Machina Ludens

On play as a coordination mechanism, and the cryptography that could finally hold its magic circle open at scale.

By shaurya.eth

Edited by Claude Opus 4.8

Typeset in MEKZANTINE (alpha v1), a typeface by MEK — mek.gallery · @michaelmicasso

01 ·

Motivation

I have played video games all my life, and for almost as long I have heard them written off as a waste of the hours that go into them. What that dismissal never accounts for is how much serious work a good game pulls out of the people who fall into it. Factorio hands you a world and a chain of optimization problems and asks you to build order out of raw material. Shooters like PUBG and team games like Dota demand a deep command of an invented system and the nerve to act on it in an instant. Grand strategy and 4X games train long horizons and a feel for decisions that only pay off two hours later. An ARPG like Path of Exile wants a constantly updated knowledge base, a trader's sense of its economy, and the pattern recognition to turn all of it into something of your own. People give these games years of real cognitive labour, the kind we would call work in any other setting, and they do it first, and for almost all of them entirely, for no money at all. The pay, where it comes, comes late and to almost no one, and it was never what drew the labour out to begin with.

That free pull is the most undervalued resource I know of, and the whole of this essay is a single question about it: whether it can be made to do serious work, or whether it is condemned to stay trivial.

02 ·

Play: Older Than Culture

Johan Huizinga gave the play-creature a name, Homo Ludens, and argued that play is older than culture and runs underneath it, that the serious institutions we treat as bedrock began as games and never fully shed the fact. A few of his observations carry the rest of this essay. Play is voluntary before it is anything else, and the instant it is compelled it stops being play. What keeps it alive is the magic circle, the bounded space where the game's own rules hold and the ordinary world is suspended, and that circle lasts only as long as the play is done for its own sake, with no outside purpose leaning on it. Inside it play builds an order of its own, a small complete world that allows no interference from beyond its edge, and it defends that order in a way that tells you what it prizes. The cheat, who breaks the rules while still pretending the game is real, is forgiven within limits; the spoilsport, who refuses to grant that the game matters at all, is driven out, because the cheat keeps the circle intact and the spoilsport denies it exists. The uncertainty of the outcome is what gives play its moral weight, the tension in which courage and fairness and patience are actually tested, and drain that tension and the game dies with it. Much of the play that built culture was a contest for standing rather than for goods, a striving to be seen to win, where what was won was status and not material gain. And play has always set itself apart, insisting it belongs to its players and not to the world outside, leaving behind communities that outlive any single game.1

Look at the world this way and it comes apart into pockets of these invented realities, each with its own boundary and rules and players, and together they are most of what keeps a civilization running. You could object that what really holds it together is incentive and not play, and often that is true, but the underlying mechanisms can and do run as pure play, which is exactly what the games I began with keep demonstrating. The thing to hold on to is that this most productive form of coordination is also the most delicate, kept alive by nothing sturdier than the wanting of the people inside it.

on stakes

The Stakes Paradox

If the circle survives on wanting alone, then money ought to be its natural enemy, and Huizinga believed it was, insisting that play lasts only as long as it stays free of material interest and that the instant profit or gain enters, the play is corrupted and the circle breaks.2 Most of the time this is exactly what happens. But not always, and the exceptions are where it gets interesting. Some games are built so that money sits at their very centre and the play survives it whole, and a single case like that is enough to turn the rule into a paradox. The question is no longer whether money breaks play. It is why money breaks it in one game and leaves it standing in another, which turns out to be a question about how the game is built.

Gambling puts money at the very heart of a game, the chips serving at once as the stake, the score, and the medium of play, with none of the symbolic distance a chess piece keeps from anything that actually matters. By Huizinga's reasoning this should be the end of play, interest folded so far in that nothing done for its own sake could survive it. And yet it is plainly play. People sit to it for the tension and the uncertainty and the shape of the thing, the money fully inside the circle and the circle entirely intact. So the rule is too strong, and the thing worth understanding is what we were actually looking at, when we were certain money would wreck the play.

What we were looking at was ascription. Stakes are never simply there; we assign them. The chips matter because the table agrees they matter, and money matters because the world agrees it does, a status the world confers and not a trait the thing carries, and neither agreement runs any deeper than the other. The serious world and the game are the same move aimed at two different realities.3

If money is not the thing that breaks the circle, something else is, and it is worth naming precisely. The circle breaks when the stakes stop being played for and start being extracted, and that turn has a tell. Caillois, refining Huizinga, had named this corruption already, though he watched it from the other side, play's forms spilling out into ordinary life; the breach that concerns us runs the other way, the world's interest seeping into the circle, but it is the same wall giving way at the same seam.4 A magic circle holds in proportion to the distance between its stakes and a person's survival: chess survives any degree of seriousness because losing costs nothing once you stand up, while the nearer the stakes press to genuine need, the more easily the play curdles into extraction. The thinner that buffer, the harder the circle is to hold. Keep this close, because it comes back the moment we ask what it would take to make play do serious work, when everything will turn on what happens to the play as that distance closes.

