The Forge The Internet Home for Independent Role-Playing Games
About the Forge | Forums | Articles | Reviews

System Does Matter

by Ron Edwards <>
Copyright Adept Press

I have heard a certain notion about role-playing games repeated for almost 20 years. Here it is: "It doesn't really matter what system is used. A game is only as good as the people who play it, and any system can work given the right GM and players." My point? I flatly, entirely disagree.

"Whoa," you might say, "my GM Herbie can run anything. The game can suck, but he can toss out what he doesn't like and then it rocks." OK, fine. Herbie is talented. However, imagine how good he'd be if he didn't have to spend all that time culling the mechanics. (Recall here I'm talking about system, not source or story content material.) I'm suggesting a system is better insofar as, among other things, it doesn't waste Herbie's time.

"Oh, okay," one might then say. "But it's still just a matter of opinion what games are good. No one can say for sure which RPG is better than another, that's just a matter of taste." Again, I flatly, entirely disagree.

Some definitions would be good. First, I'm talking about traditional roleplaying games, in which the GM is a human, and the players are physically present with one another during play. Second, by "system" I mean a method to resolve what happens during play. It has to "work" in two ways: in terms of real people playing the game and of the characters experiencing fictional events.

System Design: Part One

(The following is based on the ideas presented at, but I'm expanding their application pretty widely.) Three player aims or outlooks have been suggested, in that a given player approaches a role-playing situation pretty much from one of them, with some, but not much, crossover possible.

Here I suggest that RPG system design cannot meet all three outlooks at once. For example, how long does it take to resolve a game action in real time? The simulationist accepts delay as long as it enhances accuracy; the narrativist hates delay; the gamist only accepts delay or complex methods if they can be exploited. Or, what constitutes success? The narrativist demands a resolution be dramatic, but the gamist wants to know who came out better off than the next guy. Or, how should player-character effectiveness be "balanced"? The narrativist doesn't care, the simulationist wants it to reflect the game-world's social system, and the gamist simply demands a fair playing field.

One of the biggest problems I observe in RPG systems is that they often try to satisfy all three outlooks at once. The result, sadly, is a guarantee that almost any player will be irritated by some aspect of the system during play. GMs' time is then devoted, as in the Herbie example, to throwing out the aspects that don't accord for a particular group. A "good" GM becomes defined as someone who can do this well - but why not eliminate this laborious step and permit a (for example) Gamist GM to use a Gamist game, getting straight to the point? I suggest that building the system specifically to accord with one of these outlooks is the first priority of RPG design.

(Note, therefore, that I might praise a given system because it matches beautifully with one of these outlooks - even if I don't share that outlook and might hate playing that game. This is an important point, because I now have some criteria to judge, instead of just yapping about "what I like.")

System Design: Part Two

Now that a system has an outlook or aim to use as a yardstick, it's time to dissect that resolution method in some detail. Here I follow Jonathan Tweet's suggestion (found in the rulebook of the excellent RPG Everway) that there are three modes of resolution in role-playing.

A given system may certainly mix and match these methods, and in fact Everway actually permits the GM to concoct his or her own smooth blend. Amber, for example, modifies its Karma system with Drama; Extreme Vengeance modifies its Drama method with Fortune; and Sorcerer modifies its Fortune method with Drama. Some systems use different methods for different sets of activities; e.g. AD&D uses Karma for magic and Fortune for combat.

Let's consider Fortune methods as the example because that's what most of us are used to. So the question becomes, given that a system is (e.g.) mostly Fortune-based, how well does it actually work during play? I suggest two things to check carefully (these terms are stolen from ecology, of all things).

I certainly can't dictate how much is too little or too much - but I do claim that if they are not appropriate for the player outlook of the game (Gamist, Narrativist, Simulationist), players will complain, rightly, that the system "bogs down" (Narrativist), is "unfair" (Gamist), or isn't "realistic" or "accurate" (Simulationist). A good system's resolution should get the job done in appropriate amount of real time. Which job, and how long is appropriate, depend on the outlook. A new RPG system has no excuse simply to rely on the old paradigm of (1) roll initiative, (2) roll to hit, (3) roll defense, (4) roll damage, (5) check for stunning, etc, etc. This is a leftover from wargaming and is strictly Simulationist + Gamist. The RPG for you might be very, very different. In Zero, for instance, the order of actions, the success of each action, the degree of success for each action (including damage), and every other aspect of resolution are determined by ONE roll per player and ONE roll by the GM, in all cases, even in large-group combat. This game's system is truly an eye-opener for those used to the older methods.

(Again: it so happens that I'm a hard-core Narrativist who enjoys Karma-based systems most, with a little Fortune mixed in. But according to the principles above, I can now judge a system according to its priorities, rather than just going by "what I like.")

Another interesting question about resolution methods is, what is actually being resolved in terms of numerical game mechanics? Consider three things: the actual event ("do I hit?"), the energy it takes to do it ("deduct 4 Endurance"), and the reward ("You did 18 damage, that's 18 EP's, mark'em down"). Food for thought: maybe an RPG needs only one of these, two at most, and can let the third just vanish - and it doesn't matter which. I'm still thinking about this issue, though; at the moment it's just a notion, not a conclusion.

In Conclusion

To sum up, I suggest a good system is one which knows its outlook and doesn't waste any mechanics on the other two outlooks. Its resolution method(s) are appropriate for the outlook: they have search and handling time that works for that outlook, in terms of both what the players have to do and what happens to the characters. (One might even suggest that the method be thematically suitable as well, as in marbles for Asylum and playing cards for Castle Falkenstein; I like this idea too, but it's not absolutely necessary.)

Perhaps the ongoing debate about "system-light" vs. "system-heavy" is a waste of time. A system is not automatically good if it is more or less complex than another. The degree of acceptable complexity comes from the game's outlook, and should be judged in that context only. A Simulationist, Fortune-based game almost has to be complex, but a Narrativist, Karma-based game is most satisfying with a simpler system.

Please consider comparing a few systems yourself before reacting too strongly to this essay. I do respect your opinion, but it's fair to consider how many role-playing games you have actually, truly played. That is, real stories and sessions with characters the players created and cared about, not demos at a tournament or running a quick combat. I suspect that those of us who've played more than five or ten RPGs in a committed fashion will agree that "system doesn't matter" is a myth.