctively wager more than they should on unlikely outcomes, because the unlikely outcomes are given similar prominence to the likely outcomes.

    This suggests that having low probabilities of awful outcomes will make situations feel more dangerous than they actually are. (We also need to be careful to not do the opposite – low probability miracles can give false hope, which runs counter to our experiential goals)


  • Risk Aversion / Ambiguity Effect
    These related biases imply that people will usually prefer to choose a more certain outcome over a more uncertain outcome, even if the expected value of the latter choice is higher overall. In a typical gaming environment, it’s often too much effort to work out exact probabilities for complex choices involving hidden information (e.g. draw a new card from a large varied deck, with a bunch of cards already discarded or hidden in private hands), so there’s a tendency to prefer options with “obvious” probabilities (e.g. a coin flip) or fixed results.

    Viewed from a different angle, this could be due to the complex probability calculations acting as a way of creating genuine uncertainty in a game with fully visible rules (math is hard and/or slow). As mentioned above, this information gap creates Anxiety.

    It’s also important to note that you should try give players a choice here. If they have to accept a random outcome from an “unknown” probability distribution, then they can blame bad outcomes entirely on chance (“stupid random game!”). If they’ve consciously opted in, then the reveal of the outcome becomes much more tense (“did I do the right thing?”).



Using the above theory, our ideal horror board game needs an ongoing sequence of uncertain threats that:

  1. Each increase your chance of losing the game
  2. Are sufficiently relatable/intuitive/immersive that it’s easy to project yourself into the player avatar
  3. Take advantage of cognitive biases to make choices seem scarier than they probably are.


Sub Terra tries to put all this theory into practice in the following ways:

  • Relatable player-to-avatar link
    While it’s unlikely that all players will be seasoned spelunkers, they will usually be a group of friends who, if transported to this cave setting in real life, would probably act much the same. This is important – rather than pretending to be “Captain John Smith”, Steve is free to play himself (in all his capricious irritating glory), which helps with immersion. Each player is fully in control of their character’s actions, and cannot directly force another character to commit an action against their will. You take risks, and you suffer the consequences.

    The fact that the game is fully cooperative is very deliberate. If the players were competing, or had hidden motives, they cease to be a group of friends and start playing assuming fictional roles (as competitive as most groups are, I’m pretty sure they don’t want to actually kill each other (usually)). This isn’t to say that competition or hidden motives can’t work in horror games – Dead of Winter and Battlestar Galactica are great examples of traitor mechanics creating strong paranoia, just they undermine this relatability aspect to immersion.

    To create further investment in your character, if you lose all your health points, you fall unconscious and are unable to perform actions until someone rescues and revives you.  Being in this state deliberately isn’t fun, but it’s nearly always your fault, as you could have been more careful.  It also provides a social reason for your friends to rescue you as quickly as possible so you can rejoin the game.  (Because the game is cooperative, this temporary elimination isn’t as bad as it might first appear because you’re still able to participate in group decision making)

    Finally, the victory scoring is asymmetric – you only score if your character escapes, which adds extra incentive to keep your character alive at the expense of others (else you can’t “win”), and also adds extra incentive to rescue others at risk to yourself (else your friend can’t “win”).


  • Intuitive environment
    While I’m not going to pretend that the Sub Terra cave is “realistic” (it’s wonderfully lit by glowing mushrooms, vents choking clouds of suspiciously green gas, and may contain terrifying shadow monsters) it does at least try to be consistent with what people intuitively understand about “caves” – for example:
  • You can’t move through solid walls
  • You can’t see very far in the darkness
  • It’s hard to swim through water
  • It’s also hard to squeeze through narrow passageways
  • If lots of heavy rocks fall on you, it will hurt
  • If there are lots of heavy rocks blocking the tunnel, you can’t move through them
  • Gas clouds stick around for a bit before dispersing
  • Some ledges need ropes to be set up before you can cross them
  • You can’t live here – you need to find the way out to survive


Not only does this help with rules comprehension (lowering complexity by piggybacking on known concepts), but it tries to make the cave feel less like an abstract numerical simulation and more like an actual place you’re moving through. I’m not saying this is perfect and just like virtual reality (unless you have a really good imagination), but it’s supposed to help draw you in just a little bit further than if the mechanics and theme didn’t match.