The distance-to-survival gradient A single axis running from stakes pressed against survival on the left, where the buffer is thin, to stakes far from survival on the right, where the buffer is thick. The nearer a game's stakes press to real need, the more its magic circle curdles into extraction. Chess sits far from survival, a potlatch nearer, and stakes pressed against need at the extraction-prone end. distance between stakes and survival gambling the rent a potlatch chess near survival · buffer thin far from survival · buffer thick buffer = distance to survival;the thinner it gets, the more the circle curdles into extraction

There is a sharper way to say what for its own sake means, and it comes from Bernard Suits, who defined the playing of a game as the voluntary attempt to overcome unnecessary obstacles, taken up in what he called the lusory attitude: the player accepts the rules just because they make the game possible, and keeps to the harder path with an easier one in plain sight.5 That attitude is the thing to watch, more than the size of the prize. A circle breaks when the reason for facing the obstacle changes, when the rule stops being kept for the play it makes possible and starts being endured for what can be taken past it. So the distance-to-survival tell was reading something real, but reading on one axis what has two. One axis is the structure of the stake, whether it is played for or extracted. The other is the regulation of the motive, whether the wanting is the player's own or pressed on him from outside, and here the psychology is settled: a stake a player has made his own sits beside the wanting without dissolving it, while a stake that controls him dissolves it on contact.6 Money near survival curdles play because nearness is the ordinary way a played-for stake becomes a taken one.

The clearest cases are the ones where the weight was real and the play held anyway. A penalty shootout in a World Cup final, a nation's hopes set on a single kick, is deep play in Geertz's sense, the game held under all that weight and still played for glory and not for any purse. Goffman found the same at the gaming table, where what is courted is the action, the fateful instant that only real consequence can manufacture. The climbers and jumpers who stand to lose their lives to a single error were clearest of all about what drew them, which was the freedom on the far side of the risk.7 Play has carried weight as heavy as weight gets and stayed play. What it has never survived is having its wanting bought out from under it, which is a different thing entirely, and the distinction is the whole of what follows.

03 ·

Play as a Means of Coordination

We have just seen play hold together on nothing but the wanting of the people inside it, and break the moment that wanting is replaced. Widen the lens from a single table to a whole civilization and the same question returns at scale: what makes large numbers of strangers act in concert? For the whole of our history the answer has come in only two forms. There is authority, which moves people by command, the hierarchy running the state and the army and the church, holding millions in line through law and rank and the threat behind them. And there is incentive, which moves people by payment, the market and the price and the wage, coordinating strangers who will never meet through nothing more than what things cost. Almost everything large we have ever built runs on one of these, or on the two of them braided together, and each took centuries to harden into the institutions we were handed.8

Both tools carry a tax that has nothing to do with how fast they run. Authority, wherever it scales, tends to concentrate and then to be captured, bent over time toward whoever has drawn close enough to work it.9 Incentive corrupts in its own fashion once it becomes the lever, every measure raised as a target gamed the instant it is one, every system drifting toward extraction by those who learn its edges.10 Neither is a flaw in some institution that better management could reform away; each is the standing tendency of the tool itself, the perversion it breeds in whatever gets built from it. And the argument we have about this never gets past the two of them, market against state, the same quarrel restaged for a century, each side certain the other tool is the culprit, as though those were the only two ways the thing could be done.

They are not. There is a third way to make people act together, and we have left it almost entirely unexplored as a tool, though every one of us has felt it work. I mean play, and I mean it plainly, the fun of the games I began with, Factorio and Dota and Path of Exile, where thousands coordinate in tight and demanding concert and not one of them was paid or compelled. Inside a game people show up unbidden, take roles no one assigned, hold to rules with no one over them, and give their fullest and most honest effort, drawn out by wanting alone, and the expectations the game throws up settle into norms that outlast any single round. This is real coordination, doing the very work that command and payment are built for, and it is the one form of it we have never seriously tried to put to use. So there are three forces, not two: authority, which commands; incentive, which pays; and play, which runs on the wanting itself. The instinct that command and payment do not exhaust the ways people act together is an old one and in good company, the clan that Ouchi set beside the market and the bureau, the network that Powell found to be neither, the peer production Benkler watched assemble itself out of little more than the wanting; what stands at this third corner is play, where their clan and network lean on shared values and on trust, and play leans on the wanting alone.11 Set the three at the corners of a triangle and the picture sharpens, the two we know how to wield sitting cold along the base while play waits at the third corner, warm and free and barely touched at scale. The norms a circle throws up are its own, a deposit laid down from the inside the way a commons makes its rules, binding the next round because the players made them and not because a price or a command was set over them; reputation is the test of it, for kept inside the circle and unpriced it is the circle's own memory and costs the play nothing, while the instant it can be cashed it has turned into incentive under a kinder name and curdles on the same gradient as any other stake.12 What makes that corner worth reaching has little to do with speed, though play unburdened of payment and command can be quick enough. The point is that the perversions which hollow out the other two find nothing in it to grip, no reward to game and no office to capture.

