I thought I'd start the column off with something big: What is a game? If it isn't really what we thought it was, then what is it and how can a new definition help developers? Or rather, and let's be honest, how can it result in more net awesomeness in the games we play?
The book the Rules of Play looks very deeply into the definition of a game, and I think they reach a pretty good conclusion as to the most complete definition: "A game is a system in which players engage in an artificial conflict, defined by rules, that results in a quantifiable outcome." I think most game developers would agree with this. The problem with it is that besides constraining the notion of what can or cannot be a game in Ludology, it's not too practical for game developers.
A good rule for a creative field has to be a workable constraint that can be broken by appropriately skilled developers - a benchmark to ensure quality but one that doesn't discount experimentation. Like in Cormac McCarthy's The Road, he forgoes most punctuation and uses run-on sentences to great effect. He's very obviously writing "incorrectly," but he knows enough as a writer and a poet so that he knows which rules to break and when.
What sort of rules a game will follow or break are, mostly, decided when a game is pitched (this is a normal shooter, this is a shooter with a gimmick, this is a game where you play as a marshmallow, etc), but game development is rough, ever-changing work. The rule you thought you'd be able to be break in a cool way just doesn't work in reality. The single-save system you thought would create tension might make the game too hard and cause players to save constantly, turning your expected gameplay on its head. Sometimes it's because of the technology available, but most often the things that are reworked the most are the things players touch the most, and thusly many studios put an emphasis on playtesting. A working definition of what a game is has to include the influence that the act of playing has upon the game and its defining characteristics.
What's interesting to note also is that many purists don't consider gambling to be a game. Some people add the stipulation that a game, as a sort of make-believe, ceases to be a game once it affects the real world. This constraint rules out the more abstract and less useful musings of Johann Huizinga as having a place in a working definition. In his fantastic book, Homo Ludens, he argues that war and even life itself can be defined as a game; it makes sense, since many games mimic the rules of real life situations. But maybe this sort of crazy talk actually has more bearing than the sterile definition. Playing chess may very well improve the tactical aptitude of a king on the actual battlefield, and perhaps flight sims can prepare people to pilot drone planes.
But the purists and the Huizinga-ites each sort of teeter into their own extremes. The fact is that games do have some impact on real life, but not so much that one could theoretically say that life itself is a game. When my roommate jumps up from his chair and speeds to the kitchen to develop a meal at whirlwind speeds, he's doing so because in one minute he's going to be fighting the undead with twenty four other people. When I surprised my parents by ordering a pizza for dinner to eat by the computer, late into the night, I was doing so because I was involved in an important clan match and couldn't let down my team by leaving. So to say a game cannot impact reality is a bit misleading. While it's true my gun might not be pumping bullets into a real Nazi, I am really spending three hours pretending to pump bullets into a Nazi, and that happens in the real world.
There can't be any question that games have an effect on reality - it's what makes us gamers. I can tell stories about exploding shurikens, backflips and ninjas. I've pulled off shots I never thought possible, and just barely made some landings. I've pulled ahead in a race and broken into a sweat countering punches. These are the real experiences that our medium generates whether they're designed with intent or not. Just as this column should impart something to the reader, a game has to say something and create a certain experience. This is the sort of end result that should be helpful for game developers to study, or at least they should know what end result they're aiming for and how to achieve it.