Dark Provinces: the prisoner’s dilemma

The prisoner’s dilemma is a foundational concept in game theory used to study cooperation and self‑interest. In its simplest form, two individuals must independently choose whether to cooperate (partner) or defect (to act selfishly, or free-ride). A dilemma arises because:

• Mutual cooperation (partnering) produces a collectively good outcome.
• Mutual defection (free-riding) produces a collectively disastrous outcome.
• But free-riding is individually tempting, because if one free-rides while the other partners, the free-rider does best of all.

Even though cooperation would benefit everyone, rational self‑interest pushes individuals toward defection or free-riding —especially when trust is uncertain.

The key insight is that individually rational decisions can produce disastrous group outcomes. The prisoner’s dilemma is not about villains or bad people; it is about incentive structures. When costs and benefits are unevenly distributed in the short term, cooperation becomes fragile even among well‑intentioned participants.

Importantly, real-world prisoner’s dilemmas are rarely one‑shot. In repeated or group versions, additional dynamics emerge: trust, reputation, retaliation, exhaustion, inequity, and burnout. Nuanced uses of the prisoner’s dilemma therefore focus less on individual transactions and more on how systems slowly corrode or stabilize cooperation over time.

Where the prisoner’s dilemma lives in Dark Provinces

Dark Provinces explicitly states that its core loop is a repeated prisoner’s dilemma embedded in a cooperative survival game. This is not metaphorical: the prisoner’s dilemma appears as a formal decision point every challenge round, embedded right in the rules.

Each challenge forces every player, independently and secretly, to choose between two roles:

• Partner (cooperate): commit a trait card, share the risk, contribute to the group’s success.
• Free‑Ride (defect): commit no card, avoid immediate cost, but still benefit if the group succeeds.

This choice occurs after discussion but before resolution, and before the vote is held, it is simultaneous and private. No player knows who will cooperate until the reveal. That simultaneity is critical: it recreates the uncertainty that makes the prisoner’s dilemma function.

Payoff asymmetry: why free-riding is rational

Dark Province's "Universal Outcome Table" makes the prisoner’s dilemma mathematically explicit.

On a successful challenge:

• Partners expend two trait cards and gain:
• +1 Glory
• +1 Trust
• Free‑Riders expend one trait card and gain:
• +3 Glory
• −1 Trust
• +1 Stress

This asymmetry is deliberate. Free‑riders gain three times the glory of partners while avoiding the immediate cost of expending an additional trait card, which represent your health in the game. Glory is the primary individual scoring and recovery currency for your health, so in the short term, free-riding is often the better individual choice.

Critically, the game does not hide or soften this incentive. The game is explicitly structured such that free-riding is often rational and is intentionally rewarded. This mirrors the prisoner’s dilemma’s central tension: the strategy that helps you most right now harms the system that keeps you alive.

Even in failure, the asymmetry persists:

• Partners suffer more stress than free‑riders.
• Free‑riders still lose trust, but avoid the heavier cost of commitment.

This means partners generate fewer rewards relative to free-riders on success— and it is also the riskier strategy on failure.

Why cooperation still matters: cumulative systems

If the Dark Province’s prisoner’s dilemma rewards stopped at individual payoff, the game would collapse into opportunism. It does not, because it’s long‑term systems convert repeated defection into systemic collapse. Four cumulative tracks enforce group‑level consequences:

Trust (individual, persistent)

Trust loss from free-riding is small per challenge, but trust gates leadership, credibility, and several trait synergies. Players with very low trust cannot be leader and and lose the reward distribution and other powers that come with the leader position. Over time, habitual free-riders lose influence even as they hoard glory.

Stress (individual, escalating)

Stress accumulates from challenges and use of expedient p traits. At 10 stress, a blow‑up occurs: an adaptive q trait of that player permanently mutates into an expedient p trait. This models burnout and erosion of healthy behavior. Stress does not care why it was accumulated—only that pressure is escaping somewhere.

Reputation (collective moral ledger; cumulative external judgment)

Reputation functions as an additional cumulative system that translates repeated cooperation or exploitation into the group’s standing in the world, extending the prisoner’s dilemma beyond the party itself. Unlike trust, which is individual and directly affects coordination, reputation is collective and external: it records how the group’s patterns of action are perceived by factions, institutions, and the broader social environment. In game‑theoretic terms, Reputation introduces an iterated audience effect: choices made to optimize short‑term individual payoff (especially free-riding supported by exploitative or deceptive traits) may succeed tactically, but gradually degrade the group’s credibility, invoking harsher penalties, fewer options, and hostile reactions in future challenges. Because reputation modifies outcomes, faction responses, and final scoring multipliers, it ensures that cooperation is not only a private bargain among players but also a public signal assessed over time. This mirrors real repeated prisoner’s dilemmas in social systems, where free-riding does not only erode trust among peers but also reshapes how outside agents allocate risk, tolerance, and enforcement. In Dark Provinces, reputation therefore closes the loop between internal incentives and external consequences: a group that survives by habitual self‑interest may still persist for a time, but it does so under worsening conditions, until the cost of lost legitimacy outweighs the gains of defection and the system collapses under its own moral debt.

Doom (collective fail clock)

Doom advances on failures and panic events. If doom reaches its threshold, the campaign ends in total loss regardless of individual scores. No amount of personal glory can outrun doom.

These tracks ensure that while free-riding is locally optimal, excess free-riding destabilizes the system that makes success possible.