Published May 2, 2026
Prediction markets for the other 90%. Gaming, not gambling. Reputation, not speculation. Intelligence, not extraction.
1
The Wrong Path
Prediction markets are no longer a fringe curiosity. Weekly volumes have climbed into the billions. Growth over the past two years has been explosive. Kalshi and Polymarket are no longer niche products for internet obsessives — they are becoming serious financial and cultural objects. The market has validated one thing beyond dispute: people want to express beliefs about the future in a structured, tradable way. The question is no longer whether prediction markets matter. The question is what kind of category they will become.
The dominant path is already visible. And it should make us uneasy.
What began as an elegant mechanism for aggregating dispersed information is drifting toward something narrower and more extractive. Bots increasingly sit inside the system, arbitraging mispricings faster than ordinary humans can react. Very short-duration contracts reward reflex over judgement. AI agents are beginning to crowd the human out from the very category that was meant to reveal collective human wisdom. Problem gambling rates among 18-to-25-year-olds are rising. And the surrounding culture is moving in the direction that all money-first systems tend to move: from forecasting to trading, from trading to adrenaline, from adrenaline to addiction.
Source: “Prediction Markets: The New Signal”, Social Capital, February 2026
That is not an accidental side-effect. It follows from the design. Once real money becomes the core stake, the centre of gravity shifts. The product starts serving the people most willing to risk capital, most willing to obsess over micro-movements, and most willing to treat every opinion as a position. The category reaches the 10% and increasingly optimises around extracting from them.
Kalshi’s endgame has been stated plainly: financialise everything, turn any difference of opinion into a tradable asset. That is one destination. It is coherent. It is also a vision in which more and more of everyday human judgement gets pulled into a market logic built for speculation.
But that is not the only path prediction markets can take.
There is another possibility: prediction not as an extractive financial product, but as a mass-participation system for judgement, reputation, and learning. Not a place where the most capitalised players dominate, but a place where ordinary people can build a record of calibrated thinking. Not a casino wrapped in the language of truth-seeking, but a social product where consequence comes from effort, identity, and public memory.
This is where WePredict begins. Not with the 10% who want to risk money, but with the other 90% who do not. Not with speculation, but with participation. Not with extraction, but with compounding intelligence.
Gaming, not gambling. Reputation, not speculation. Intelligence, not extraction.
| Path | How it works | Who it reaches | Where it leads |
| Path 1 — Real money | Kalshi, Polymarket. Cash staked on outcomes. | The 10% willing to risk capital | Extraction, casino drift, bot dominance |
| Path 2 — Free play-money | Manifold, Metaculus. Chips handed out freely. | Broad in theory, hollow in practice | No consequence, no identity, stays niche |
| Path 3 — WePredict | Earned Mu + Predictor Score + social Circles. | The other 90% | Participation, reputation, intelligence |
Further Reading: The WePredict Reader: A Summary of 10 Essays
2
The Third Path
WePredict is the third path. It combines earned play-money with public reputation and social consequence. None of the three is sufficient alone. Together, they create something neither of the first two paths achieves: broad participation with genuine stakes.
Mu: earned, not given
Mu is WePredict’s currency. It cannot be purchased by individuals. It must be earned through daily engagement — opening NeoMails, completing Magnets, maintaining streaks, showing up. A Mu balance of 3,000 is not a sign-up bonus or a free pile of chips. It is a record of attention, consistency, and time.
When someone stakes Mu on a prediction, they are not spending something they got for nothing. They are spending something they accumulated through repeated participation. That changes the psychology entirely. A balance represents weeks of showing up. Spending it feels consequential because it cost something real.
This is the principle most play-money systems missed. They treated currency as a convenience. WePredict treats it as earned scarcity. Free chips create play. Earned chips create stake.
The earned-only model is also legally correct under India’s Regulation of Gaming Act 2025. Purchased Mu that can be lost through chance counts as gambling. Earned Mu is clean. The integrity argument and the compliance argument are the same argument. Mu being earned is not a limitation. It is the point. A crowd whose participation was purchased is a polluted signal. A crowd whose currency was earned is not.
Predictor Score: reputation that compounds
Predictor Score lives in WePredict Public. It is a persistent, public record of forecasting accuracy built on Brier-score-style calibration logic. The point is not to reward dramatic certainty. It is to reward honest probability, expressed consistently over time.
The Score is designed to be slow. It exists from day one, but it takes months to become meaningful — and that is precisely the point. A Score built across 400 markets over 18 months is worth more than one built across 20 markets in a week. Time is built into the architecture. A high Predictor Score cannot be bought, manufactured, or shortcut. It can only be earned through calibrated forecasting, repeated over time, in public.
