A sandbox reference for Atmaverse contributors who understand the prisoner's dilemma and want to design aliens that feel real.
You've got a species concept that works perfectly as a game-theoretic construct - a fascinating utility function, an interesting relationship to cooperation, a novel position on the Atma question. Now you need it to be an organism. It needs a body, a planet, a sensory world, an evolutionary history, and a reason why its game-theoretic properties emerged from natural selection rather than being magicked into existence by an author who needed them for a plot.
The Atmaverse standard is Greg Egan / Peter Watts levels of biological plausibility. That doesn't mean you need a biology degree. It means you need to understand a handful of principles about how evolution actually works, so that your species' interesting strategic properties emerge from its biology rather than being stapled on top of it.
The single most reliable method for designing a plausible alien species is to start with the planet and work inward. The environment determines what survival strategies are viable. The survival strategies determine the body plan. The body plan determines the sensory world. The sensory world determines the cognition. The cognition determines the game theory.
If you reverse this chain - starting with "I want a species that always defects" and working backward to biology - you'll usually end up with something that feels contrived. But if you start with "what kind of planet would produce organisms whose cognition doesn't support iterated cooperation" and work forward, the game theory falls out naturally and the species feels inevitable.
Every major Atmaverse species demonstrates this principle, and the diversity of starting environments is what produces the diversity of game-theoretic positions:
Crucible (apocritans): A high-gravity ferrous super-Earth where standing up is an engineering problem. Bioaccessible metals everywhere - in the soil, the water, the atmosphere. Every organism incorporates environmental metals into its structure because the alternative is structural failure under 1.8g. This produced a species whose bodies are progressively constructed from biological and metallic components, whose identity is located in the process of construction itself, and whose technology is extractive and metallurgical because their biology is extractive and metallurgical. Their instrumental, transactional approach to federation cooperation follows directly: a species that builds itself from environmental resources treats all relationships, including interstellar ones, as resource-access questions.
Regez (ansale'wit): A tidally locked super-Earth around a red dwarf, volcanically hyperactive, crust riddled with conductive mineral water. Chemosynthetic metabolism powered by hydrothermal chemistry. Every medium - air, rock, water - conducts electromagnetic signals. This produced a species whose cognition is electromagnetic, whose primary sense is EM perception, whose communication is involuntary broadcast, and whose identity is distributed across a planetary fungal network. Their strict identity boundaries and their horror at the Atma program both follow from the environment: when your self is partially constituted by the network you're embedded in, severing that network connection is a specific, nameable harm that a species with portable selfhood might not even recognize.
Arachne IV (Weavers): A warm, ocean-dominated super-Earth with volcanic archipelagos, 24% oxygen, and 1.2g. Complex tidal-zone ecology with spatially variable resources. This produced colonial organisms whose subunits tightened into integrated paddle-array bodies with distributed cognitive strands, whose reproduction involves engram exchange through ingestion, and whose identity is located in the cognitive pattern rather than the substrate. Their Atma program - preserving species by archiving cognitive patterns - is their reproductive biology scaled up. The environment that selected for engram exchange also selected for the philosophy that made Atma seem obviously right.
Practical method: When you have a game-theoretic concept you want to explore, ask yourself: what kind of environment would produce an organism with this property? What selection pressures would make this cognitive or behavioral trait adaptive? Work from the planet to the body to the mind to the strategy, and the result will feel grounded.
Not everything about Earth life is universal. When designing aliens, you need to know which biological features are likely to show up on any life-bearing planet (convergent) and which are accidents of Earth's specific evolutionary history (contingent).
Convergent features are the ones that physics and chemistry demand. Eyes are convergent - photon detection is so useful that eyes have evolved independently over 40 times on Earth. Bilateral symmetry is convergent for organisms that move in a consistent direction. Streamlined bodies in fast swimmers, cephalization, wings for flight - these features are reasonable to include in alien designs because the physics that makes them useful doesn't change on other planets.