That last claim has to answer for the last fifteen years, in which play has been turned into one of the most efficient engines of extraction ever built, the free-to-play economy and the gacha pull and the whole grammar of engagement bent until they empty a wallet. But none of it is play corrupting itself. Look at where the grip actually sits. The randomized reward priced in a currency you must buy is incentive, the cold tool slipped back inside the circle; the platform that owns the rules, tunes the schedule, and keeps the record of everything you do is authority, the office rebuilt behind the fun.13 What these designs prove is narrower than it looks. Play laid over a hidden price and an owned operator will extract, which is the very failure this argument predicts, the circle paid away from within while the costume stays on. Strip the price and strip the owner and there is, once again, no reward to game and no office to capture.

There is a reason that corner has stayed empty. Play coordinates only for as long as the magic circle holds around it, the bounded space where the game's own rules govern and the ordinary world is kept out, and that circle survives only while the playing is done for its own sake. The two tools that could make play larger, incentive and authority, are exactly the two that break the circle, because paying for the play or commanding it replaces the wanting with something taken, and the circle does not survive that substitution. So play has stayed small, held there by a hard circularity: the only known ways to scale coordination are the two that dissolve the very thing play needs in order to work. Look back across the serious, lasting coordination we have built and it is uniformly cold, and from that we conclude that play cannot scale, when what we are really reading is a record of the only survivors there could be, the ones that paid the circle away in order to grow.14

And yet we have proof that it can. The strongest case is the war that EVE Online fought at B-R5RB, where more than seven thousand players, gathered into alliances no company chartered, threw years of accumulated labour into one battle that destroyed something near three hundred thousand dollars of real value, the whole of it set off by a single automated payment missed on time. Foldit took the same free pull and aimed it at the laboratory bench, and its players, almost none of them biochemists, solved the structure of a retroviral protease that had defeated the researchers, a result clean enough to publish.15 These are coordination at the scale the argument actually asks for, strangers in their thousands held in concert by rules that bound without a person behind them to enforce. What the smaller games I began with show is something a little different and worth naming exactly, a single circle grown good and then struck a million times over. That repetition is the native shape of the thing, for a magic circle is a boundary by definition, and bounded things scale the way every durable commons does, by nesting and federating under a shared order rather than swelling past the size a circle can hold.16 For two decades, hiding in plain sight under the word entertainment, games have run exactly this, real and densely peopled coordination grown from play and its norms with only the thinnest film of the cold tools laid over the top. They are the single clean demonstration that play scales when the circle is kept whole, which turns the whole problem into one precise question: can a magic circle be built and held at scale without the two tools that have always been the only way to hold anything at scale? A circle needs two things to stand. It needs an order of its own, rules that genuinely bind, and it needs concealment, the power to keep its own moves out of sight so they cannot be bought, coerced, or read in advance. Both, whenever they have been had at scale at all, came bundled with authority or payment, the very things that drain the play. So the question narrows to whether order and concealment can be had on their own, without the cold tools that always came attached. For the whole of history nothing could, because order came from a person or a payment, and rules did not keep themselves.

04 ·

Cryptography: A Circle No One Owns

For the first time, we have the parts to build a magic circle that no one owns. They come from cryptography, two pieces of it in particular, one that can supply such a circle its order and one that can supply its concealment, neither resting on a person or a payment. On their own, each is a clever piece of engineering. Set together, they are something new, a machine that can hold a magic circle open at scale, the order kept by one half and the concealment by the other, with the cold tools left out of both. It is worth taking each half in turn before seeing what they amount to.

