I Need To Leave

A turn-based puzzle game about quietly leaving parties early.


I Need To Leave is a turn-based puzzle game about quietly leaving parties early. Duck and dance around monstrous party guests, reach the exit and leave without being spotted. The game uses simple pattern recognition puzzle mechanics to explore feelings of social anxiety and enochlophobia.

The games was developed independently over the course of a month, using Construct 3.


Project insight - mechanics breakdown

With ‘I Need To Leave’, I wanted to create a game about social phobias without relying too heavily on text or dialogue. I’m sure many people have experienced uncomfortable situations they have been unable to escape from. All of the game’s core mechanics were designed with a focus on expressing this theme:

1. Player movement

The player must navigate a grid-based play area, reaching the exit without being spotted by other party guests. The player-character, not brave enough to engage directly with the ‘enemies’, has no ability to stop or remove them. They can only use movement and waiting to avoid them en route to the exit. Player movement is turn-based. The player can move in one of four directions each turn, or choose to ‘wait’ and not move at all, although they can only ‘wait’ three times per level.

Turn-based puzzle games promote a more methodical play-style, allowing the player to take time and think through each move. They also don’t require much mechanical skill or practice from a player, making them more accessible to a wider audience.

Turn-based movement also has thematic significance in ‘I Need To Leave’. The player-character is slowed by anxiousness, desperate to stay out of sight. Their movement must be shrewd and calculated – simply running out of the room would create a scene, exactly the opposite of the player-character’s desire to stay unnoticed by anyone who might judge them for leaving the party early.

Players only have a set number of moves to complete each level. If they don’t reach the exit in that number of moves, their attempt to escape is foiled; a friend might come looking for them, someone might engage them in conversation or ask why they aren’t dancing. This mechanic also adds an additional challenge to a game that was otherwise quite susceptible to ‘brute force’ solutions.

2. Enemy design

Guests, depicted as monsters to reflect the anxiety, fear and sense of ‘otherness’ the player-character feels at the party, roam and rotate around the play area. The player is spotted whenever they end their turn in a space that is within a monsters’ line of sight. Being spotted triggers an enemy dialogue line, as the fellow party guest prevents the player-character’s escape or questions their motive for leaving.

When the player moves, the monsters do too. Different monsters all have different movement and line-of-sight patterns, made up of a maximum of four looping steps. These stay consistent from level to level, allowing the player to predict a monster’s movements and plan their tactics accordingly. There is a passivity about the monsters’ movement. They are not actively trying to seek out the player – it’s up to the player to avoid them. They’re at a party after all; their main aim is to have fun, even if the player-character isn’t. Gameplay becomes rhythmical, with player and monster dancing around each other.

With a puzzle game, it’s important to introduce new mechanics at the right time; not too early for the player to be unready for the additional challenge, and not too late for the player to have already become bored with the game. As ‘I Need To Leave’ is quite short (just 24 levels), I wanted to avoid introducing too many different types of enemies. I originally designed five for the game, and cut this to just four during playtesting.

3. Level design

Levels are designed to be claustrophobic, tightly-packed and unwelcoming.

The game’s play area is small. I originally envisaged it as an 8x8 grid subdivided by impassable walls, much like a maze. I wanted each level to look like the floorplan of a house hosting a houseparty. However, it was difficult to create interesting and challenging levels with this approach without decisively changing how the enemies detected the player (logically line-of-sight cannot pass through walls…). As a result I shifted to a more free-form level creation process, ditching the floorplan template and instead going for a more abstract representation of areas in a party.

Levels are tightly constrained, with pathways rarely wider than one tile. Often there is only one sequence of moves a player can make to complete a level, again alluding to the claustrophobia of crowds. One undesired side effect of this decision is that some levels do lend themselves more to trial and error rather than calculated movement.

Since the player has no ability to stop or remove the monsters, the board doesn’t open up as pieces are removed, like in chess – the space stays constrained, requiring the player to avoid, rather than directly face, their fears.