On my blog, I like to analyze game rules and determine whether they are effective and desirable, and I like to analyze games and suggest new game rules that would improve the game, and I like to design game rules for new games. However, thinking of role-playing games, it is important to note that there are different goals which one might be trying to achieve with a game or a game rule, and sometimes a game rule might be well suited towards achieving one goal, but badly suited for another. So I thought I'd describe some of the classic role-playing game goals.
Probably the most classic goal is the goal of realism. When a rule is being judged according to the goal of realism, it is judged according to whether the effect of the rule is as similar to real life as possible. Before D&D was invented, its predecessors in the hobby game field were wargames, which descend from the concept of actual wargames used to study actual military tactics. The traditional goal of old wargames was realism. This is to say that such games were unconcerned with playability and fun. These are required considerations in the design of any game. But with the focus on realism, these are treated as constraints, with the desire being to maximize realism while maintaining the desired level of playability. I always enjoy reading the Designer Notes of old Avalon Hill wargames, with the detailed discussions of how and why the game rules were chosen in such a way as to maximize historical accuracy.
A second, somewhat different goal is that of genre simulation. When a rule is being judged according to genre simulation, it is judged according to whether the flavor and situations resulting from the rule match the source material of the game as closely as possible. The source material is typically works of fiction, and the difference between realism and genre simulation is the difference between how the world works in fiction, and how it works in real life.
Going back to the original Dungeons & Dragons, it was clearly designed with some core game rules meant to simulate the fantasy genre. The idea of hit points that increase massively as characters gain experience is clearly not realistic, but meant to simulate the idea that the fantasy characters are larger than life, and that great mythological heroes are not going to be killed by a single stray arrow. And, of course, aspects of the game like the magic system couldn't possibly be realistic because the setting of the game is not based on real life, but on fantasy.
When I think back to old D&D, however, my memory of the rules discussions in the old game magazines was that they tended to focus on realism as a goal, rather than genre simulation. It seems to me that in 1st edition D&D, and a lot of other games, genre simulation was used more as an inspiration of the rules rather than a goal of rules. That is, the fantasy genre was used as inspiration to create the fundamental game rules, but once those were created, the original source material was no longer used to guide the development of new rules. Rather, the goal was what I might call "self simulation". The game became its own genre, and the focus of new rules and supplements was to logically extend the game as defined by the basic game rules.
An example of self simulation would be the third edition D&D idea of putting law vs. chaos as magical concepts on a par with good vs. evil, so that you would have lawful weapons that damaged chaotic creatures and so on. The goal here certainly isn't realism, nor is it based on genre simulation (there are certainly fantasy works that involve the forces of law vs. chaos, but not, I think, in the style of D&D, where you have lawful good and lawful evil in some sense banded together against chaotic good and chaotic evil). Rather, this is driven by looking at the alignment system and trying to extend it to its full logical extent. If you already defined that there are 9 equally valid alignments, and they "great wheel" cosmology which implies again that all of the alignment axes are equally important, then it creates a pleasing sense of logical completeness to put law vs. chaos on equal terms with good vs. evil in the game rules.
Going back to early D&D rules discussions, however, and the frequent focus on realism. D&D didn't seem to be founded on the basis of realism, but it did seem to be surrounded in an atmosphere that considered realism a highly appropriate goal. The fact is that the D&D rules were the most realistic RPG rules in existence, since they were the only RPG rules in existence. So it is hard to meaningfully discuss the intended goal of the core game rules compared to other games. However, certainly could be said that many of the individual rules could be easily recognized as realistic in intent. For instance, discussions of keeping track of your exact monetary status, of tracking the passage of time so that you know when your six-hour supply lantern oil is used up, and so on.
What I always found strange was the way that game rule concepts would focus on realism while simultaneously incorporating without question certain highly unrealistic game mechanics, creating very strange results. For instance, the idea that when you are hit by a fireball, it would be realistic to check to see whether all of the wooden items on your person catch on fire. This ignores the complete unreality of the character himself surviving the fireball without being totally disabled by serious burns. If the core game mechanic being modified here was trying to be realistic, it would make sense to add extra rules to make it more realistic. But when the core game mechanic is not trying to be realistic, I think this kind of rule just sacrificed playability for no real gain.
(to be continued when I return from vacation).