Learning

About this model

The diagram represents a simple game learning mechanism, where the outcome of the games played (either win or lose) affects the chance of winning in subsequent games. Initially, resources flow from a source labeled "Play" to two gates which then direct resources to either a "Win" or "Lose" pool, based on simulated win and lose probabilities. The "Games Played" pool accumulates resources, representing the number of games played over time.

Central to this model is a "Win Chance" register, which dynamically adjusts the win and lose probabilities based on the number of games played. Each game played increases the win probability slightly, reflecting a learning effect where the more games played, the better the chance of winning. This mechanic is facilitated by state connections altering the win and lose probabilities in response to changes in the "Games Played" pool, effectively simulating a player's improving skill or learning over time. The diagram captures the fundamental dynamics of game outcomes affecting future performance, providing a basic model of learning and adaptation within a gaming context.

“Learning” is a public Machinations diagram by Harry Ashton, shared in the Machinations Community.

You can open the original diagram in the Machinations app, press Play to watch how resources move through it, and change its values to see how the system responds. Machinations is a visual tool for designing, simulating and balancing game economies and other systems, where resources flow through sources, pools, drains, converters and gates.