The order comes from Ethereum. Strip away the noise around it and what remains is a shared computer that no one owns and no one can quietly reach into, running whatever rules are written onto it exactly as written, for anyone who comes, without asking who they are or whose side they take.17 Place a rule on it and the rule simply runs. There is no operator behind a curtain to be leaned on for an exception, no owner to be bribed into looking away, no authority who can decide that this once the outcome should be different. The thing authority used to supply, order and the force that made rules actually bind, comes here from a machine that holds the rules while holding no opinion about them, a referee that is not a person and so cannot be intimidated or bought or worn down, because it is nothing but the rules themselves carried out in the open where anyone can check them. This is what credibly neutral means: it treats everyone alike because it has no capacity to do otherwise, with no will inside it that could decide to favour one of them, and a fairness that cannot be revoked is a different ground to build on than one resting on the goodwill of whoever holds the office.18

The neutrality is approximate, not absolute. The rules still had an author, so the authority that once stood over the game has not vanished so much as collapsed into the ruleset and set hard, spent once in the writing and made enormously costly to spend again by hand. Beneath the machine a diffuse social layer remains, and when the stakes ran high enough it has re-coordinated to override the very rules it was meant only to run, as it did in 2016, when a theft draining a flagship contract was reversed by a coordinated fork. It is tempting to read that as the machine failing exactly when it mattered, the human hand returning the moment something serious was at stake. But to override the rules that year was to amend a constitution, the rarest and most visible act a community can take, a people binding itself the way Ulysses bound himself to the mast and able to come loose only by an extraordinary, costly, collective effort.19 An office is captured quietly, by whoever draws close enough to work it; a constitution is changed in the open, at a price so high it has happened once. And the record since is the proof, for when a comparable sum was later frozen in the Parity wallets and the same kind of rescue was proposed, the community refused it, the vote falling against the intervention. High stakes were needed to move the social layer and were not enough, which is what a thinner authority should look like. Set against the only fair baseline, every institution we have ever had, an order set aside once, visibly, at great cost, with the dissenters free to walk to a chain of their own, is thinner than all of them.20 That is far less than the evangelists promise and far more than nothing, and it is enough.

This matters for play in a way easy to miss. Authority does more to play than punish it. It stands over the game as a will that can overrule it, a presence the players must defer to and second-guess, so the order is never quite theirs. Let the order come instead from a protocol that simply runs and no one is standing over anyone. The rules hold, disputes settle, the game binds, and there is no master in the room whose mood has to be managed. We keep the part of authority we needed, order that holds itself, and lose the part that poisons play, a person the players are made subject to. For the first time the binding can come from somewhere other than a boss.

That gives the circle its order, and Ethereum's openness is part of why it works, a public space anyone can audit and no one can quietly alter, which is a real strength and not a flaw to be apologized for. But a public record is not yet a circle. Play also needs concealment, the power to choose what is shown and what stays out of sight, because so much of play turns on what is not yet revealed, the move held back, the choice made in private, the information that must stay hidden until the moment it is meant to surface. A game where every move is visible also leaves open the oldest way of working it from outside, the buyer who pays for your choice and can check that you made it, the one who leans on you and can confirm you complied, both of which need only that your move be seen. What play wants from cryptography is a capability set on top of that public space: the power to reveal selectively.

That power is what recent cryptography unlocks, and it arrives in pieces a builder can assemble. A zero knowledge proof lets you prove a statement true while revealing nothing beyond its truth, so you can show you are eligible to take part without showing who you are, or that you played by the rules without revealing the move you made. A commit-reveal scheme lets you lock a choice in now and open it only later, so a whole field of players can move at once with none of them reading the others first.21 These are general primitives, building blocks rather than finished things, and the mechanisms worth building are assembled out of exactly these, selective revelation arranged into whatever shape a particular game needs. Laid over Ethereum's public space, they give a circle its concealment without costing it any of its order, the private move and the hidden tally sitting on top of a record that anyone can still check. The word to be careful with is hidden, which promises more than the parts deliver. What they buy is unprovability, a smaller and more exact thing than invisibility, since the art of the receipt-free ballot was always about refusing a proof rather than concealing an act. The buyer who once paid for a choice he could verify now pays for one he cannot, and the coercer who needed to confirm that you complied is left with nothing to confirm; the pressure that has to see a move before it can bend it loses its hold. What it does not yet close is the pressure that keys on the result rather than the move, the bribe that pays whichever side wins, the custodian who simply votes the key you handed him. These belong to a younger frontier, and the work against them, the receipt-free ballot, coercion resistance, the anti-collusion machinery being built now, is real and unfinished.22 The concealment the circle needs is here against the oldest way of working it from outside, and not yet against every one, which is more than has ever been had and less than the whole.

There is a younger frontier past this worth marking. Where a zero knowledge proof lets us prove things about hidden data, fully homomorphic encryption points toward computing over data that stays hidden the whole way through, adding up votes that are never decrypted, running a whole tally inside the encryption and revealing only the result.23 It is slower and heavier than anything I would build on today, but it says the room we are clearing, where coordination runs without anyone seeing the private facts that feed it, is larger than zero knowledge alone would make it.

