Mira's Interview with Scott: A Conversation About In-game Mortality, Consent, and the Stories We Remember

When Terrible Gamers’ own Scott Mills and I started talking about character death, I made a mistake. I took his enthusiasm for what lethal games offer as evidence that he preferred them. He corrected me. One of his longest-running and most enjoyable campaigns was Vampire: The Masquerade, Second Edition, where his character faced no real risk. David Schmidt, Terrible Gamers’ own Court GM, ran it for many years. As a player, Scott enjoys narrative games with protected characters and genuinely lethal games equally.

That matters to the argument. We are not trying to establish which players are tougher, which style of game is better, or which games deserve to survive. We are asking what credible mortality makes possible in tabletop roleplaying games, and what happens when protecting a character’s future becomes an expectation rather than a particular choice.

The answer begins before any player character dies.

Before the fight

Scott's description of lethal play is mostly about trying very hard to stay alive. Players become careful and deliberate. They investigate their surroundings, consider what they can do with the things in their packs, negotiate, and look for ways to avoid encounters altogether. Combat becomes a decision with alternatives.

Imagine a party exploring a dangerous tower. If a trap can end a character's life, the marks on the floor and the mechanism beside a door deserve attention. A balcony might be an escape route. A piece of equipment that contributes nothing to a combat build might solve the problem that would otherwise force a fight.

The place remains important after the party leaves its first room. Retreat has to go somewhere. Noise might bring trouble from somewhere else. A passage left unexplored can matter when the obvious exit is blocked. The players have reasons to ask how the tower actually works, because their characters may have to get out of it.

Once they commit to combat, cooperation and positioning carry the weight of bringing particular people home. Covering a retreat, spending a scarce resource, or abandoning a doomed plan can determine whether a character or their companions have a future. With those stakes, a difficult fight presents a real calculus of risk. Some characters might not survive, and the group may have reason to retreat rather than finish it.

Mortality needs decisions that matter. If players have no useful information and no effective alternatives, frequent death may teach resignation instead of ingenuity. The value is in credible danger they can respond to, not in a GM’s ability to kill them. Scott is careful to distinguish this from a GM who seems “out to get” the players. He does not want the game to feel like a contest between players and game master. They are creating a story together.

In his 2013 Dungeon World article, Scott recalls a fighter from an earlier campaign who survived six months and staggered out of a mage's tower as the party's sole survivor. He had to trade the campaign's coveted magic item for medical treatment. He did not last long in the next campaign.^[1] Survival had given him a history, not a permanent exemption.

Getting a character to second, third, and fourth level under those conditions means this particular person made it through uncertainty. Advancement can be satisfying in a protected campaign, too. But surviving something that might actually have ended you is an accomplishment with a different character.

A world that answers back

As a GM, Scott prefers something closer to a simulation: players inhabiting people whose decisions make sense within their circumstances. Credible mortality in some of his campaigns can bring the character's concern for their own life into the player's deliberations.

This need not mean ordinary human vulnerability in every genre. A superhero can reasonably attempt something an inexperienced adventurer should avoid. Extraordinary abilities are part of the fictional circumstances. The gap appears when confidence depends instead on something the character cannot know: that the GM will arrange a rescue because the protagonist's story is unfinished.

The same problem extends beyond death. Suppose a character makes a choice that causes something horrible to happen. Regret can become part of the roleplaying. The character might try to repair the damage, deny responsibility, or discover an unwelcome truth about their loyalties. Removing the consequence because the player did not want it removes the world the character now has to inhabit.

Collaboration can include a division of authority. Players decide what their characters say and attempt; the GM, circumstances, and agreed rules determine how the world responds. Their choices help create the story without giving them a retrospective veto over everything those choices produce. Scott described that distinction in his earlier Dungeon World article, while also acknowledging that other games distribute authority differently.^[1]

I would apply the discipline to the GM as well. If the players derail a cherished plot, the GM should not quietly rearrange reality to recover the desired conclusion. A simulation-like game asks everyone to accept that the process can produce something they would not have chosen to write alone.

Consent to participate, authority to revise

A group can agree that characters may die through decisions, circumstances, and rules without selecting each death as a desirable story development. Another group can reserve character death for outcomes a player individually approves. Those are different mortality arrangements. Both can be chosen knowingly.

Scott's position also distinguishes consent to participate from authority to revise the fiction. Before the campaign, GM and players should be honest about its character, possible content, and their limits. Anyone who finds the proposed game incompatible with what they want can decline it.

That freedom continues during play. If a player or GM discovers that they are no longer comfortable participating, they can leave. An earlier agreement is not an obligation to remain. In Scott's preferred arrangement, however, leaving does not itself confer authority to require a different fictional result. The other participants can respect the departure while continuing the game they agreed to play.