  • Shifting randomness
    Sub Terra uses a shuffled stack of around 60 cave tiles to incrementally create the cave. While the composition of the stack is known in advance, you still don’t know exactly what’s around each corner, and the probabilities of this are constantly shifting.

    Every turn, a hazard card is revealed and resolved from the top of a shuffled deck, causing bad things to happen. Again, the composition of the deck is known in advance, but because it’s a draw without replacement, the odds of a certain card appearing will shift as more are revealed.

    These contribute to the Anxiety felt by the players, and the shifting probabilities help prevent each action being easily “solvable” (and even if they are solved, your underlying biases will affect your judgement).
  • Low probability of awful events
    While most tiles and hazard cards are just incrementally bad, some of them (“cave-in” tiles/cards) have the potential to take down a player who’s on full health if they’re standing in the wrong place at the wrong time, even though this is unlikely. (In Sub Terra, you aren’t eliminated if you run out of health, but you do fall unconscious and need to be rescued by another player, which is a big setback for the team.)

    The possibility of this outcome makes these tiles and events feel more dangerous than they actually are, through the availability heuristic. It also makes revealing a cave-in card when there are multiple players on cave-in tiles a very exciting moment – a subsequent die roll is required to see which tiles actually cave-in, and the impact on the game from an unlucky roll is large. This doesn’t feel particularly unfair, though, as players usually have a choice to move off these tiles by risking a small amount of health (‘exerting’).
  • Invincible, deadly, unknown monsters
    Unlike action horror games like Zombicide which threaten to drown you in tides of zombies, the point of Sub Terra is to make you feel isolated and powerless. In my opinion, the best horror films are those that reveal as little about the monster(s) as possible, and are those where the protagonists are hopelessly outmatched and have to run or hide rather than fight directly (Alien, The Descent, It Follows…).

    To capture this feeling, the monsters in Sub Terra are kept deliberately abstract to allow the players’ imaginations to fill in the gaps. They’re impossible to directly remove (other than by a single character “class”, who can’t be everywhere at once), and direct contact with them will immediately remove all your health and knock you unconscious. To compensate for this, they move very slowly – it’s very possible to outrun them, but to do so usually pushes you into taking more risks, which can in turn slow you down. The only real way to deal with them if you’re trapped is to try and hide, which takes a full turn, requires a dice roll to work, and is harder to pull off the more players are nearby.
  • Time pressure
    Against all of this adversity, you need to escape the cave in a finite number of turns. This is a finite and dwindling resource that you’ll need to allocate wisely between fast risky exploration, cautious steps into the unknown, and rescuing trapped friends. At the end of the game, each player has a 50% chance of becoming lost in the darkness at the end of each turn – it’s technically possible to have an unlimited number of turns to escape, which gives a small sliver of hope to even the most dire of situations.

    The game doesn’t yet have a real-time element to it, but that’s certainly something that could cause additional pressure – shared turn timers felt like they’d cause more problems than they’d solve. If you feel up to the challenge, try playing the game with a strict 60 minute time limit! (or less!)



So that’s how the design of Sub Terra mechanically tries to make you feel scared! Given that it’s a board game, with the usual static board game limitations, I think this works pretty well. This is definitely a rich seam of design space to mine, and I’d love to see more games out there that implement these techniques far more skillfully than mine.

In my next post, I’ll discuss cooperative games, the problems these games usually have, and how Sub Terra tries to avoid making the same mistakes. See you then!

Tim Pinder

Tim Pinder

Game Designer, Sub Terra

Indie gamedev robot. All opinions, fatalities and apocalypses are the responsibility of my creators, who should have worked harder on the control problem