From one sketched screen to a playable game

Core mechanic

The light is the clock

The question. The sketch had torches and oil that keep the light alive. So what is the light really for, and what happens when it runs out?

What I decided. No timer on screen. The torch is the light, the timer and the life bar at once. A full torch burns for 60 seconds and the circle of light shrinks with it. Below 25 % it pulses and turns red. Oil adds 11 seconds, a new torch fills the flame back up.

Why it matters. One connection replaced three interface elements. It also made every later rule easy to read, because from now on everything in the maze either gives light or takes it.

Controls

One thumb is enough

The question. How do you move through a maze on a phone held in one hand, without covering half the screen with buttons?

What I decided. Drag anywhere and an invisible joystick follows your thumb, with a small dead zone against accidental drift. Movement is smooth, not tile by tile. A quiet line says “Drag anywhere to move” only until the first touch. On desktop, arrows or WASD.

Exploration

Darkness forgets

The question. If everything you've already seen stays visible, the maze becomes a map within a minute. Where is the tension then?

What I decided. Ragnar sees only what his torch lights and what walls don't block. Once he walks away, corridors fade back into darkness. No minimap, no arrows. You have to remember the way.

Guidance

Being lost needed an answer

The question. Removing the map created a new problem. A lost player stops having fun long before the torch dies. How do you help without taking away the feeling that they found the way themselves?

What I decided. The hint lives inside the world. Draft from the stairs bends the flame and its sparks towards the way out. The direction comes from real walking distance through the maze, so it never points into a wall. Near the exit a cold glow shows through the dark, and players get three chalk marks to draw arrows on the floor.

Obstacles

Traps became decisions

The question. The sketch had traps. But a trap that only punishes feels like bad luck. Could it ask for a decision instead?

What I decided. Spider webs block the ideal route. Burning through one costs 5 seconds of flame. Each has a detour that costs no flame up front but takes longer, and in the dark you can't see how much longer. With a weak flame you can't burn at all. Puddles slow you down and make the torch burn four times faster. Now the light connects to every choice.

Character

A monster that talks first

The question. A maze needs a threat. How do you teach a rule without a tooltip, and without punishing a player who didn't know it yet?

What I decided. A draugr blocks one corridor, never the only way to the key or the gate, and a sword is always within reach. Bump into it unarmed and it first insults you, then louder, and only the third time it hits back for 5 seconds of flame. Ragnar then thinks out loud that he should maybe find a sword first. The humour carries the instruction.

Onboarding

A story instead of a tutorial

The question. Why is a viking in a dark maze in the first place, and how do you say it in under half a minute?

What I decided. A four-panel comic played as a short film: a battle, a victory feast, one more horn of mead, and Ragnar waking up in the maze. The skip button is visible the whole time.

Iteration

Atmosphere, not noise

The idea. Rats, beetles and bats to make the maze feel alive. Harmless, only there for mood.

What went wrong. In the first version they crossed the player about every four seconds. Atmosphere turned into noise and pulled attention from the maze.

What I changed. Now they flee from the light and slowly drift away when they are far from Ragnar.

Rating

Rewarding the right thing

The question. A run takes about a minute. What should the result measure, so that it pushes players towards the way the game is meant to be played?

What I decided. Stars compare your time with the ideal run of each map, the shortest route from start to key, gate and stairs. Get close to it and you earn five stars. The slower you are, the fewer you get, but finishing always earns at least one. Your best time for each map is saved and shown under the result, so there is always a reason for one more try.

Why time, not flame. Counting the flame left at the end would teach players to hoard light and avoid every risk, which is the opposite of what the game is about. Counting time rewards confident exploring and quick decisions at every web and every crossing.

Iteration

The last free slot

The idea. The inventory had a free fourth slot. I wanted a one-use emergency guide, hidden in a dead end, that shows the way to the exit, never to the key. Using it before you have the key is your own problem.

How it changed. The first idea was a raven. It became a Raidho rune stone that releases a blue light orb. The orb races ahead faster than you can follow, leaves a fading trail of sparks and flashes red when it passes a draugr. Ragnar's line when he picks it up depends on whether he already has the key.

Art direction

Look and sound

The question. How do you get one consistent world out of tools that each generate something slightly different?

What I decided. A flat vector style for Ragnar, the draugr, items and tiles, then clean-up by hand: backgrounds removed, everything recoloured to one palette, the flame redrawn in code so it can react to the light. For music I chose a darker Nordic track that doesn't loop the same motif. Sound can be turned off on the first screen and the game remembers it.