Contingent features exist because of specific historical accidents. The vertebrate body plan with exactly four limbs is contingent - there's no physical reason land animals need exactly four limbs. DNA as the information-storage molecule is likely contingent. The specific amino acids in Earth proteins are probably contingent.
For species design: Feel free to include convergent features in your aliens. But don't default to Earth-specific contingent features without a reason. The Weavers have a radial paddle array because they descend from colonial siphonophore-like organisms. The apocritans are hexapod arthropoids because their lineage was arthropoid. The ansale'wit build humanoid Images as vehicles, but their actual organism is a fungal network that bears no structural resemblance to any Earth animal - the humanoid shape is a choice, selected for mechanical utility, and they can grow different ones. Each body plan follows from evolutionary history, not author familiarity with humans.
Evolution doesn't design organisms from scratch. It modifies what already exists. Every feature is a modification of something that came before, repurposed and adapted under new selection pressures. This means organisms are full of compromises, dual-use structures, and features that only make sense when you know the evolutionary history.
This is actually great for worldbuilding, because the most realistic alien species are the ones that feel slightly jury-rigged - where the same structure serves multiple functions, where you can see the evolutionary history in the design compromises.
Atmaverse examples: The Weavers' engram exchange evolved from the colonial ancestor's reproductive mode. Early colonies reproduced through fragmentation and budding; as they integrated into unified organisms, budding became more controlled, eventually differentiating into Sex 1 (backup) and Sex 2 (reproduction). The intimacy-through-ingestion behavior is a domestication of what was originally a purely reproductive process - which is why the Atma program, designed by a species whose deepest intimacy is ingestion, involves a probe that consumes the target species. The technological lineage from "colonial budding" to "interstellar uplift probe" is traceable step by step.
The ansale'wit Image is a repurposed growth capacity - the same organism that evolved to thread fungal hyphae through rock learned to thread them through a chitin-analogue shell instead, co-opting substrate colonization for locomotion. The EM cognition is a byproduct elevated to a primary function: it started as metabolic noise from electrochemical activity in metallically doped tissue and became the basis of their entire social and cognitive architecture.
Apocritan metallic integration began as passive environmental absorption - organisms incidentally incorporating iron from the mineral-rich environment - and was refined over tens of millions of years into active biomineralization with specialized cells and enzymatic pathways. Their tool use emerged as a continuous extension of this process: the transition from "finding useful rocks" to "shaping useful rocks" to "smelting ore" has no bright line. There is no boundary between apocritan biology and apocritan technology, because the technology is what the biology became when it exceeded what the body could do unaided.
For species design: When you're designing an organ or body structure, think about what it was before it was what it is now. The most convincing alien anatomies are the ones where structures have been repurposed from ancestral functions, and you can tell.
The way an organism perceives its environment determines the kind of cognition it develops, which determines the kind of social structures it builds, which determines its game-theoretic properties. This chain is the most important thing to get right for Atmaverse species, because it's where the biology connects to the politics.
As demonstration:
Ansale'wit perceive primarily through electromagnetic fields. Their cognition generates EM signatures as a byproduct, the way human neurons produce faint electrical fields - except amplified by evolution and conducted through metallically doped tissue. They feel each other's cognitive activity the way you feel where your own hand is. This produces a social cognition where mutual awareness is the baseline, privacy is an effortful achievement (like holding your breath), and the decision to be private is itself public. Their strict identity boundaries - same brain, different network, different person - are a cultural corrective to biological leakiness. Their horror at Atma is a direct consequence of perceiving selfhood as partially constituted by relational context.
Weavers perceive through composite 360-degree vision (eyespots on every paddle) and rich chemoreception, but their social cognition is built on engram exchange - integrating another's cognitive patterns through ingestion. This produces a social world where deep communication requires literal consumption of the other's cognitive material, surface communication (gesture, grinding sounds) is considered adequate for discussing the weather, and every non-Weaver interaction is operating in a mode the Weavers consider shallow. Their patience with other species is the patience of someone communicating through earplugs.
