As a programmer/designer/part-time-robot, I usually think of games as complex mathematical systems, with players trying to rationally extract as much value from their actions as possible.  After all, this is how the field of artificial intelligence has treated games for the last sixty years and into the foreseeable future (at least until a neural net somewhere decides the best way to win is to delete all opponents and wipes us all out).

However, this robotic approach is very clearly not how human players learn and play games.  That’s not a dig at puny human brains (so tiny) – it’s more that millions of years of evolution has equipped humans with a bunch of really useful mental shortcuts to quickly absorb and process certain kinds of information.  Taking advantage of these shortcuts can turn complex mathematical systems into games that are approachable, intuitive and fun. Theme, story and visuals can do far more than just look pretty.

In this article, I’ll explore some of the ways games can use visual and thematic elements to make gameplay more intuitive and to manage player attention.  I’ll then discuss how and why these techniques were used in Sub Terra II, and show off some more of Diana’s illustrations along the way.

Let’s take a look!



Do all games need a theme?


Well, some games don’t.  Abstract games with simple rules like Tic-Tac-Toe, Go or Blokus choose not to, and their minimal aesthetic is part of their charm.  But add a few more rules to the mix and you end up with something like Chess, Shogi or Onitama, which all feature pieces and actions with distinct names and effects.  Why?

Mark Rosewater (Magic: The Gathering head designer, whom I cite endlessly) defines “resonance” as “[building] a game component on top of information the audience is already familiar with”.  In the case of western Chess, you have six different pieces:

  • Pawn
    The weakest piece, but much more numerous.


  • King
    The most important piece on the board.  Defend at all costs. Go patriarchy!


  • Queen
    The second most important piece on the board, and the most powerful.  (The history of this piece is really interesting, actually – the piece was introduced and grew in power over the last millennium to reflect changing attitudes towards female monarchs, including resistance from ye olde bigots calling these new variants “madwoman’s chess”)


  • Bishop / Knight / Rook
    More important than pawns, less important than royalty.  (“Rook” isn’t immediately recognisable, but started life as an armoured war chariot and morphed into a tower)

The name of a piece encodes a lot of information about its relative importance, if not too much about its actual abilities.  As a traditional game in the public domain, chess pieces can vary wildly in appearance, but in general pawns will be smaller than the others, and the royalty larger.  As you learn to play, you can use these external concepts to anchor the specific rules for each piece.

This is a simple, obvious example.  Here are some more:


  • Health points are removed when you get hurt, and something bad happens when you run out
  • Money can be exchanged for goods and services
  • You gather wood from forests, and stone from mountains
  • Fighting the bad guys is risky, but can make them go away
  • Having more workers lets you do more things, but you have to pay more to keep them

These echo basic, universal truths about how our world functions – they’re so obvious I feel silly listing them.  But because they’re so obvious, you don’t have to teach or remind players how they work, which can dramatically lower your game’s barrier to entry.  (consider: can I spend health to buy items? Will fighting bad guys heal me? Is there water in this desert?)

You’re not just limited to universal truths, though.  Take zombies.


  • Form large hordes
  • Are slow (OK, sometimes scarily fast)
  • Are stupid
  • Want to eat you
  • Aren’t that threatening unless they gang up
  • Are relatively easy to destroy, if you aim for the head
  • Are already dead (or mostly dead), so attacking them is guilt-free

If the zombies in your game function like this, then the rules surrounding them suddenly become easier to learn, as players just have to map the mechanics onto their already-acquired mental image.


Pop-culture tropes can be a great way to hide complexity from the player.  However, you have to be careful that the tropes you’re referencing are generally understood (or are easily understandable) by your target audience.  Zombies are very recognisable villains to most of us, but players exist (such as my parents) who haven’t seen any zombie films or shows, and will not know how they’re expected to behave.  They’re pretty intuitive once you understand the “mindless, shambling horde of dead people” thing, though.

Compare them to vampires:


These have had many different interpretations over the years, and it’s not going to be immediately clear whether your vamps are the ones from Dracula, Blade, Twilight, Buffy or Sesame Street.  Don’t get me wrong – we all need more vampire games in our lives – but you probably can’t let the concept of “vampires” do as much heavy-lifting as “zombies”, because first you’ll need to explain what kind of vampires they are.



Resonance isn’t just limited to text and tropes.  Picture This: How Pictures Work by Molly Bang relates simple visual elements in illustrations to the emotions they make us feel, including:


Vertical shapes are active and exciting (unstable, fighting against gravity)
Diagonal shapes are dynamic and transitional (implying motion or tension)
Flat, horizontal shapes are stable and calm (at rest)

High objects are stronger and happier (conquering gravity)
Low objects are weaker, grounded and constrained (conquered by gravity)

Relatively large objects indicate strength and power
Relatively small objects indicate weakness

Lighter images feel safe and welcoming (illumination, visibility)
Darker images feel serious and dangerous (night, unknown)

Sharp pointed shapes feel dangerous
Soft curved shapes feel safe and protective

These principles can be used to make players both feel a certain way about specific game elements (“these spikes are dangerous”) and effectively compare different game elements (“the giant hell-beast is more dangerous than the tiny hell-beast”).


