How Arcade Difficulty Increases

The complete list

Every way an arcade game can get harder falls into one of seven categories. Five are good, two are traps, and knowing which is which saves a great deal of tuning.

Speed and its wall

Speed is the first dial everyone reaches for and the most limited. Increasing scroll or object speed compresses the player's reaction window uniformly, which raises difficulty fast and then stops working entirely.

The wall is human reaction time. Once the interval between a hazard becoming visible and reaching you falls below about a fifth of a second, no amount of skill defeats it - the game has stopped measuring ability. Our games raise speed by roughly eleven percent per difficulty step and then cap it.

Density and pattern complexity

Density - more hazards per unit of distance - raises difficulty without touching the reaction window, which makes it far more sustainable. It also shifts the demand from reflex toward routing, which is usually more interesting.

Pattern complexity is better still, and it is the only dial with no ceiling. Same speed, same count, harder arrangements: two gates with openings on opposite sides, a slick patch before a lane blocker, a formation with a hidden hanging spike. A novel arrangement is challenging even to an expert in a way that a faster version of a known arrangement is not.

Spatial and perceptual constraint

Narrowing the playable area raises required precision without changing any timing. Lollipop Dash tightens its corridor; Candy Tunnel narrows as you descend. Both need a floor, because a corridor that narrows forever ends in a wall.

Reducing available information is the subtler version. Candy Tunnel shrinks your glow radius at the same depths at which it narrows, so you see less of a tighter space. This is powerful and easy to misuse: reducing information is fair only while the player still has enough to act on. A game that hides what you need is not difficult, it is arbitrary.

The two traps

Control degradation - adding input lag, unreliability or imprecision - makes a game feel broken rather than hard. The player's failures stop being about the game world and start being about the interface, and there is no skill that addresses that.

Increasing randomness is the subtler trap because it does increase failure rate. But it makes a game harder to learn from rather than harder to master: a player cannot improve at a pattern that is different every attempt, so thirty runs teach what one run would have. The failure rate goes up and the skill ceiling does not.

Stepped, not continuous

A continuously accelerating game is smoother and worse. Discrete steps at known thresholds have three advantages that outweigh smoothness.

• The player gets a moment to reorganise at each boundary.
• The schedule is learnable, which turns escalation into a plan rather than an ambush.
• Runs become comparable, so improvement is visible rather than obscured.

We publish every threshold on the relevant game page for exactly this reason. A player who knows a step arrives at five hundred metres can prepare; one who is surprised by it every run learns nothing from thirty attempts.

Key it to progress

What the curve is attached to matters as much as its shape. Time-based escalation punishes careful play arbitrarily - a slow player faces the same difficulty having travelled less. Random escalation makes runs incomparable. Progress-based escalation makes difficulty at any point in a run fixed and learnable.

All five of our endless games key difficulty to progress: distance, depth or height. None uses a timer and none uses a random roll.

Common mistakes

• Relying on speedIt hits a hard wall at human reaction time, after which the game measures nothing.
• Adding randomness to raise difficultyIt raises the failure rate without raising the skill ceiling, because patterns cannot be learned.
• Degrading controlsFailures become about the interface rather than the game, and no skill addresses that.
• Unbounded spatial or perceptual constraintA corridor that narrows forever ends in a wall; a light that shrinks forever ends in darkness.
• Continuous rather than stepped curvesSmoother and worse: no reorganising moment, no learnable schedule, no comparable runs.

Practical tips

• Learn the published step thresholds for any game you want to improve at.
• Expect pattern complexity, not speed, to be what eventually ends your runs.
• Use the boundary moment at each step to reposition and reassess.
• If a game feels unfair rather than hard, check for randomness or control degradation.
• In narrowing games, commit to a side early - precision demands rise before speed does.

Frequently Asked Questions

Read next

• how endless candy games increase difficulty — the endless case in detail
• designing fair candy collection levels — fairness alongside difficulty
• why deterministic hazards beat random ones — the randomness trap
• Candy Tunnel — spatial and perceptual constraint together
• Candy Factory — difficulty with zero randomness