Tuesday, August 15, 2006

Conflict: Resolve or Evolve

I have a problem with the way most RPGs handle conflict. Here's the short form: GMs often abuse dice mechanics in a mistaken effort to artificially increase tension. I think this is a holdover from the RPG industry's wargame heritage. If you want to improve the tension in a game, make the game more difficult: reduce the player's odds of success, increase the complexity and make the system harder to master, or complicate resource management.

While these techniques may be useful in a strictly tactical game, they often sabotage story. Conflict drives every good story. Good conflicts follow the laws of cause and effect. They arise naturally from the story. Their resolution should feed back into the story, influencing the narrative flow. Conflict must remain central to the story.

Let's say my character, Bob, tries to stab an NPC named Fred. At the topmost view, Bob's success or failure is unimportant. What matters is a) Bob's motivation and b) the ripple effects from Bob's actions. From this viewpoint, spending thirty minutes rolling each thrust and parry is a waste of time and energy.

Let's look at an example. Bob needs to sneak into the infamous Castle of Four Winds. He crosses the moat unseen and starts to scale the tall, slick walls. The GM wants the moment to feel dramatic, so he decides that Bob must make ten successful climb rolls to reach the top.

My reaction: "ARGGGGG! Thump! Thump! Thump! (the sound of my head slamming repeatedly into the table)"

OK, what's wrong with this situation. First, the tension comes from game mechanics, not from the story. As a player, I'm focused on the dice and stats. Sure, the GM may try to play up the drama by describing the feel of the slick stones under Bob's fingers, or the way the wind howls around our hero. That's great--for a small number of rolls, but how can anyone maintain the tension over all 10 rolls?

The rules of good storytelling insist that the tension increases as the scene progresses. The GM's description should get more dramatic with each roll. This is very hard; most of us fail most of the time. For example, GM may make the initial description too dramatic, leaving no room to grow; or he may make a weak initial description, one that has no dramatic weight. Later descriptions can either become repetitive and boring, or they rapidly escalate to ridiculous proportions.

The worst case comes from what I call "Traditional D&D Combat." Here, you have dozens upon dozens of rolls to resolve a single conflict. Say Bob takes a good hit. The GM wants to build a visceral sense of tension, so he describes the ork's axe slicing through the meat in Bob's shoulder. However, Bob only loses 1/10th of his HP. Soon he's running around with a half-dozen graphically-described-but-mechanically-minor injuries. The game starts to resemble Monty Python's "Black Knight" sketch.

This brings up a related point. While the actual description is important, it's the player's sense of risk that really matters. All too often there is a disconnect between the GM's description, and the game mechanics. The GM can describe an edge-of-your-seat, cold-sweat, higher-brain-functions-have-stopped danger--but, if the player knows he has a 98% chance of success, mere words won't carry much weight.

To sum up, we want our GMs to create an impromptu, reasonable sounding, realistic yet still dramatic set of descriptions that closely match the actual risk faced by our character over a series of random die rolls. Good luck.

But, the climbing example has an even bigger problem. Good conflict needs to be the cause of interesting effects. Here, I mean "interesting" in the sense of "May you live in interesting times." As a quick rule of thumb, the player's situation should become more difficult and more complicated. Ideally this happens both when they fail and when they succeed, but I would settle for interesting failures. Success, after all, tends to drive the characters deeper into the story.

Let's look at the climbing example again. What happens if Bob succeeds? He gets to the top of the wall. Ho, hum. Still, he's now inside the Castle of Four Winds. Any number of bad things might happen to him there.

Let's look at the other side, what happens if he fails?

One possibility, Bob falls to his death. While that may instill a real sense of risk, story-wise it is a pitiful way to resolve conflict. Dying generally ends the story (at least as far as Bob's concerned). It doesn't heighten dramatic tension; it brings everything to a sudden halt. Don't get me wrong, death can create good story (e.g. Boromir's death in Fellowship of the Ring). In RPGs, these tend to come from a collaborated effort between the player and the GM. It's something you work for over several gaming sessions. A character's death should be glorious. Random deaths are, by definition, pointless.