Colour can also have specific associations for certain audiences, though be aware that the meaning of certain colours can change drastically based on where you are in the world (lots of guides exist – here’s one).  However, colours used to reference natural concepts are more universal (e.g. red = heat, fire, blood).

If your game involves humans (or humanoids), then you can also use pose, body-language and facial expressions to indicate how these characters function.  A confident pose can imply safety and power. A dynamic pose can imply speed and agility.

Well-known equipment and tools can be used to imply functionality as well: shields keep you safe, swords can hurt, guns have range.

This should all seem obvious, which is kind of the point – you want your audience to immediately and subconsciously connect an illustration with the gameplay behind it.  If you’re working with an illustrator, they’ll definitely know all this! What they won’t know (unless you tell them) is how your specific gameplay components work, and what each illustration needs to convey to the player.



Board games will have lots of illustrated components, each conveying complex information, often arranged unpredictably.  If we’re not careful, this can be quite overwhelming.

Clearly communicating relevant information is the main responsibility of graphic design (very roughly: illustration = fancy pictures; graphic design = symbols/text/layout).  If you’re working with a graphic designer, they’ll know how to emphasize information appropriately, provided you let them know which gameplay elements need emphasis.

Ideally, we would like players to expend as little effort as possible in order to:

  • Search
    Locate specific gameplay elements on demand.


  • Prioritize
    Order gameplay information by importance, or to easily locate important information.

The key to optimizing search is visually grouping similar elements and visually separating different elements (Gestalt theory).  Here are a few different ways to achieve this:

Screens handle colour via an additive red-green-blue model (RGB), printers use subtractive cyan-magenta-yellow-black (CMYK).  However, for our purposes it’s best to follow the brain and think in terms of hue-saturation-lightness (HSL).

Hue is one component of this, and is what most people will think of when you use the word “colour” (red, green, blue etc.), and can be illustrated by a colour wheel:

A single dominant hue is a good way to uniquely identify a particular gameplay element.  Components with the same hue will be treated as having some identical property (e.g. they all belong to the same player).  Components with adjacent hues (blue/cyan) can feel like they have similar properties, while components with sharply contrasting hues (blue/orange) should feel radically different and be easier to mentally separate.

Unfortunately, ~5% of your audience will have some form of colour blindness.  The different forms this can take means there’s no colour-based solution that can accommodate everyone, so the best approach is to use unique high-contrast iconography to separate critical elements.  However, the most common type (Deuteranopia) confuses red and green hues, so using blue and non-blue hues should still be effective separation for most players.  (Here’s a site that can preview images for different forms of blindness)

I’ll cover saturation and lightness more in the following section, but these can also be used to visually separate elements, just a bit less efficiently (i.e. this requires larger saturation/lightness swings, and two saturations of the same hue will still feel somewhat related).

Here’s a useful list of 22 reasonably distinct colours (including accessibility filters) which might be a helpful starting point if your game has an excessive number of different characters/factions.

  • Position
    If your game contains a lot of similar components, keeping a consistent layout across those components makes it easier to locate similar information (e.g. action cards with a cost-circle and effect-box).

    However, some components won’t always have a defined orientation (e.g. tiles) – in this case, you’ll need your layout to have full rotational symmetry, which usually means using the center of the component to hold all information (Kerala) or adding some kind of border (Castles of Burgundy).

    Related pieces of information on the same component should be placed close to each other.  Different types of information should be in different places in the layout.


This applies to both the illustrations printed onto components (cards, tokens, boards) and the components themselves (meeples, discs, cubes).  Similar symbol and component shapes imply a similar gameplay purpose.

  • (For example, Castles of Burgundy uses differently-sized circular, square and hexagonal tokens as clear separators between component types)

Prioritization is a little more involved.  Here are a few different ways to raise the priority of information:

Larger elements attract more attention than smaller elements, and are visible at greater distances.

Separating an element from similar objects or background details will draw attention to it.

Elements closer to the center of a (fully visible) component will draw more attention than those off to one side.

However, note that components aren’t always fully visible – a hand of cards is usually held in a fan with only the corners showing, so information that’s important in this state should be placed in the corner.  Dominion is a good example of this: the cost of a card is placed in the lower-left corner (visible when in the market but deliberately obscured in hand, as now irrelevant), while the value of money cards is placed in the top-left corner (clearly visible in hand, very relevant here).

  • Read Order
    If the component has a defined orientation (e.g. cards are always viewed vertically), then the top of the component is higher priority than the bottom, as players read things from the top down.  (There’s a bit more to it than this – see this article on Gutenberg, Z and F layouts for more details)

Increase Saturation / Lightness
Saturation and Lightness are the other two components of colour, next to Hue (covered above).  Saturation controls how vibrant the colour is, and Lightness controls brightness:

Contrast (Colour / Shape)
Sudden breaks in a pattern attract attention.