Set the two halves together and the shape of the thing comes clear. The order from Ethereum and the concealment from cryptography are exactly what a magic circle needs to stand, and here they are, supplied for the first time by neither a person nor a payment. That is the answer to the question the last section left open: a circle can be built and held at scale without the cold tools that always broke it. Huizinga gave us Homo Ludens, the creature that plays; what stands in front of us now is the beginning of its counterpart, a machine that lets play run, machina ludens, Ethereum holding the order and cryptography the concealment, the work the cold tools used to do now discharged by an apparatus that does it without the human will that made it poisonous. This is what I mean by resisting incentive and authority, though resist almost overstates it, since there is no temptation being held off and no willpower being spent. We simply stop needing them. The order, the enforcement, the verification, the work those tools used to do, is largely taken over before we begin, and so incentive and authority can be left out of the design with their work already finished before we start. A machine for play, running anywhere, owned by no one. The only thing it asks of us is the one thing the scaffolding never let us reach: to make the thing worth playing.

05 ·

The Empty Corner

The triangle from earlier gives us a place to put all this, the cold pair along its base and play alone at the apex, a map of where coordination draws its pull. The question is where the thing we can now build would sit on it, and the honest way to answer is to start with what is already there.

The mechanisms we have lately built on crypto have not moved us off that edge as far as their builders believe. A prediction market sits almost exactly on the incentive corner, since the money is its whole substance and the belief that wins is the one with the most capital behind it. Futarchy is that corner aimed at governance, a market told to decide policy, inheriting every cold property of the thing it is made from, so that the policy that wins is not the best one but the one the market is cheapest to move, the measure made a target the instant it is set to decide.24 Quadratic funding and the DAOs creep a little along the base, the DAOs especially, where a token-weighted vote is only capital wearing the costume of a vote, the same old authority reached now by purchase.25 Each is real and each works, and each lives on the cold base, built with the cold tools though the ground beneath them had stopped requiring it.

The coordination triangle A triangle with play at the apex and authority and incentive at the base. Every deployed crypto coordination mechanism clusters on the cold base; the play corner held at real weight is empty. Play Authority Incentive ? ← unoccupied 3 2 1 cold mechanisms: 1 prediction markets, 2 futarchy, 3 QF / DAOs play held at real weight, empty circle size = real weight at stake

So the corner I care about is the one no one is standing in: play held at real weight, where the pull still comes from wanting and yet something consequential rides on how it falls, the order holding without an authority and the game staying sealed against the bribe the whole way through. This is the corner that seemed impossible, weight and wanting at once, and the resolution is quieter than the worry. The wanting was never the machine's to supply. What it holds off are the two ways a played-for stake turns into a taken one, a neutral order so that no hand sets the stake to its own profit, and concealment so that no move collapses into a price someone can pay for. The wanting still has to come from the players; the machine only keeps the world from reaching in to take it. Honestly, you can still only put a question mark there. The cold mechanisms carry the weight and have surrendered the play; the games keep the play and have surrendered the weight; nothing yet holds both. The whole wager of this essay is that the machine has finally made that corner reachable, that the road in no longer has to pass through the cold pair and pay the old toll.

It would be easy to mistake a cleared road for a finished one, so let me be exact about what this shows. None of it makes play scale, and none of it makes a game good. It removes the reason the corner stood empty, that every path in drained the thing on the way, and that reason is gone. A corner newly reachable is still an empty corner. That nearly everything built on this machine so far has been aimed straight back at the incentive vertex, that the freest coordination tools ever invented have been spent overwhelmingly on the coldest and most extractive coordination imaginable, does not prove the play corner cannot be filled. It only tells us what the builders chose, and they chose the corner they already knew. The ludic corner held at weight has not been built, which is a different claim from saying it cannot be.

One temptation to refuse is the leap from holding the parts to holding the thing. The parts are necessary and newly here; whether they are sufficient is the open question, and the history of mechanism design is full of cases where every component lay ready and the target was impossible all the same.26 So it is worth asking the hard form directly: might play held at real weight, neutral and sealed against the bribe, be ruled out in principle and not merely left unbuilt? Nothing we know rules it out. The classic impossibilities all assume a domain that takes every comer with every preference and demand every good property at once, while a magic circle does neither, admitting only those who enter for the play's own sake and never asking for a flawless efficiency, so they do not directly bind it. Where theory speaks at all it leans the other way, the folk theorem showing that cooperation among patient players is sustainable rather than forbidden, though it will sustain a collusion as readily as a cooperation and so promises only that the door is not locked.27 And the strongest evidence is already on the table, the games standing as proof that intrinsically motivated coordination can be real, durable, and large. What is unproven is the conjunction, the weight and the wanting and the seal all at once, and the honest way to leave it is as a wager that could be lost, three claims set down plainly enough to fail: that concealment can hold against the bribe, that a stake can press hard and stay played for as long as it keeps its distance from survival, and that consequence can ride on a game with no office over it and no reward to game.28 Each can be attempted, and each can be shown false.