Apocritans perceive primarily through chemoreception - dense antennal arrays detecting hundreds of distinct pheromone compounds at concentrations far below human instrumentation. Their social world is a continuous chemical information layer that shapes mood, attention, hierarchical perception, and motivation below conscious deliberation. This produces a social cognition where "what I think" and "what the pheromone environment is shaping me to think" are not distinguishable from inside - which is why workers experiencing pheromone-free environments for the first time describe it as "clearing" or "waking up" or "becoming loud," and why some of them don't experience the freedom as improvement.
For species design: When designing a species' senses, think about the downstream implications for social cognition. The richest species designs are the ones where the game-theoretic properties are direct consequences of the sensory world.
How a species reproduces determines more about its social structure, power dynamics, and game-theoretic properties than almost any other single factor.
The key variables: how many offspring per event, how much parental care, whether reproduction is sexual or asexual or something else, whether there are distinct reproductive roles, and how mate selection works. Each combination produces different selection pressures on social behavior, and those pressures have direct game-theoretic consequences.
Weavers: Sex 1 (mutual pearl backup between partners) and Sex 2 (novel reproduction from hybrid engram material) are both built on engram exchange. Partnership is literal mutual backup - you carry each other's potential resurrection. Offspring begin life carrying cognitive fragments from both parents, knowing their parents' way of thinking from the inside. This produces a social structure where trust is absolute within partnerships (you are placing your continuity inside another organism's body), where knowledge and communication involve literal cognitive integration, and where the Atma program - archiving and reconstituting strangers' patterns - is the natural extension of what they do for loved ones.
Apocritans: Queen-controlled reproduction with epigenetic caste determination. Any egg can become any caste; queens control which outcome occurs by controlling nursery chemistry. Workers are technically capable of reproduction but socially prohibited from it; unauthorized reproduction is a capital offense for the drone. The food chain creates an inescapable dependency loop: adults can only eat larval excretion, larval excretion carries the queen's pheromone signature, chronic pheromone ingestion shapes workers' neurochemistry to experience the queen's authority as natural. The entire caste system is embedded in the food supply. This produces a civilization where the game-theoretic properties are almost entirely determined by the reproductive biology: the queen's biological controls over reproduction, food chemistry, and caste determination create the political structure, and political authority follows from rather than precedes biological control.
Ansale'wit: Spore dispersal for baseline reproduction (hardy dormant packets, sexual recombination with nearby spores, no inherited memory), cross-pollination at organized galas for selective genetic contribution (community-negotiated, socially significant), and network propagation for cognitive inheritance (persistent electrochemical patterns in the substrate that echo a dead individual's signature, picked up by new spores as vague past-life memories). This produces a species where identity is partially heritable through the environment rather than through the genome, where "home" is the network that carries your ancestors' cognitive echoes, and where displacement from that network - the ansale'wit diaspora - is a loss that cannot be fully compensated by building new networks elsewhere.
For species design: Work out your species' reproduction strategy early, because it constrains everything else. And look for the game-theoretic consequences - they're almost always there.
Every organism exists in an ecological web - predators, prey, competitors, parasites, mutualists. The most convincing alien species are the ones where you can glimpse the rest of the ecosystem around them.
This matters for game theory because inter-species cooperation isn't unique to sapient civilizations. It's everywhere on Earth, driven by mutual benefit in iterated interactions - the same logic as the federation, scaled down. Understanding why cooperation evolves between species gives you tools for designing how your species relates to others.
The apocritan food chain is a powerful example. The adult dependency on larval excretion isn't a standalone biological quirk - it's an ecological relationship between life stages that has been co-opted by the queen caste as a control mechanism. The larvae are effectively a different ecological niche than the adults (omnivorous solid-food consumers vs. obligate liquid-diet consumers), and the dependency between them structures the entire civilization. On Crucible, this would have coevolved with the broader ecology: what do the larvae eat? What competes with them for those food sources? What predates on vulnerable pupae? The answers would shape apocritan hive architecture, military behavior, and resource extraction patterns - all of which have downstream effects on their game-theoretic properties in the federation.
