To Link Test or Not in Burning Wheel Gold

Balancing Decision Fatigue, Time Budgets, and Fictional Pacing

Note: This post has content disclaimers.

Over on the Burning Wheel Discord a conversation came up about an example of a task and intent in Burning Wheel Codex 🔍:

To capture the sentry: “I sneak up behind him, muzzle him and choke off his windpipe.” Test Stealthy plus Brawling (or Martial Arts) plus appropriate Wises.

As written, this is a single test; the task and intent are clear and achievable. But I want to explore this further

Discussing a Standard Test

Let’s work through the Obstacle (Ob 🔍).

The “sneak up and muzzle him” is more complicated than “sneak up;” there’s space for more things to go wrong. As the Game Master (GM 🔍) I would set the Ob higher than if the intent were simple to sneak up on the guard.

With the intent of “sneak up and muzzle him,” I would frame this as a Standard Test. Looking at Burning Wheel Gold 🔍 (page 15), I’d assign an Ob of 4 or 5, though I’m leaning towards an Ob5. An Ob4 is a risky act, an Ob5 is an act that requires expertise, and an Ob6 is an act that requires heroic effort..

The consequence of failure? The guard blurts out a warning before you incapacitate them.

Comparing to a Linked Test

Let’s compare this to what could instead be a linked test. Each test must have it’s own task and intent. The character wants to “sneak up on the guard to get the drop on them” and then they want to “muzzle the guard.“ Is this appropriate to state two different tasks and intents, when one is dependent on the previous? For this example, I’ll assume it’s appropriate and proceed as such.

What are the failure conditions of each of those tests? For sneaking up, I want to say that the guard would notice the character and forcefully ask “Who goes there?”

For the “muzzle the guard”, the failure would depend on the previous test. If the guard noticed the character, then attempting to muzzle them would be quite different than if the character succeeded.

For the “sneaking up,” I would have the character make a stealthy test versus the guard’s observation; both the guard and the character would roll. Since we’re dealing with a linked test, the result would impact the “muzzle the guard” test.

What would the “muzzle the guard” test be? Is it a versus test? I think so. It would be brawling versus brawling.

Reflecting on Both Approaches

As I was writing the linked test, I found that I needed to think through more decision points. This linked test, in essence, slowed down the camera to follow more closely the narrative beats of sneaking up and muzzling the guard.

It also necessitates more decisions: you’ll need at least two consequences, and the second consequence is somewhat contingent on the first. You also need to determine two different obstacles.

Whereas the “sneak up and muzzle him” test had less decision points. I needed to think of a reasonable obstacle. And the camera followed at a different pace. We’re not watching the character take a few silent steps, pause, then lunge at the guard. Instead we show the character right behind the guard, slipping their hand over the guard’s mouth.

And when we compare the two approaches, the linked test would likely have two easier Ob to overcome compared to the more challenging Ob for the “sneak up and muzzle him” task and intent.

Which Approach to Use?

It depends.

And in this way, Burning Wheel Gold breaks with traditional games. In Dungeons and Dragons (D&D 🔍) you’d sneak up and make an attack roll. But in Burning Wheel Gold, you’re focusing on resolving a stated task and intent. And you have each characters' Beliefs, Instincts, and Traits (BITs 🔍) to help inform you.

As a GM, I want to minimize the chances of decision fatigue, so I’m going to favor the “sneak up and muzzle him” over the linked test. But if the players want to help me navigate that linked test, I’m game.

However, as a player, I might propose that I want to attempt to sneak up on the guard so my character can get the drop on them. I’m assuming that will both be an easier test, but also one from which the character could pivot.

Or, I might look and say “I don’t want a test where my character alerts the guards, then I need to scuttle my attempted muzzling of them.” Instead, I want to hazard my character for something more. To put them in a potential state of urgency.

This circles around something that Jason Cordova and Judd Karlman talked about in The Sixth Ring: Loss episode of the Trophy Podcast. One of their points is on playing to lose versus playing to win.

In some moments, players play to avoid losing. They will take the cautious approach; inching their way towards a larger goal. Putting one cautious foot forward to test the situation. That, in a way is analogue to the above example of the linked test. This approach also requires more table time. At a minimum two rolls, but also a likely reassessment between the two tests.

In other moments, players play to win. They grab for something, lunging towards their goal. This is analogue to the “sneak up and muzzle him” task and intent. This test likely consumes less table time.

Consider the Competing System Rewards

It’s hard to imagine that either these two approaches directly tie into a character’s BITs, but the opposition of the guard is likely a barrier towards achieving their goals. Moving through the opposition and towards goals helps increase opportunities for Artha awards.

If we consider that each session has a finite time budget, you can look at the decision regarding the two approaches as either favoring more tests or awarding more Artha. Both are valid, as evidenced by the Fight!, Duel of Wits, and Range and Cover subsystems. Those subsystems slow down the fiction even faster, but are great opportunities to wrack up numerous tests.

Postscript

Compared to their low-level D&D counterparts, characters in Burning Wheel Gold are rather durable. With fragile characters, it makes sense to play to avoid losing. The stakes are high that you’ll lose your character.

But in Burning Wheel Gold I’ve found a joy in the lunge of a character reaching for something. Yes, the character’s life can get complicated, but it’s unlikely that they’ll be removed from play. Instead, I end up having challenges I need to resolve (and write them down as Beliefs, which in turn yields Artha). Risk creates the grist for the virtuous cycle that is the Artha rewards system of Burning Wheel Gold.

update

Paulo De Tiège posted on Reddit the following addition to this post:

An interesting post as usual, thank you for linking this. The post focused on the mechanical aspect of Artha generation vs test opportunity, but I'd also focus on the narrative aspect of it all.

If the narrative is focused on having the players talk to the duke, I'd use a single test to make this a small footnote in the story of them getting to talk to the duke; however, if the narrative is a heist to steal the Duke's jewels, I'd absolutely use more linked tests to put the microscope on the individual steps and increase the dramatic potential. ("Oh no! The guard raises the alarm! Mad scramble for the exit after having yanked the jewels back out of their hands!")

This does, of course, tie in with BITs, as you suggested.