I personally found the definitions of Roger Cailos and Chris Crawford to best fit my personal understanding and experience of what a game is. Both definitions try not to create an actual and concrete definition but instead create a series of term which can act as a rubric when questioning the nature of a particular activity. Crawford’s use of the term “Representation” expands upon Cailos’ “make-believe”; taken together, they strike at the very center of what a game is, the ultimate expression of the human capability for abstract thought.
One of the major complaints of this definition, however, was that of the nature of puzzles and simple games. Salen and Zimmerman ask whether tic-tac-toe requires any kind of make believe. I argue yes, that since tic-tac-toe does not represent any kind of physical concept, there is an increased demand upon the mind to accept it as a meaningful interaction between players. The most imaginative fantasy game, while more complex, is based upon pre-existing concepts and physical laws. For instance, the sword wielded by the players carries with it unstated implications of violent physical interaction, while the simple symbols of tic-tac-toe must be explained in their entirety to a first time player.
Thus I would say that a game could most simply be defined as the human capacity for abstraction applied to the broad arena of “play” in order to create a structured experience.
Probably my greatest issue with Greg Costikyan’s definition of games is the word “game tokens”. While other definitions certainly included terms that might be seen as restrictive, this is almost intentionally so. Cailo’s use of the term “make-believe” for instance, could be argued either as referring to the mind’s ability to assign value and meaning to abstract concepts while the term “game tokens” requires much more thinking and arguing to imply anything other than bright plastic pegs on a cardboard playing field. This serves to damage his intentions of labeling games as “art” as he suggests a very narrow definition which does away with out-of-the-box interpretations (an oftentimes essential aspect of art).
I am convinced, due to their structured nature that puzzles can be considered games. After all, if a puzzle with, say, 5 ways of winning would still be considered a puzzle, couldn’t a game simply be seen a puzzle wherein the number of possible winning combinations approaches infinity? Likewise, I am convinced that RPG’s like second life are not totally games but rather fit the definition of “play” into which “games” can be created.
Finally, I am completely convinced that due to their function of creating interaction of persons through purely abstract means, that games can easily be considered art.