If all else fails, just point at the thing you want players to notice.  This doesn’t have to be an actual arrow – perspective lines can also draw the eye in a certain direction.



How does Sub Terra II use all this?

The resonant heart of Sub Terra II is the fantasy of the intrepid explorer.  The Indiana Jones franchise is the most currently visible example (maybe Jumanji?) – though Indy was referencing earlier tropes popularized in the serials and magazines of the 1930s, which themselves reference the adventurous science fiction of Jules Verne and Arthur Conan Doyle.

All this source material has been riffed on in countless ways across a wide variety of media, so it’s likely that players will immediately understand their quest (explore volcano temple, grab idol, escape) and the hazards they’ll face along the way (traps, lava, collapsing ruins).

The core mechanics of the game deliberately resemble what the game’s characters would be able to do in this particular situation.  Actions take time (action points) to perform, and encompass looking, moving, resting, digging and fighting. The temple is dangerous, with different obstacles able to hurt you in different ways, depleting a pool of health.  When that health runs out, you fall over and have to crawl back to safety. Eventually, the volcano will erupt, and a deadly wave of lava will chase you out of the temple like a really visible molten boulder. All of this should make the game easier to learn.

Each playable character has two unique special abilities.  The illustration for each character tries to convey both what those abilities can do and how you should probably play that character.  The hope is that first-time players will pick someone they like the look of and be able to perform reasonably well just by role-playing.

As an example of this, here are the Guide and the Marksman.  Take a second to think how these characters might behave.

The Guide works best as a frontline explorer.  You’ll often run off by yourself and map out a lot of the temple on your own, as your abilities help you squeeze through tiny gaps and reveal tiles at a faster rate.  However, you’re quite fragile, and will need the support of others if you want to survive. The illustration tries to convey this by depicting a happy confident kid off on another adventure, looking around in wonder as he uncovers more of the temple.  However, he’s still just a kid, without any training, armour or sense of caution, and is probably going to get into trouble if he runs off by himself (this happens a lot).


The Marksman works best as a stationary midrange defender.  You’ll find a safe spot to hole up with good sightlines, and protect the rest of your team from distant enemies before they can get close.  The illustration tries to convey this with an “alert sniper” pose – she’s not moving, but she’s on the lookout for danger, and can quickly raise her rifle to shoot when she finds some.



The most visible part of Sub Terra II is the temple, which is constructed over the course of the game by placing these square tiles:



Each tile has different properties and connections.  Players move their meeple avatar from tile to tile, carefully uncovering more of the temple, with the aim of reaching the inner sanctum.

On any given turn, this is the temple-related information that’s most relevant to each player, in order:


  1. Their current location
  2. The locations of the other players, and their status (active/downed)
  3. The locations of any enemies
  4. The route from their current tile to any other tile
  5. The relative threat of nearby tiles

Players and enemies are represented by wooden meeples (or plastic miniatures) that sit on top of the flat temple tiles, making them very easy to locate at all times.  Like most games, each explorer meeple is a unique colour that matches the character board in front of each player, making it relatively easy to see where everyone is [1,2].  Enemy meeples are a similar size but a noticeably different shape [3].

To make pathfinding easier [4], each tile has a thick dark border region on sides that cannot be moved through.  The area where players can stand is considerably and consistently brighter, and fairly grid-like thanks to the man-made temple setting.  The default light brown floor colour is extended through easily passable tiles (bridge, lava, ruins) to indicate their lack of obstructions.

Tiles that cause more problems have floor colours and textures that begin to deviate from this (traps, guardians).

When a ruin tile is blocked, the rubble token covers the passable floor area and makes the tile appear much darker. When the token is removed, the tile visually changes to the lighter state.

Indicating relative threat is a bit more involved [5].  Safe tiles (normal, bridge, key) all primarily use a default brown colour in their illustrations (low saturation, orange/red hue), though still use brightness to differentiate themselves.

Moderate threats make their centrally-placed symbols more visible, deviating slightly from the default by becoming more saturated (lava) or shifting hue (ruin).

High threats more noticeably shift their hue (guardians).  These want to grab your attention even if you’re nowhere near them, as you’ll need to plan around them.  Players also need to be able to quickly find all guardian tiles in play when spawning new enemies, so the guardian symbol is deliberately very high contrast and different to the other symbols in the game (white, not black).

All of these effects are relatively minor, but cumulative.  Each composition needs to strike a balance between aesthetics (“this looks good”), resonance (“I understand what this is”) and intuition (“I understand what this does”).  We’re still some way off manufacturing this, and I imagine things will continue to shift as we proceed.



I hope that all made sense!  I’m definitely not an artist, so I’m sure there are a bunch of techniques I’ve missed or misinterpreted – please berate me in the comments.  But hopefully you’ll now think about theme and visuals as valuable design tools instead of just interchangeable fancy hats atop your math-mannequin.


Next week I’ll be discussing player motivations – the research that’s been done to sort us neatly (or not-so-neatly) into various boxes, and how this can be used to make games more engaging.  See you then!