decision making, conflict resolution, and the metagame

[punkbohemian]

I've been thinking about resolution mechanics on more of a metagame level lately. To give you a frame of reference for this discussion, think of games like d20, Tribe 8, and GURPs along with games like Mouse Guard and Burning Wheel. I would like to propose three different approaches for comparison.

Classic Traditional

All participants roll initiative. Actions are declared and resolved simultaneously in the order of "fastest" to "slowest".

Traditional Revised

All participants roll initiative. Actions are declared in the order from "slowest" to "fastest", but are resolved in the order of "fastest" to "slowest".

Simultaneous Approach (e.g. Mouse Guard)

No initiative is rolled. Instead, all participants secretely decide actions and all are revealed and resolved simultaneously.

Recently reading Mouse Guard is what inspired me to think about the different psychological dimensions of these three models. That is, an aspect of resolution is the decision making process where the player chooses a particular action. This decision making process is independent of the game itself. So, if you compare games like D&D and Tribe 8, it seems like the games are vastly different, but the process of player decision making (when mechanical resolution is required) is almost identical. They both follow the Classic Traditional approach. This works well enough as the strategic elements of combat in both games pretty much boil down to kill-the-other-guy-before-he-kills-you.

I have never seen the Traditional Revised approach in an actual game. I merely present it to allow a comparison. It does add an additional element to decision making that is not provided by the Classic Traditional approach. That is, if you know what your opponent(s) is/are doing that round, you can implement that knowledge into your own decision making process to possibly counter or minimize the effects of their actions while maximizing the effects of your own. This probably does very little good for games like D&D, where most actions are about dishing out as much damage in the shortest time possible, but it's a different story when you consider it alongside the Simultaneous approach.

If you are not familiar with Mouse Guard, the mechanic basically boils down to a game of Rock-Paper-Scissors, though I'm vastly oversimplifying for the sake of the discussion at hand. Any action is strong against some actions and weak against others. Generally, you want to be the one who picks the strong action and your opponent to be the one who picks the weak one. Also, a game like Mouse Guard has an additional element (that I find to be quite clever), which is that your success (or degrees thereof) is not just dependent upon defeating your opponent, but also minimizing the damage to your own party. It adds a more complicated dimension to the cost-benefit analysis of the decision making process. These two elements are key to being able to take advantage of the simultaneous revelation and resolution of actions. Such an approach would be of little benefit to D&D because most actions can rarely cancel each other out (not to mention, it would make a nightmare out of using a battle map).

The thing about the Simultaneous approach is that it's primarily driven by luck (as is the case with RPS). Sure, you might occasionally see a situation where it would seem quite likely that your opponent would make a certain choice, but for the most part it's a pure gamble. The decision making process becomes much less strategic the less one can form an educated guess as to their opponents' intended actions. However, if you were use the Traditional Revised approach with a mechanic where some actions cancel out others and is canceled out by others, it would only exacerbate the problem. That is, if you know your opponents' actions before you choose your own, then you will always know the right move, making the decision making process too simple. I am not really seeing an in between.

One compromise that I had considered was using the Traditional Revised approach to determine the action order, but every subsequent round the person that goes first rotates up (or down) the list. So, ABCD in round one, DABC in round two, CDAB in round three, etc. However, the only difference there is that you spend your time flip flopping between two extremes: knowing and not knowing. While it balances out the total resolution, every individual opportunity for decision making is still either totally obvious or a crap shoot.

This does not get me any closer to establishing a gradient. Think of a game like chess. Before your opponent makes his/her first move, you have no idea what s/he is intending. But right before s/he makes the last move, you can pretty much see what's about to happen. In the middle is the actual "game" part of chess. With these resolutions mentioned above, you're either looking at a freshly set up chess board, or a chess game one move away from victory. I'm not seeing a approach to decision making that falls in the between.

Anyway, I thought I might throw this dilemma out to the community to see if anyone else had some thoughts.

[Ron Edwards]

Hi there,

Your Traditional Revised showed up just as you describe in Sun & Storm (1991) and Street Fighter (1994). It's logical but the few bits of reported play I've encountered claim that it runs into what I call handling-time problems, requiring just a bit too much attention and effort per unit of fictional action.

The fully Simultaneous technique has been seen in simple form for a while, usually with one or both sides rolling once for an entire clash or even whole encounter. You can go all the way back to Tunnels & Trolls in the mid-70s to see it as single-roll side-against-side for combat rounds. It fell out of fashion but showed up dramatically in Over the Edge (1992, I think).

For more nuanced, multi-participant versions of the Simultaneous technique, the first game I know about that busted it out in functional form was Zero (1997), on which I based order/action for the final version of my game Sorcerer. Hero Wars (2000) totally did it, and Orkworld the same year. These are all iterative versions like Over the Edge, where a single roll vs. roll resolves some or all of a conflict depending on what the roll did, and then keeps going into new rounds if necessary. And since then, a lot of the games built here have gone with interesting tricks off the simultaneous platform, especially since The Pool (2002) kind of broke open our thinking about it. I'm not sure even how to summarize it after that, except to say that whole little subfamilies have sprung up mainly based on how much of a conflict gets resolved in a single exchange, whether one side or both roll (or draw cards or whatever), and how speaking (narration) is organized relative to the outcomes.

I apologize for the terse and dry response, but I'm hitting a quick time window. I'll be more chatty next time.

Best, Ron

[Ron Edwards]

Whoops! One more terse post.

In this particular forum (Game Development), you're required to provide a link to a working design document for reference, in order to start a thread. You can read about that in the sticky post.

If you'd prefer simply to discuss the techniques instead, then the place for that is the Actual Play forum, in which case you're required to describe some time or times of real-actual-play experiences as part of the post.

Let me know which one will work better for you with this topic. Please don't post a new thread in either forum or anything else that requires effort - whichever you prefer, I can move this very thread there or let it stay here.

Best, Ron