For species design: Even if you're only writing up one species, think about what else lives on their planet. The answers will suggest behavioral adaptations that make your species feel like part of a living world, and those adaptations often have game-theoretic implications that enrich their role in the federation.
If you're starting with a game-theoretic concept and need to find the biology, here's a method:
Start with the strategic property you want to explore. State it clearly. "A species that cooperates through mutual threat rather than mutual benefit." "A species where individuals can't defect because they literally share a nervous system." "A species that's federation-compatible in one life stage and species #0 in another."
Ask what cognitive architecture would produce this. What kind of brain (or brain-equivalent) would make this strategic behavior the natural default? Does it require specific sensory capabilities, memory structures, emotional responses?
Ask what sensory world would produce this cognition. Senses don't just gather information - they shape the categories the brain thinks in. An EM-perceiving species has a fundamentally different model of self and other than an acoustic one, which produces different social reasoning, which produces different game theory.
Ask what body plan would support these senses. The body plan follows from the sensory requirements. Dense chemoreceptor arrays need large exposed surfaces (apocritan antennae). EM perception needs metallically doped conductive tissue (ansale'wit fungal core). 360-degree composite vision needs distributed eyespots (Weaver paddles).
Ask what environment would select for this body plan. Why would evolution favor this collection of traits? The answer is almost always environmental: a specific combination of pressures that makes this body plan more successful than alternatives.
Check for consistency. Does the chain hold together? Are there contradictions? Contradictions aren't always problems - sometimes they produce interesting compromises that make the species feel more real - but they should be acknowledged rather than ignored.
The "humans but with one thing changed" problem. A species that's basically human psychology in an alien body isn't speculative biology - it's a costume. The most interesting species are the ones where the alien biology produces genuinely alien cognition. If you can swap your species out for humans with a cultural quirk and the story doesn't change, the biology isn't doing enough work.
The monoculture problem. Earth has one sapient species and it has thousands of cultures. Your species should have internal diversity too. The apocritans have competing hive-ships with different queens, different internal cultures, and different relationships to the federation. The ansale'wit have the homecoming faction vs. the diaspora-identity faction, with shipboard dialects that have diverged from planetary ones. Monolithic "planet of hats" species are a red flag that the biology hasn't been thought through deeply enough.
The "too perfect" problem. Real organisms are messy. They have vestigial structures, suboptimal compromises, developmental disorders, and ecological dependencies that create vulnerabilities. The Weavers' weakness in continuous mathematics (calculus, estimation, order-of-magnitude reasoning) is a cognitive consequence of their native binary paddle-fissure numeracy - they compensate through brute-force simulation, which works but which human engineers find "almost philosophically offensive." Include the jank. It's what makes a species feel biological rather than designed.
Forgetting scale. A spacefaring species needs enough population to run a civilization and enough industrial capacity to build spacecraft. This means cities, infrastructure, education systems, supply chains. The apocritans have kilometer-scale hive-ships that are self-sustaining industrial ecologies. The Weavers have architectural complexes on Arachne IV that are, in their entirety, computing systems - walls that process, floors that sense, buildings that think. If your species has starships, it has a civilization that supports building starships, and that civilization should be visible in your writeup.
For speculative biology: All Tomorrows (Nemo Ramjet), Expedition (Wayne Barlowe), Snaiad (Nemo Ramjet), The Future Is Wild (Dougal Dixon). For the intersection of biology and game theory: The Selfish Gene (Richard Dawkins, specifically the chapters on evolutionarily stable strategies), The Evolution of Cooperation (Robert Axelrod). For what "hard sci-fi aliens" can look like when done well: Blindsight (Peter Watts), the Orthogonal trilogy (Greg Egan), A Fire Upon the Deep (Vernor Vinge), Children of Time (Adrian Tchaikovsky).
For the interactive game theory foundation: The Evolution of Trust by Nicky Case.