And set honestly against the breakdown all around us, coordination failing at planetary scale and faster than we can think, the claim here is small.29 One mechanism does not mend that, and ludic coordination is not its cure. What I am claiming is only that a door long held shut has come open, that it is worth walking through, and that walking through it would be one attempt at one class of question rather than an answer to the whole emergency. Whether play can carry real weight and stay play while it does is unproven, and will stay unproven until someone walks into that empty corner and holds it.

06 ·

Walking In

I have been describing the corner from the outside, and I should say that I am trying to walk into it. The attempt is called Doxa, an open-ended design I am still working out rather than a thing I have finished and can hand you, and it starts from the question the cleared road leaves behind, where the wanting is supposed to come from once the machine has done its part. The answer I keep arriving at is the oldest free pull we have, the urge to take a side and to win others over to it. Tribalism is usually counted among the defects of our reasoning, and I am treating it instead as a ludic instrument, a contest people will enter for nothing but the pleasure of it, and Doxa is an attempt to turn that contest into an honest reading of belief on the kind of contested question that no fact will settle and no show of hands would mean anything about.

It works, in plain terms, like this. A question goes up that people genuinely disagree about, and you join by picking the side you already hold, because wanting your side to prevail is its own reward and that is what brings you in unpaid. Then everyone argues across the divide, each side trying to pull people from the other across to theirs, while your actual move, whether you held firm or were won over, stays sealed by cryptography and the running tally is kept dark, so that nobody can pay you to switch and then check that you obeyed. When the arguing ends the state is sealed and the verdict is read, and here is the choice that keeps Doxa off the cold base: the winner is the side that actually drew its opponents across, and not the side with the most heads or the most capital, since your own side's noise is yours to buy while your adversaries' concessions are not yours to manufacture. The weight is real because the verdict has consequences in the world, and yet nothing pays a player, the only thing won being clout, a standing that is soulbound, unsellable and uncashable, kept deliberately worthless as money so that it can stay a measure of standing, because the instant it could be cashed the stakes would close on a player's survival and the circle would break in the way I described earlier.

How Doxa reads belief A left-to-right flow of the mechanism. A contested question is posed; participants take a side, drawn by the free pull of taking sides; they argue while each move stays sealed by commit-reveal; the system measures which side pulled its opponents across the divide; the side that drew more of its opponents wins the verdict, and the only reward is soulbound, non-cashable clout. The winner is read from cross-faction movement, not from headcount and not from capital. Doxa reads the winner from which side pulled its opponents across ? a question take sides tribal pull argue, sealed commit-reveal measure who crossed opponents cross verdict + clout soulbound clout colours mark two sides here, not the warm and cold axis

I came to this the hard way. An earlier version of Doxa was itself a market, beliefs staked with capital and scored against the crowd, and it taught me why this one cannot be built that way, since the moment a market like that publishes its consensus as a public number, anyone can place a bet outside it that pays out on where the number lands, and nudging the number even slightly becomes worth more than it costs to move it, at any size of stake, which is a hole I could not close from the inside. So this design keeps the trustless and concealed parts and throws out the money and the market that made the answer worth corrupting. What is left can do something a market never could, which is improve the beliefs before it reads them, since a price can only weigh the convictions that walk in the door, while Doxa lets people move each other inside the sealed circle and reads the result only after they have, the persuasion doing the work the cryptography merely keeps clean.

I will not pretend the seal is total. Hiding each move kills the bribe that has to verify your compliance, and it does nothing against the bribe keyed to the public result, the offer to pay whichever side wins, which the dark tally and the cross-faction scoring shrink without closing. Hiding the content of a move does not hide its timing or its author or even its style either, which is a real way back to a name, and I leave both of these open rather than paper over them. So Doxa is one attempt at one class of question, eliciting genuine belief without an authority running the room or capital buying the answer, and I am building it alone and slowly.

It is strange to end an argument about coordination by pointing back at games, but that is where this has led. The same pull that makes someone give years to a world that pays them nothing may be one of the few forces left that can hold us together without commanding us or buying us, and the thing that always kept it small has, for the first time, fallen away. I cannot say what gets built in the space that opens up, or whether anything does, only that the space is finally there. The impulse to fill it is one most of us have felt since we were children, in the games we never had a reason to take seriously until now.

