The Gardens of Grey Nax

Cooperation as Dominant Strategy

7 min read

Cooperation as Dominant Strategy

Synthesized from: Watts & Strogatz small-world network research; Matthew Colville, “Towards Better Rewards” (MCDM, YouTube); prisoner’s dilemma game theory; and vault notes on behavior as communication and the gift economy

The Core Argument

The question of whether a community cooperates or defects is not primarily a question of character. It is a question of incentive architecture — what the system rewards, what the communication layer says, and whether cooperation or defection is the dominant strategy given those conditions.

You cannot ask people to cooperate. You have to build a system where cooperation is what the incentives point toward.

The Nash Equilibrium Problem

In the prisoner’s dilemma, two players each choose independently to cooperate or defect. The Nash equilibrium — the state where neither player can improve their outcome by changing strategy unilaterally — points to mutual defection, even though mutual cooperation produces better outcomes for both.

This isn’t because players are selfish or short-sighted. It’s because the reward structure makes defection the rational move regardless of what the other party does. The system selects for defection.

Run this game at scale, iteratively, with market signals as the communication layer, and you get late-stage capitalism. Not as a moral failure — as an adaptation. Companies that internalized externalities couldn’t compete with companies that didn’t. The reward structure selected for defection over generations. The surviving entities are optimized defectors. “Pay the fine, don’t fix the car” is not a psychopathic choice; it’s the Nash equilibrium of a system that prices human life below liability exposure. Ford Pinto. Boeing. Opioid manufacturers. Each decision individually rational within the system’s logic. The communication layer said defect, so defection was selected.

This is behavior-as-communication (Behavior as Communication) operating at institutional scale. The corporation isn’t broken. It’s responding correctly to what the system is asking of it.

The Communication Layer

Matthew Colville states the design principle directly: the behavior a game rewards is the behavior a game encourages. This is the Nash equilibrium in design language. The reward structure is the communication layer. Change what gets rewarded, change what behavior the system selects for.

Classic D&D gave experience points for gold found — not monsters killed. This made players intensely curious, proactive, and exploratory, because curiosity and proactivity were the dominant strategy for advancement. The same players, in 5th edition’s milestone leveling system, become passive — because the reward structure no longer differentiates between engaged and disengaged behavior. Show up, roll dice, level up at the same rate as everyone else. The players haven’t changed. The incentive architecture has.

The implication is uncomfortable: passive players are not a failure of character. They are a rational response to a system that doesn’t reward active engagement. The fix is not to motivate people better. It’s to redesign what the system rewards.

This applies directly to community design. A community where showing up has no differential reward from not showing up will produce low engagement — not because residents don’t care, but because the system communicates that caring doesn’t matter.

The Small-World Wrinkle

Watts and Strogatz’s network research adds a structural dimension. In simulated prisoner’s dilemma games, cooperation tends to emerge and stabilize in highly clustered networks — small groups with repeated interactions. The familiarity and iteration that produce tit-for-tat strategies require enough density of contact that defection has social consequences.

Introduce shortcuts — weak ties connecting distant nodes — and the dynamics change. Cooperation can collapse: defection spreads faster through weak ties than cooperation does, because defectors exploit the asymmetry of connecting to cooperators who don’t yet know their reputation. This is the keyboard warrior phenomenon: people are generous to neighbors and cruel on Twitter, because the network structures carry different incentive architectures.

The implication for Wellspring: the internal community needs enough clustering — enough repeated interaction and genuine relationship — that cooperation norms can stabilize. The outward-facing connections (weak ties to the broader Durham neighborhood) carry real benefits but also import the defection dynamics of larger anonymous networks. The design challenge is capturing the connective benefits of weak ties without exposing the internal community to the defection dynamics they can carry.

Behavior as Communication as Escape Hatch

Here is where Behavior as Communication does something the standard game theory framing cannot: it makes cooperation the dominant strategy even when the other party defects.

In the standard prisoner’s dilemma, if your opponent defects, you lose by cooperating. Defection is contagious because the costs are real. But the behavior-as-communication frame changes the meaning of defection. The neighbor who fails to do their commons maintenance share isn’t making a strategic choice to exploit you — they’re signaling something: they’re overwhelmed, they’re struggling, they need something they don’t know how to ask for.

If you respond to that signal with curiosity and care rather than retaliation, you aren’t being naive — you’re playing a different game. You’ve changed the communication layer. The response to apparent defection is: what is this telling me? That question keeps the door open for cooperation in a way that tit-for-tat retaliation closes.

This only works inside a community with sufficient relational density to read the signals accurately. A stranger’s defection is opaque; a neighbor’s defection is legible. The village has to know its members well enough to tell the difference between strategic exploitation and a cry for help. Which is, again, the design problem: The First Step and the Desire Path.

Structural Change Precedes Cultural Change

The most important implication: you cannot build a cooperative community by asking people to cooperate. You have to change the reward structure so that cooperation is what the incentives point toward.

Wellspring’s structural work is, in this frame, game design:

  • Frozen carrying costs remove the zero-sum competition over housing expenses that makes neighbors into adversaries in conventional housing markets. Your neighbor’s stability doesn’t threaten yours.
  • Cooperative governance makes the community’s outcomes genuinely shared. Your investment in the commons improves your own situation. Defection from collective maintenance costs you directly.
  • Gift economy infrastructure (heritage library, tool lending, skill-sharing) makes generosity self-reinforcing rather than self-depleting. Giving increases the total available to receive. See Lewis Hyde — The Gift.
  • Mutual aid without ledger-keeping means helping someone who isn’t helping back is still the right move — because the returns are diffuse and long-term, not immediate and bilateral. See Mutual Aid.
  • Relational accountability means the cost of defection isn’t retaliation but honest conversation — which is lower stakes and more repairable than punishment. See Relational Accountability.

None of this requires residents to be unusually virtuous. It requires a reward structure that makes ordinary people’s ordinary self-interest align with collective flourishing. That’s the design problem. That’s what the structural work is for.

The Colville Corollary

Colville’s cards — explicit goals with explicit rewards — are a tool for making the reward structure visible to players. The underlying principle generalizes: people are not unmotivated, they are under-informed about what the system is rewarding. Making the connection between behavior and reward legible changes behavior.

For Wellspring, this suggests that community norms shouldn’t just be stated — they should be embedded in what the community visibly rewards. The person who checks in on a struggling neighbor, who maintains the garden, who teaches a skill, who shows up for someone’s hard day — that behavior should be legible and honored, not just assumed. Not through a formal points system (which would commodify the gift), but through the community’s stories about itself, its rituals, its memory of who showed up and when. The community of memory (Habits of the Heart (Bellah)) is partly a record of cooperation made visible over time.

What This Changes

The standard framing of community building: find good people who want to cooperate, put them together, hope it works.

The design framing: build the incentive architecture so that cooperation is the dominant strategy, and ordinary people — not saints — will cooperate. The communication layer will say: the behavior this community rewards is showing up for each other.

That’s the game. Design it correctly.

Graph View

Backlinks