What is the Point of System?

This is the first in an open ended series of posts titled "What is the Point..." where I dissect some aspect of gaming and discuss my feelings on the subject, the good, the bad, etc.  While I have two others planned (Experience Points and Plot), I am open to suggestions for things that you want to hear my thoughts on.  Please leave a comment if you have something you'd like to hear about.

So, today's topic is "What is the point of System?" Why should (or shouldn't) we use systems within games?  A game system can be an enabling tool, a ridiculous straitjacket for descriptive action, a way of splitting narrative focus, a tool for in-character bullying or all of the above at the same time.  I'm not necessarily going to advocate one system over another, but talk about some of the general potentials that systems offer and potentially the circumstances under which a system is not a good idea.

But to take a step back, what is a system within the context of game?  Does a system have to include miniatures and dice, charts and tables, THAC0 and hit locations in order to be considered a game system?  For the purposes of this discussion, a system must have at least one of the following characteristics, but with the exception of very experimental games, usually contains all three.
1. A system must be a way in which narrative disputes are resolved.  When two participants describe different possible outcomes of the narrative, the system has a way of indicating which outcome happens.
2. A system must include some way of describing the world and its participants.
3. A system must define the boundaries of the game space and dictates what types of player action affect the direction of the narrative.

Each of these characteristics opens up multiple possibilities for greater understanding of system in game and different answers to the question what is the point.  At the same time, I think it's important to point out that by my definition, there is no such thing as a systemless game.  Even a game where characters have no sheets and all disputes are discussed between participants has a system, just one based on description and consensus.

So what are some of the opportunities that system opens up?  What should one consider when deciding how system incorporates into a given game?

I. Deferred Responsibility and Neutrality
A system that includes a method of narrative resolution has a significant advantage.  Basically, it allows the energy of the Storyteller(and players) to be directed towards creative work rather than devoted to narrative arbitration.  In a system where resolution is primarily negotiated between participants, including systems where there is significant negotiable variation(equipment bonuses, situational penalties, variable difficulty numbers) a lot of time and energy goes into the negotiation of resolution.  This time and energy could theoretically be devoted elsewhere if the system deferred that responsibility.
This is also useful because it allows a greater degree of neutrality.  Nobody can achieve true neutrality.  Bias exists in every single decision that is made.  A system allows the Storyteller to defer decisions to a theoretically neutral arbitration system that does not have the same biases(it may have its own).  Additionally, I have found as a Storyteller that systems and randomness can make it easier to say no(which can be a very important skill in running games).  If the system does not indicate that something is possible, it is a lot easier to justify saying no.[another future post will determine whether saying no is a good or bad thing in game].
Overall, by taking decision making out of the hands of storytellers, more of the storyteller's energy can be devoted to creative aspects of running the game rather than arbitrative aspects.

II. Matching System to Narrative
One important aspect to system is that different systems match different underlying worldviews.  Choosing a system that matches the general tenor of the world of game or at least hits on themes present within game is important.  This can be very well shown by example.  I was once playing in a game that used the same system for two very different linked worlds.  One was set in the universe of Buffy the Vampire Slayer, one in the world of Lovecraft's Chthulu mythos.  The system was identical for character creation and resolution, but one simple decision differentiated the narrative feel of the universes.  When the results of a roll indicated a tie, final results varied by universe.  In the Buffyverse, where plucky heroes defy insurmountable odds, the heroes won on ties.  In the Chthuluverse, where the very fabric of the universe conspired against the protagonists, the heroes lost on ties.  This minor variation defined the difference between the two universes.
A system choice can be even more defining.  In the dramasystem used in Hillfolk, the system defines the emotional and practical stakes of every interaction and defines a system of resolution based on those stakes.  Because the story is essentially about the differing agendas of people in society together, the system is about the ways in which those agendas interact.  While there is no randomness to the system, the inherent political economy of it emerges in the interaction.
Finally, a system choice can go beyond just the ways in which conflicts are resolved.  Some systems are very heavily weighted in the description of the universe to certain types of player interaction.  These can be leveraged well in narrative creation if engaged with.  So, for instance, in the classic dungeons and dragons system, there are different character roles that engage with different narrative challenges.  In story planning, this can be used to naturally shift focus between characters as different sorts of challenges come up.  Alternatively, systems may encourage different methods of resolution that shift play style.  The game Grimm, where the characters are children stuck in a dark fairytale world, makes challenges significantly easier if characters work together and thus makes challenges about how the characters use their abilities synergistically rather than individually.

III. Randomness as Epiphanic
So, here is where I make a case for randomness.  Many contemporary story games do away with random elements, sometimes entirely. While I think that can be useful for telling certain types of stories, I think that randomness has a place as an epiphanic element of storytelling.  Randomness has the chance of pulling you out of your comfort zone, forcing you to sometimes find narrative justifications for the unlikely.  This can drastically and memorably alter the shape of the story, and that, ultimately, is why gaming is different from writing, because there are parts of the story that pull in different directions.  If the dice tell me that David scores a critical hit on Goliath, then that story ends differently from the pre-determined series of events.  Epic success and Epic failure push the participant out of the ordained narrative path is the way an epiphany takes you out of your humdrum life.

Overall, system should be a choice.  There are a lot of parts and when planning a game, you should take into account the things that your system does to the story.  But you should also remember that there is no such thing as a game without a system.  Even 'systemless' games have some form of arbitration.  By making system an active choice, Storytellers have an opportunity to make system work for them, not against them.