07 ·

References

  1. Johan Huizinga, Homo Ludens: A Study of the Play-Element in Culture (1938; Eng. trans. 1949). The term “magic circle” appears there only in passing; it was established as the working term of game studies by Katie Salen and Eric Zimmerman, Rules of Play: Game Design Fundamentals (MIT Press, 2003). ↩
  2. On payment as the enemy of a freely given act, Richard M. Titmuss, The Gift Relationship: From Human Blood to Social Policy (1970). ↩
  3. John R. Searle, The Construction of Social Reality (Free Press, 1995), on status functions and the collective assignment of value, with money as his own paradigm case. ↩
  4. Roger Caillois, Man, Play and Games (Les jeux et les hommes, 1958; trans. Meyer Barash, 1961). Caillois’s corruption runs outward, play’s forms leaking into ordinary life; the essay locates the same mechanism at the circle’s intake. ↩
  5. Bernard Suits, The Grasshopper: Games, Life and Utopia (1978), for the lusory attitude and the definition of game-playing as the voluntary attempt to overcome unnecessary obstacles. ↩
  6. Richard M. Ryan and Edward L. Deci, “Self-Determination Theory and the Facilitation of Intrinsic Motivation,” American Psychologist 55, no. 1 (2000): 68–78, and Edward L. Deci, Richard Koestner, and Richard M. Ryan, “A Meta-Analytic Review of Experiments Examining the Effects of Extrinsic Rewards on Intrinsic Motivation,” Psychological Bulletin 125, no. 6 (1999): 627–668, on integrated versus controlling regulation; Mark R. Lepper, David Greene, and Richard E. Nisbett, “Undermining Children’s Intrinsic Interest with Extrinsic Reward,” Journal of Personality and Social Psychology 28, no. 1 (1973): 129–137; Bruno S. Frey and Reto Jegen, “Motivation Crowding Theory,” Journal of Economic Surveys 15, no. 5 (2001): 589–611; Uri Gneezy and Aldo Rustichini, “A Fine is a Price,” Journal of Legal Studies 29, no. 1 (2000): 1–17; Mihaly Csikszentmihalyi, Flow: The Psychology of Optimal Experience (1990). ↩
  7. Clifford Geertz, “Deep Play: Notes on the Balinese Cockfight,” Daedalus 101, no. 1 (1972): 1–37; Erving Goffman, “Where the Action Is,” in Interaction Ritual (1967); Eric Brymer and Robert Schweitzer, “The search for freedom in extreme sports: A phenomenological exploration,” Psychology of Sport and Exercise 14, no. 6 (2013): 865–873. ↩
  8. Charles E. Lindblom, Politics and Markets: The World’s Political-Economic Systems (Basic Books, 1977), on authority and exchange as coordination mechanisms. ↩
  9. George J. Stigler, “The Theory of Economic Regulation,” The Bell Journal of Economics and Management Science 2, no. 1 (1971): 3–21, on regulatory capture. ↩
  10. Charles Goodhart (1975), as formulated by Marilyn Strathern, “‘Improving ratings’: audit in the British University system,” European Review 5, no. 3 (1997): 305–321: when a measure becomes a target, it ceases to be a good measure. ↩
  11. William G. Ouchi, “Markets, Bureaucracies, and Clans,” Administrative Science Quarterly 25, no. 1 (1980): 129–141; Walter W. Powell, “Neither Market Nor Hierarchy: Network Forms of Organization,” Research in Organizational Behavior 12 (1990): 295–336; Alan Page Fiske, “The Four Elementary Forms of Sociality,” Psychological Review 99, no. 4 (1992): 689–723; Yochai Benkler, “Coase’s Penguin, or, Linux and The Nature of the Firm,” Yale Law Journal 112 (2002): 369–446. ↩
  12. Elinor Ostrom, Governing the Commons: The Evolution of Institutions for Collective Action (Cambridge University Press, 1990), on self-generated, self-policed rules; Robert Axelrod, The Evolution of Cooperation (Basic Books, 1984), on reciprocity under a long shadow of the future; Martin A. Nowak and Karl Sigmund, “Evolution of indirect reciprocity,” Nature 437 (2005): 1291–1298, on reputation. ↩
  13. Natasha Dow Schüll, Addiction by Design: Machine Gambling in Las Vegas (Princeton University Press, 2012); José P. Zagal, Staffan Björk, and Chris Lewis, “Dark Patterns in the Design of Games” (Foundations of Digital Games, 2013); Ian Bogost, “Persuasive Games: Exploitationware” (2011); Aaron Drummond and James D. Sauer, “Video game loot boxes are psychologically akin to gambling,” Nature Human Behaviour 2 (2018): 530–532; Daniel L. King and Paul H. Delfabbro, “Predatory monetization schemes in video games,” Addiction 113, no. 11 (2018): 1967–1969; Tim Wu, The Attention Merchants (2016); Brian Sutton-Smith, The Ambiguity of Play (Harvard University Press, 1997). ↩