The closest analogy is a chess rating. It compresses an entire history of play into a number that took time to build and can be damaged by a careless period. Unlike a chess rating, the Predictor Score also travels — it follows the participant across every public market they enter, and can be displayed inside private Circles as a signal of broader credibility. Inside a Circle, the leaderboard shows who called it right among friends. Alongside it, the Predictor Score shows who has built a record of calibrated judgement in the wider world.
Circles: social consequence without cash
WePredict Private runs inside closed Circles — WhatsApp groups, family chats, office Slack channels, cricket groups, alumni networks. People already predict together in these rooms. They argue, boast, tease, and remember. What they lack is a system that turns those habits into a persistent record.
Being wrong in front of people who know you is real consequence even when no money changes hands. Losing a prediction while a friend wins on the same market, in a group that watched both of you, is genuinely felt. The social frame is what makes play-money serious. The group is the room. The leaderboard gives it memory.
**
Earned scarcity through daily engagement. Reputation that compounds publicly in markets. Social consequence inside Circles where being wrong in front of people who know you is genuinely felt. Free chips create play. Earned chips create stake. That is the design.
3
WePredict Private
The most immediate business for WePredict is not the giant public market. It is the smaller, more intimate, more repeatable room: WePredict Private.
Private runs inside Circles. A Circle can be almost anything — a family WhatsApp group, a hostel floor, a cricket gang, a college alumni chat, a product team in Slack, a company offsite cohort. The insight is simple: people already predict together. They argue, boast, guess, and remember. What they have never had is a system that turns those habits into a persistent scoreboard. WePredict Private supplies that missing ledger.
Two modes
Poll mode requires no Mu. Participants vote before the outcome is known, the result resolves, and the Circle leaderboard updates with a permanent record of who said what. No prior balance needed. No prerequisite. Just a poll — with a receipt. That phrase captures it precisely: the receipt is what WhatsApp’s native poll has never provided. The group finally has memory. Who called it, when, and whether they were right. Predictor Score does not update in Poll mode — that belongs to Public markets. But the leaderboard compounds from day one, and that is enough to make the habit form.
Basic market mode uses parimutuel mechanics and requires Mu. Everyone stakes into a shared pool; correct predictors split it at resolution. No market maker is needed. No hidden liability is created. The rules are transparent and intuitive — structurally identical to how informal office pools already work. Parimutuel is the right mechanism for private groups: clean, fair, and exposure-free.
LMSR (Logarithmic Market Scoring Rule) — the Advanced mechanism — is not available in any Private tier. LMSR requires a subsidised market maker. In a private Circle, that role falls to the admin. The liability is invisible, potentially unbounded, and entirely wrong for a casual group running a cricket market. Parimutuel removes that exposure completely.
Three tiers
All tiers are priced directly — no intermediate currency, no complexity.
| Tier | Who | What’s included |
| WePredict Private | Any group — free | Poll mode and Basic market mode, capped Circle size. The entry product. |
| WePredict Private Plus | Individuals — subscription | Larger Circles, Predictor Score display, richer analytics, custom market categories, streak tracking. |
| WePredict Private Enterprise | Companies — annual subscription | Slack integration, calibration dashboards, admin controls, anonymity settings, audit trail. |
Enterprise Mu distribution
Enterprises can purchase Mu in bulk and distribute it to employees or customers — seeding participation in their Circles without requiring individuals to earn it first. The enterprise pays WePredict directly; recipients use the Mu exactly like earned Mu, in the same single wallet, with no distinction.
The individual never purchased Mu with their own money, so India’s Gaming Act exposure does not apply. The enterprise is buying engagement infrastructure, not enabling gambling. The analogy is airline miles: the airline sells miles to a card company; the cardholder spends them without having bought them directly.
Enterprise Mu distribution is a separate B2B arrangement — distinct from the Private Enterprise subscription.
**
Private is the near-term business because it solves three problems at once. Poll mode solves cold start — anyone can play immediately, no Mu required. Parimutuel keeps the mechanism clean — no admin liability, no complexity. The leaderboard gives groups a memory they have never had. A poll with a receipt.
4
WePredict Public
If WePredict Private is the room where forecasting becomes social, WePredict Public is the arena where forecasting becomes credentialed.
One mechanism: LMSR, managed by WePredict
Public markets run on LMSR only — the Advanced mechanism — with WePredict itself acting as the professional market maker. This is precisely where LMSR belongs. In Public, the liquidity pool is centrally managed, the mechanics are understood, and the liability is visible and deliberate. What would be dangerous in the hands of a casual Circle admin is manageable — and powerful — when handled by the platform.
Earned Mu is the mandatory entry ticket. No purchased Mu. No free public participation. Everyone in the public arena earned their way in through daily engagement. That preserves both the legal cleanliness and the integrity of the signal.