So, maybe the fall merely injures Bob. That seems better, but it's still not great. Massive injuries either incapacitate the character, taking him out of the story (similar to death, only temporary), or the injury has no real effect on play. Bob quaffs a healing potion, or Raymond, the party priest, says a few magical words and suddenly we're off slaughtering orks again. In traditional RPGs, injuries rarely add to the drama in any significant way.

Also, GMs (myself included) tend to put the players in life-threatening situations and then flinch. Back to the climbing example, even if Bob fails, the average GM will give him second (and third, and fourth...) chances. "Ok, roll your dexterity to see whether you catch yourself before you plummet to your death...Hmm, roll again to see if you can grab a protruding branch on the way down..." This gets back to the perceived risk issue--if you know the GM won't kill you, much of the risk evaporates.

This leads to two rules for creating good conflict:

1: Conflict must cause interesting complications



Before introducing an element of conflict, think about how things might go wrong. Remember, this can include complications even when the character otherwise succeeds. If the player kills an NPC in combat, does that NPC have a brother (or sister, spouse, offspring or parent) that will come looking for revenge? If the character successfully scales the wall--what dark secrets lurk on the other side.

For failures, it helps to look at conflict as a complete task, not as the component actions. I roll to see whether Bob reaches the top of the wall undetected. Failure does not necessarily mean Bob failed at the act of climbing. Maybe something happens that prevents him from reaching the top. For example, a guard might spot him when he's only half-way up.

2: Roll the dice once; resolve or evolve



Roll the dice. When they land, either accept the result and narrate the resolution (resolve), or add a complication that ratchets up the conflict (evolve). This evolution may lead to additional die rolls. Having the guard spot Bob is a perfect demonstration of conflict evolution. The conflict goes from "I need to get up this wall undetected" to "I'm stuck halfway up this wall, and the guards have bows."

Good conflict evolution should force the character to make hard decisions. In the climbing example, the character must decide whether to keep climbing (probably fighting his way up to the top), or to abandon his attempt and try to flee.

Conflict evolution can also be subtile. For example, instead of having a guard spot the character, maybe the character is nearly to the top of the wall, when a lovesick scullery maid decides to stop just above him. She leans against the wall and stares out into the night. She hasn't seen our hero yet, but it would be hard for him to sneak past her. Does he wait? Clinging to the side of the wall is exhausting and dangerous. The longer he stays, the greater the chance of someone spotting him. Does he try to sneak past? That also risks discovery. Does he try to kill her quickly and quietly--thus removing a potential witnesses? Is her lover coming to join her, thus further complicating the scene?

Of course, creating and evolving good conflicts places a major burden on the GM, and most GMs are overburdened as it is. To partially alleviate this, I propose pushing much of the responsibility onto the players.

Here's the general flow:


  • The GM can always add an external complication to the conflict. This represents something that complicates the situation, but is not a failure on the player's part (e.g. the scullery maid above). It's OK to have a group of guards walk past on patrol. Having the guards automatically spot a hidden character is uncool.


  • If the player fails, he may add a complication to temporarily avoid his fate. The complication must have a serious impact on the rest of the scene, and the GM must approve. Possibilities include an injury that affects performance (e.g. hand smashed and useless), equipment failure (weapon broken or possibly just disarmed), or the introduction of outside elements (the guards spot Bob halfway up the wall). If the player does not add a complication, the GM is free to narrate any result. "Oh, goody. You fall to your death."


  • Turn-about is fair play. The GM can add complications to NPCs to mitigate their failures. However, the GM should only use this option on major, named NPCs. Cannon fodder and faceless thugs can just die anonymously.



This would probably work best using a narrative mechanic where the winner of a roll gets to describe the results (e.g. The Pool). But, I suspect this system could be strapped on to any game engine. It may rattle a bit at highway speeds, however.

No comments:

Adventures in the Wondrous City of Cathlarian

 So, sometime before Christmas, my son asked me to run a D&D game for him. I've been mostly running narrative focused games lately. ...