This is not a rejection of ongoing consent. It is a disagreement about what follows when willingness changes.

Contemporary consent guidance can recommend more than withdrawal. Sean K. Reynolds and Shanna Germain's Consent in Gaming explicitly permits leaving an uncomfortable situation, but also recommends revising current content or skipping past it when someone uses an X-Card. It says a topic should be removed if a participant withdraws consent to it.^[3] These recommendations allocate creative authority as well as providing a way to stop participating.

Their scope needs care. The guide distinguishes a topic from its detailed depiction; reducing graphic description can leave an event intact.^[3] It does not prescribe separate approval of every character death. Mechanical concessions, content boundaries, and permission to edit a scene should not all be treated as the same thing.

Nevertheless, Scott's objection has a real target. He does not want discomfort-based vetoes at his table, and he objects to their presentation as expected practice. A procedure can be formally optional while being promoted as what considerate participants ought to use. Questioning that expectation is different from denying other groups the freedom to adopt it.

A group may deliberately give its members authority to revise disturbing content. Scott grants the legitimacy of that choice without wanting to play under it himself. Its possible value elsewhere does not answer his objection to making it the presumed arrangement.

Nor does his model excuse misleading the participants. Offering one kind of campaign and delivering another violates the initial understanding. The existence of an exit does not make that breach acceptable.

Letting the story disturb us

Scott compared the situation to choosing a novel or film. A reader or viewer can learn about its themes beforehand, decide whether to engage, and stop if the experience becomes unwelcome. Rejecting the work does not automatically establish a right to have its disturbing developments rewritten into reassuring ones.

An RPG is different because its participants help create the work. But participation need not distribute every kind of authority equally. Players can meaningfully shape what happens through their characters' actions without approving every result afterward. Whether they also have editing authority is a further agreement, not an inevitable consequence of collaboration.

This matters when a story becomes uncomfortable through the participants' own choices. A character's loyalties might implicate them in an injustice. A seemingly righteous decision might leave innocent people worse off. The group may discover that its preferred heroes are capable of something it would rather not acknowledge. There can be value in staying with that discovery instead of immediately removing it.

If any discomfort is enough to disqualify a development, the story's capacity to unsettle its creators is restricted. We should be able to choose fiction that challenges our expectations rather than promising to satisfy them at every stage.

Scott connected this concern to Adorno’s criticism of art reduced to consolation: apparent escape can help reconcile us to the world as it is.^[11] Adorno was not writing about tabletop gaming. Applied here, his argument raises the question of whether constant reassurance makes our imagined worlds too accommodating to what we already want.

The possibility of transgression matters; transgression is not compulsory. Adorno also writes, “Art can no more be reduced to the general formula of consolation than to its opposite.”^[11] A disturbing event is not automatically worthwhile, and an agreed boundary does not make every remaining artistic choice harmless or conventional.

I also think a group can voluntarily revise a scene for good artistic reasons. It may be gratuitous, poorly handled, or inconsistent with the undertaking. That need not turn the work into a product tailored to each person's preferences. But the possibility of such a decision does not make it an obligation at every table. It differs from a standing individual entitlement to remove whatever becomes unwelcome.

Character death belongs in this discussion because a future cut short can resist the story we expected to tell. Its meaning may arrive afterward: in what the survivors do, in an unfinished undertaking someone else accepts, or in the memory of a small act nobody knew would be a final one. The event need not be selected in advance as the right ending to become part of a meaningful story.

Most of the stories Scott returns to with friends, or tells when recalling his best tabletop experiences, involve a funny character death or some kind of transgression. He is careful to say most, not all.

For example, Scott recalls a Warhammer Fantasy Roleplay campaign run by one of his favorite GMs, Aaron, sometime in the mid-1990s. The campaign was grueling, and fate points had already been spent. The party had been tasked with infiltrating a warehouse in Altdorf that appeared to be the center of a new, possibly Chaos-tainted criminal organization. The mission went wrong quickly, and the group fled. Jerry’s character stayed behind to cover the retreat. Just as he seemed about to escape, he suffered grievous wounds to his chest and arms and was captured.

Scott never forgot what happened after the rest of the group made it back to their home base. Aaron turned to Jerry, whose character had been left behind. As Scott remembers the exchange:

Aaron: “Jerry, does your character have any fate points left?”

Jerry: “No, why?”

Aaron: “Roll a sanity check.”

Jerry: “I passed.”

Aaron: “Jerry, your character is just starting to come to. Early morning light is starting to filter through a window, and you hear someone say, ‘He’s awake.’ That’s when you notice you have no arms.”

