Onboarding & Difficulty Curves

The 'learning game' is what a brand-new player actually plays (pattern recognition: options, feedback, am I okay?), while the 'real game' is the tuned deep experience. They are separate design problems.

Can a new player make their first meaningful decision within 90 seconds? Not a good decision — one where they understand 2-3 options and roughly why they'd pick each.

Teach the single core action first; everything else is an exception introduced when relevant. Rules needed in round three should be taught in round three.

Making the first move nearly forced so the player learns the interface by succeeding; board games can do this with a suggested setup or 'first game' variant.

A learning setup that deletes components beats a rulebook sidebar explaining them — gate complexity behind progress (Gloomhaven, Pandemic Legacy).

Every action should produce a visible, immediate consequence. If nothing on the table changes after a turn, the player learned nothing.

Challenge ramps up, dips briefly, then ramps higher. Dips let players consolidate skill and feel powerful; no dips = exhaustion, no ramps = boredom.

Plot skill against challenge: too much challenge → anxiety, too little → boredom. The playable band is narrow and moves as the player learns.

A diagnostic: if this transition takes more than about a quarter of the game's length for a first-timer, onboarding is too slow.

A new player mathematically out of contention by minute ten has a bad first impression. Soften with catch-up mechanisms, hidden scoring, or simultaneous rounds.