No spread. No rake. No house edge.
Real-money prediction platforms monetise participation through trading fees and spreads. WePredict Public does the opposite. The market maker role is a service — providing liquidity and price discovery — not an extraction point. No spread, no rake, no house edge. Fair pricing, always. Users understand immediately that the platform is not skimming every act of judgement. That is a real competitive advantage, and a clean one.
Where Predictor Score becomes a credential
WePredict Public is where the Predictor Score becomes externally validated at scale. Inside a private Circle, your leaderboard rank matters socially. In Public, the Predictor Score is tested against thousands of participants across a diverse and open pool — and it becomes something more portable.
The Score is slow by design. It exists from day one of launch, but takes months to become meaningful. That is not a weakness — it is the architecture. A Score of 82nd-percentile calibration across 400 markets over 18 months cannot be bought, faked, or shortcut. It can only be built through honest forecasting, repeated over time, in public. That compounding record is the credential.
The Score can also be displayed inside Circles — visible on the private leaderboard alongside group rankings. A participant’s broader public record becomes aspirational for Private-only users. The Score is the hook that pulls the habit of private prediction toward the discipline of public forecasting.
Revenue: premium features, not participation taxes
Monetisation in Public is direct and clean. Not through participation fees, but through subscriptions for premium features: deeper analytics, advanced leaderboard tools, historical calibration views, enhanced profile tools, and professional-grade reputation surfaces. These are not chance-based. They carry no gambling logic. They are straightforward software revenue layered atop a public forecasting identity system.
| Stream | Product | Mechanism |
| Near-term | Private Plus | Individual subscription |
| Near-term | Public premium features | Individual subscription |
| 6–12 months | Private Enterprise | Annual B2B contract |
| Medium-term | Enterprise Mu distribution | B2B bulk purchase — legal confirmation needed |
**
WePredict is not a prediction market with social features added. It is a social market — a new category where reputation is the currency, public accountability is the social mechanic, and the intelligence produced is the business model. It is a system in which attention, earned currency, social consequence, and compounding reputation are built into the core. Private gives groups a memory. Public gives individuals a credential. Mu connects both. Prediction markets for the other 90% is not a positioning statement. It is a design decision — made at every layer of the product, the economics, and the law.
5
Red Team: Will It Actually Work?
Every monetisation model looks clean on paper. The honest test is whether it survives contact with the sceptic. Here are the four hardest questions about WePredict’s revenue model — stated as the sceptic would state them, with the best current answers.
Will Private Plus actually convert — or will everyone stay free?
The sceptic’s case: nobody pays for tools that make a friend group slightly more fun. Poll mode already gives the group the key novelty — the receipt, the memory, the leaderboard. Basic parimutuel gives them the second novelty — a stake, a pool, a payout in Mu. Once those two modes exist in the free tier, most Circles may simply settle there. They may like the product, use it, even talk about it, and still never pay. That is what consumer software does to attractive light-premium products all the time: it turns them into high-usage, low-revenue habits.
The honest answer is that Plus converts only if the free tier is deliberately incomplete in one specific dimension: social scale. Not feature deprivation in general. Circle size. If the free product supports the natural size of the group most people care about, they will stay free. The conversion trigger cannot be “better analytics” or “custom categories” alone. Those are nice extras. They are not the buying moment. The buying moment comes when the group outgrows the room. The monetisation works only if the cap lands just below the size at which a real group becomes meaningfully alive — close enough that growth creates genuine pressure, not so low that the free product feels artificially broken. That calibration cannot be assumed. It has to be tested in the first cohort of active Circles before Plus is commercially launched.
Will Private Enterprise sell without a proven consumer product first?
The sceptic’s case: enterprise buyers want proof that the thing already works in the wild. Without consumer traction, the sales conversation starts from zero credibility. Enterprise risks becoming an idea sold through decks rather than a behaviour bought through conviction.
The honest answer is that Enterprise can sell early, but only as narrow software with a narrow job: internal forecasting, team calibration, Slack-native prediction rituals, auditable group participation. Sales teams forecasting quarter close. Product teams forecasting launch outcomes. Leadership teams recording assumptions before decisions. That is enough to begin. But it also means Enterprise’s early market is much smaller than the full vision suggests. If the consumer side remains weak, Enterprise does not collapse — but it loses narrative power and much of its long-term upside. Consumer traction first, enterprise sales second is the right sequencing — not because Enterprise cannot be sold earlier, but because the conversations will be shorter and the conversions faster once Private has demonstrated that groups actually use it.
Is Enterprise Mu distribution a real revenue stream or a theoretical one?
The sceptic’s case is that the idea is seductive but fragile. There are three ways it can fail simultaneously.