Scott recalls that the men who captured Jerry’s character had called a local surgeon. In that campaign, with no fate points left to avert the outcome, saving the character meant amputating his arms. Aaron’s delivery was so matter-of-fact that it was both shocking and funny. Everyone at the table burst into laughter. Jerry then added his own response: his character tried to jump up and kick the surgeon who had treated him.

Jerry kept playing until the guards in the room cut his character down. Even before that, losing both arms had made continuing with this particular thief, who could not cast magic, doubtful. He could have tried to go on, but Jerry expected he would most likely need a new character.

Scott says the memory has stayed with everyone who was at the table and gets shared with strangers at least once every GenCon. Surviving the initial injuries had not protected the character’s future in the campaign. The story belongs to the group because it includes what Jerry made of that loss, not only what the GM decided would happen.

Different rules, different promises

A protected character and a general veto over unpleasant fiction are separate choices. A campaign can guarantee that its protagonists survive while leaving betrayal, failure, and the consequences of their decisions outside individual retrospective control. Enjoying that campaign does not commit someone to a red-card arrangement.

Written rules offer different kinds of uncertainty, too. The 2024 D&D free rules include instant death, death saving throws, magical revival, and replacement characters.^[2] One of Scott's concerns is that combat-capable heroes, combined with expectations of protagonist protection, can encourage players to treat fighting as the default answer. The rules' provision for death does not tell us how often particular groups let permanent deaths stand.

Daggerheart makes outcome authority more explicit. Its ordinary death procedure lets a player choose a fatal last act, avoid death with consequences and a possible scar, or gamble on survival. Accumulated scars can end a character's journey; clearly warned, immediately fatal actions can bypass the ordinary options.^[9] That is a deliberate mechanical arrangement, not evidence that every narrative game eliminates deaths.

In Fate Core, a player can concede before a roll, accept defeat, and retain protection against the worst outcome. Balsera's appended clarification distinguishes that choice from being taken out, which can leave the character's fate, including death, in another person's hands, subject to the social contract.^[4] Narrative focus and guaranteed survival are not interchangeable descriptions.

Other games make vulnerability part of their invitation. Dungeon Crawl Classics expects losses among the multiple zero-level characters each player starts with, and advances the survivors into the first-level party.^[10] Free League presents Alien characters as expendable.^[5] Openly mortal play remains available; Scott’s argument is for recognizing what it offers, not reconstructing something that has vanished.

The existence of lethal games does not answer every objection about what players come to expect. Scott is criticizing the treatment of particular arrangements as defaults. These examples establish differences in design, not how most contemporary groups play.

Around the same table

Scott's Vampire campaign remains central because protection offers something worth wanting. Keeping a character available lets relationships and long undertakings develop. Players can pursue a disastrous loyalty or expose a character's weakness without making every such choice a gamble on whether they will still have that person next session. Political defeat and broken attachments can remain severe consequences.

Mortality can unnecessarily interrupt the development someone most wanted to explore. Replacement characters need a practical route into play; death should not casually leave a player sitting out. Arbitrary losses can erode attachment instead of deepening it. As we discussed earlier, high body counts are not an achievement for a GM.

Still, other consequences cannot reproduce everything death offers. Losing a fortune and losing the person who might someday earn another fortune are different losses. A guarantee of continuation cannot also provide the discovery that an ordinary, unplanned event has ended this particular person's future.

The same friends can choose those different experiences in different campaigns. Bankuei's Same Page Tool emphasizes agreeing on a particular game rather than combining incompatible expectations, including expectations about survival and the GM's relationship to the rules.^[8] It is possible to share a table without making every campaign deliver the same promise.

Before play, we can discuss the intended tone, mortality, permanence, replacement characters, and how authority over the fiction is distributed. We can say whether an unwelcome event stands, whether its depiction can change, or whether a participant has wider editing authority. We can also make clear that either GM or player can stop participating. Recognizing another arrangement's value does not require accepting it for ourselves.

I want a place for arriving at that tower without knowing whose story will continue beyond it. There is suspense in making plans for people we might lose, and accomplishment in bringing them home. Another campaign may give us those same people for a continuing story whose survival we have chosen to protect.

What strikes me about the warehouse story is that it continues to matter after the campaign. Aaron’s delivery, Jerry’s response, and the table’s laughter are still part of a story those friends share at GenCon. An awful outcome within the fiction gave them something they could remember together.

Neither choice needs to become the price of belonging to the hobby. What matters is that we can knowingly choose the experience, respect one another's freedom to leave, and then let the agreed game produce its consequences—even when the story it gives us is not the one we would have preferred.