It may seem strange to hit this topic now; I haven’t even defined what I think a game is! For now, I’m going to put that question off, but say that there are three types of games:

1. Puzzles – These games involve one (or more players) trying to solve some system. These games can be played with some goal: ‘how do I solve the puzzle?’ or in a more exploratory manner: ‘what exactly can I do?” Games that are puzzles include:
   • Computer Games: Platformers, Adventure Games, Simulations (Civilization/Sim City)
   • Physical Games: Rubik’s Cube, Mind-Benders, Logic Puzzles
2. Competitive Games – These games involve multiple players competing against each other within the constraints of some rule system with defined victory conditions. Some of the players CAN be AI simulated; as such, these games can be played as puzzles.
   • Computer Games: Starcraft, League of Legends, Street Fighter, HALO, Simulations (Civilization/Sim City).
   • Physical Games: Chess, Basketball
3. Collaborative Games – These games involve multiple players working together (potentially somewhat competitively) to create something.
   • Roleplaying Games
   • Team-building games?

If these distinctions seem meaningless, consider what sort of questions a player might ask about each type of game?

The question that comes to mind for puzzles is “How do I solve X?” (How do I kill some boss/find some item/whatever)

The question that comes to mind for competitive games is “How do I get better at X?” with the sense of “How do I win more games?”

The question that comes to mind with collaborative games is “How do I get better at X?” with the sense of “How do I make a better game experience?”

You wouldn’t really ask how to “solve” chess, or how to “get better” at MegaMan. You may ask how to “get better at a Rubik’s Cube”, but in the sense of asking for a series of rules to perform some transformation, whereas “getting better” at a competitive game is more about building instincts and learning to react to and read another person.

So first, what makes a good game in general? Obviously ‘fun’ jumps to mind, but that seems a bit of a cop-out. Here are some qualities that I’ve found make games naturally fun:

1. Accessibility of Learning Curves – A game should be relatively easy to understand the basics of. We’ve all played that game that doesn’t explain its controls well, or has a 500 page rule manual with alternate books, and there may very well be good reasons to HAVE those rules, but the payoffs have to be worth it. The best way a game can do this is to make a continuum from simple play to complicated play, where complicated play contains elements of simple play, but strung together in complicated ways. For example, in the single-player campaign of Starcraft, you start off by learning just to move units around and control them. You have no buildings from which to build units, no resource gathering, and so on. Once you’ve mastered unit- manipulation, you start to get bases, but those bases are more simple, and get more complex over time. This provides a neat transition into the multiplayer version, where everything is available to you. On the other hand, while theoretically my favorite board game (Twilight Imperium, 3rd edition) and rpg (Spycraft 2.0) are fun, I’d have to find 3-8 other people willing to learn all of the rules to take advantage of them, or I may as well play a less complicated game.
2. Depth of Difficulty – A game should also be rewarding to master. You can play through many platformers on an easy difficulty by just jumping and shooting. But there are skills you can master (wall jumping, double jumping, animation-canceling) and so on that may be necessary to beat the game on higher difficulties. The same holds for competitive games; Super Smash Bros. is playable by casual players, but competitive play has more tools available. This gives a game replay value, and also opens up the potential for it to be an e-sport. If there are no complicated moves to be pulled off, a competitive scene will not develop (there’s a reason League of Legends and Starcraft and Super Smash Bros are thriving e-sports, but not Pac-Man).
3. Visual Clarity – A game should have clear visual markers for what’s happened. Think of the “POW” and “BLAM” in comic books. You may not see what happened, but you know somebody has been struck. Visual clarity makes simple play more simple, and makes thinking about and reacting to complicated play more possible.

So what makes a good puzzle? A good puzzle is one that is straightforward to solve. Do I mean that it should be simple or easy? No. Instead what I mean is that a good puzzle is much like a good multiple-choice test: there is no guessing necessary, but you may have to do deductive or inductive work to figure the right answer out. That is not to say that a good puzzle doesn’t ALLOW guessing or experimentation; games which do are often more satisfying, but it should be solvable without them. Otherwise, you start to fall into a trap many games have; try every combination until one works out, which just isn’t fun.

What makes a good competitive game? A good competitive game should not only be visually clear, but also interesting. That is, like any sport, a competitive game should (at its most basic level) be understandable. Your novice player should be able to see a basic play and understand the basics of what happened (good or bad for your team? Displays of skill?) while still leaving complexity in the game for more knowledgeable players to geek out about (and more importantly, that complexity needs to be in the game for competitive play to be notably distinct from casual play.

What makes a good collaborative game? My first, cop-out answer wants to be ‘good players’. Good roleplayers will always make for a good rpg, regardless of the system. But I would argue that a system which is good for any roleplayers is a game which encourages roleplaying in the following ways:

1. Eschewing rules. You shouldn’t be rolling dice all of the time. Roll dice if it’s important HOW something is accomplished.  If it’s important THAT it’s accomplished (it needs to happen), but there are no consequences or there’s only one way to do it, it’s lazy to leave it up to a die roll, just leave it up to narration. If it’s entirely unimportant, just let it happen. Essentially, as a GM, you should be looking for a system which urges the philosophy ‘say yes or roll dice’.
2. Narrative encouraged. Spirit of the Century, Dogs in the Vineyard, Apocalypse World, and many other systems all ask for the player to give input about the world and about narrative, and this makes them strong systems. However, related to my above comment, good players will do this anyway, so I see these systems more as training wheels than a necessity; they teach people to narrate, but they do not make narration possible.

Of course, if you’re of a more simulationist/gaming bent, you’re looking for different things entirely, and you’re probably going to end up with a homebrew system. I recommend Spycraft 2.0 or GURPS, for starters, as they have very in depth rules.