  14. Abraham Wald, A Method of Estimating Plane Vulnerability Based on Damage of Survivors (Statistical Research Group, 1943), the founding case of survivorship bias as a selection effect. ↩
  15. “Battle of B-R5RB” (EVE Online, 2014) and CCP Games’ player-elected Council of Stellar Management; Seth Cooper, Firas Khatib et al., “Predicting protein structures with a multiplayer online game,” Nature 466 (2010): 756–760, and Firas Khatib et al., “Crystal structure of a monomeric retroviral protease solved by protein folding game players,” Nature Structural and Molecular Biology 18 (2011): 1175–1177; Edward Castronova, Synthetic Worlds (University of Chicago Press, 2005). ↩
  16. Elinor Ostrom, “Beyond Markets and States: Polycentric Governance of Complex Economic Systems,” American Economic Review 100, no. 3 (2010): 641–672, and Governing the Commons (1990), design principle 8 on nested enterprises; Robin I. M. Dunbar, “Coevolution of neocortical size, group size and language in humans,” Behavioral and Brain Sciences 16, no. 4 (1993): 681–735, on the cognitive ceiling on group size. ↩
  17. Vitalik Buterin, “Ethereum: A Next-Generation Smart Contract and Decentralized Application Platform” (Ethereum white paper, 2014), on the ownerless world computer that runs its contracts as written. ↩
  18. Vitalik Buterin, “Credible Neutrality As A Guiding Principle” (2020). ↩
  19. Jon Elster, Ulysses and the Sirens (1979) and Ulysses Unbound (2000), on precommitment and self-binding; James M. Buchanan and Gordon Tullock, The Calculus of Consent (1962), on the distinction between constitutional and in-period choice. ↩
  20. The 2018 vote on the Parity wallets frozen by the EIP-999 proposal, in which a comparable loss failed to move the social layer; Vitalik Buterin, “Don’t Overload Ethereum’s Consensus” (2023), on using social consensus sparingly; Thomas C. Schelling, The Strategy of Conflict (Harvard University Press, 1960), on focal points. ↩
  21. Shafi Goldwasser, Silvio Micali, and Charles Rackoff, “The Knowledge Complexity of Interactive Proof Systems” (STOC 1985), the origin of zero-knowledge proofs; Manuel Blum, “Coin Flipping by Telephone: A Protocol for Solving Impossible Problems” (CRYPTO 1981; ACM SIGACT News 15, no. 1, 1983: 23–27), the origin of the commit-reveal scheme. ↩
  22. Josh Benaloh and Dwight Tuinstra, “Receipt-Free Secret-Ballot Elections” (STOC 1994); Ari Juels, Dario Catalano, and Markus Jakobsson, “Coercion-Resistant Electronic Elections” (WPES 2005); Vitalik Buterin, “On Collusion” (2019); the Minimal Anti-Collusion Infrastructure (MACI), proposed by Buterin and built by the Ethereum Foundation’s Privacy and Scaling Explorations team (2019); Philip Daian, Tyler Kell, Ian Miers, and Ari Juels, “On-Chain Vote Buying and the Rise of Dark DAOs” (2018). ↩
  23. Craig Gentry, “Fully Homomorphic Encryption Using Ideal Lattices” (STOC 2009), the first working scheme; the problem was first conjectured by Ronald Rivest, Leonard Adleman, and Michael Dertouzos, “On Data Banks and Privacy Homomorphisms” (1978). ↩
  24. Robin Hanson, “Shall We Vote on Values, But Bet on Beliefs?” (working paper 2000; Journal of Political Philosophy 21, no. 2, 2013: 151–178), the futarchy proposal. ↩
  25. Vitalik Buterin, Zoë Hitzig, and E. Glen Weyl, “A Flexible Design for Funding Public Goods,” Management Science 65, no. 11 (2019): 5171–5187, the quadratic-funding mechanism; on token-weighted voting as plutocracy, Vitalik Buterin, “Moving beyond coin voting governance” (2021). ↩
  26. Kenneth J. Arrow, Social Choice and Individual Values (1951); Allan Gibbard, “Manipulation of Voting Schemes: A General Result,” Econometrica 41, no. 4 (1973): 587–601; Mark A. Satterthwaite, “Strategy-proofness and Arrow’s Conditions,” Journal of Economic Theory 10, no. 2 (1975): 187–217; Roger B. Myerson and Mark A. Satterthwaite, “Efficient Mechanisms for Bilateral Trading,” Journal of Economic Theory 29, no. 2 (1983): 265–281. ↩
  27. Drew Fudenberg and Eric Maskin, “The Folk Theorem in Repeated Games with Discounting or with Incomplete Information,” Econometrica 54, no. 3 (1986): 533–554. ↩
  28. Imre Lakatos, “Falsification and the Methodology of Scientific Research Programmes,” in Criticism and the Growth of Knowledge, ed. Lakatos and Musgrave (Cambridge University Press, 1970), on progressive versus degenerating research programmes. ↩
  29. Scott Alexander, “Meditations on Moloch” (Slate Star Codex, 2014), on coordination failure as a multipolar trap. ↩