In this series of posts I'll be looking at their ideas.
Activity-based framework
This model begins by splitting the game into states, and grouping these under four categories:
Holistic items relate to the whole game, and how it is initialised and played. This is divided into game instance (the context), game session (each player's journey through the game), play session (each turn, or section such as a level) and extra-game activities (such as high-score tables or customising the game appearance).
Boundary items specify game mechanics: rules, modes of play and goals. Goals are vital in any game, as they introduce structure. Modes of play are best illustrated by Pac-Man, where the player is chased by ghosts for much of the time; however, at certain points the play switches mode and the player chases the ghosts.
Temporal components relate to the flow of the game. This encompasses actions (active decisions by the player), events (automated responses to decisions), closures (changes of game state), end conditions (such as all players finishing), and evaluation functions (algorithms which calculate when closures or end conditions should be triggered).
Finally, structural components are about the people and technology: facilitator (e.g. banker in Monopoly), players, elements (board, pieces, weapons, environment), interface (control & feedback) and game time. This last one is not always needed.
This is an interesting analysis, but I don't know how useful it is. As mentioned above, the authors use Pac-Man as their example. However, I suspect that it is better suited to board games.