Ask a designer how hard their game is and you will get a number or a comparison. Ask a player and you will get a story about a specific moment. Difficulty is not a scalar — it is a curve, and the curve's shape does most of the emotional work.

The line

A straight ramp: every minute is slightly harder than the last. Endless runners, snake, most arcade score-chasers.

The line is honest and cheap to build, but it has a flaw: it has no landmarks. Players cannot say where they got good, only how far they got. That is why pure line games need a number to talk about — the score becomes the only narrative.

Lines work when the skill being tested is continuous (aim, timing, reaction) and fail when the skill is discrete (a rule you either know or do not).

The stair

Flat stretches, then a step up. Levels, waves, chapters.

The stair is the most legible shape in games, because each step is a memory: "I got past the third wave." It also has a specific failure mode — the plateau. If a flat stretch lasts longer than the skill takes to master, players get bored *before* the next step. Most level-based games that feel slow are not too hard; they are too flat for too long.

A step is only exciting if the flat before it was long enough to feel safe, and short enough to feel restless.

The wave

Hard, easy, hard, easy. The tension-and-release curve.

The wave is what arcade designers discovered by accident and what horror games use deliberately: a hard moment makes the following easy moment feel like relief, and the relief makes the next hard moment land harder. A game that is uniformly hard stops feeling hard — it just feels tiring.

The tuning knob for a wave is the ratio, not the amplitude. Two minutes of pressure to forty seconds of calm reads as generous. Two minutes to two seconds reads as cruel.

The wall

A sudden, large increase with no ramp. The boss fight, the difficulty spike, the level everyone quits.

Walls are usually described as bad design, which is only half true. A wall is bad when it is invisible — when the player cannot tell what they did wrong, or cannot practise the thing that killed them. A wall is good when it is a clear exam of a skill the game has been teaching.

The distinction is *telegraphing*. A wall you can see coming is a goal. A wall that appears without warning is a trap.

Why short games need shape more than long games

A three-minute game has no time to build a curve, so it borrows one: the run. Start easy, get harder, end. That single wave is enough structure for a session, and it is why arcade games from 1980 still feel well paced.

A long game with no shape feels formless even if every individual minute is fine. This is the most common failure in amateur design: each level is good, and the sequence is not.

The practical question

The question is not "how hard should this be" but "what shape is the player's confidence over time".

Confidence should rise, dip, and rise again. A game where confidence only falls is a game people describe as unfair. A game where confidence only rises is a game people describe as boring — and they will say so even while praising it.

Difficulty is the shape of confidence. Everything else is a number.