First, the buyer may not care enough. Enterprises already have many ways to incentivise behaviour — cash rewards, vouchers, points, badges, internal recognition. Enterprise Mu becomes interesting only if WePredict itself is already a meaningfully differentiated environment. If the platform is not magnetic, Mu is just another internal token, and internal tokens are notoriously weak businesses.
Second, recipients may not treat distributed Mu with the same seriousness as earned Mu. The philosophical architecture rests on earned scarcity. Enterprise distribution weakens that. A person spends gifted Mu differently from earned Mu, even when the wallet is technically identical. The psychology of effort and the psychology of receiving are not the same.
Third, enterprise distribution could create two economies inside one wallet — one built from effort, one injected from outside. The more commercially important Enterprise Mu becomes, the greater the risk that it dilutes the very integrity the product claims as its core advantage — a platform where Mu means effort, not entitlement.
The honest answer is that Enterprise Mu distribution is not yet a revenue stream. It is an option. It becomes real only if three things are true simultaneously: the platform already feels valuable, the enterprise has a concrete engagement use case, and distributed Mu does not visibly contaminate the meaning of earned Mu. Without all three, the idea remains conceptually attractive but commercially thin.
Will Public generate enough subscription revenue without a spread?
The sceptic’s case is simple: probably not, at least not early. Public forecasting products are hard to monetise even with intellectually serious users. Most users consume status and rankings passively. Only a minority pays for analytics. By refusing spread and participation taxes, WePredict walks away from the most direct monetisation lever in the category.
The honest answer is that Public probably does not work as a standalone near-term business. It works as a prestige layer, a credential layer, and a habit-deepening layer — but early on it is far more likely to be a retention engine than a profit engine. It gives Private users an aspiration. It gives Mu somewhere serious to flow. It gives the network a reputational summit. That is valuable. But it also means Public must initially be subsidised by Private revenues rather than expected to carry itself. Public earns the right to become a business. Private funds the journey there.
6
Making It Work: The Sequencing
The red team analysis implies a strict order of operations. Three things have to happen in the right sequence for the monetisation model to work.
Private revenue funds Public’s early life
The products launch together, and they should — the narrative is stronger when Private and Public visibly belong to the same system. But they are not equal from a business standpoint. Private is where the first monetisation pressure must land. It has clearer conversion triggers, simpler use cases, more immediate social loops, and more legible B2B packaging. Public, by contrast, is expensive dignity. It matters enormously to the system — it creates the Predictor Score, the credential layer, and the aspirational summit. But early on it is far more likely to deepen the product than to fund the company.
Treating both products as if they must monetise simultaneously is the mistake to avoid. Private pays the bills first. Public earns the right to become a business later. That sequencing has to be internalised before launch, not discovered six months in.
The Circle size cap is the most important commercial decision in the model
Not one of the important decisions. The most important. Analytics, custom categories, streak features, profile tools — all of these help. None creates the decisive conversion moment as reliably as a group hitting the wall of the free tier. The cap defines where social energy spills from “good enough for free” into “worth paying to continue.”
That means the cap cannot be chosen generously or aesthetically. It has to be empirically tuned against the first cohort of active Circles. Set it too high and Plus never converts — the free product is so comfortable that nobody upgrades. Set it too low and the free product never becomes socially alive enough to create a habit worth paying to continue. The cap is the commercial fulcrum of the near-term revenue model. Get it right and the self-serve funnel works. Get it wrong and WePredict becomes a lively free toy with no business underneath it.
The starting point, based on how Indian WhatsApp groups naturally size and behave, is 25 active members on the free tier — large enough that the product feels genuinely social, small enough that a real cricket gang, extended family group, or office team hits the ceiling naturally within weeks. The cap applies to active members, not total members: anyone who has voted or staked in the last 30 days. That keeps the pressure real and the upgrade fair.
Enterprise sales follow consumer proof points, not the other way round
The Slack use case for internal forecasting is real and partially independent of consumer scale. But enterprise sales cycles are long, and the first conversation sets the tone for all subsequent ones. A conversation that opens with active Circles, return rates, and leaderboard engagement is a different conversation from one that opens with a deck and a vision.
The discipline is to resist running enterprise sales in parallel with consumer launch. Use the first six months to build consumer proof points. Then open enterprise conversations with evidence rather than aspiration. The pipeline will be shorter to close, the contract values higher, and the product’s credibility in the room much easier to establish.
WePredict does not need every revenue stream to work at once. It needs one consumer monetisation loop that converts, one enterprise product that solves a narrow and real problem, and one public layer that compounds identity and aspiration without being forced to over-earn too early.
The architecture can be ambitious. The sequencing must be disciplined.
