There was a crowd on a New York street corner, which is not unusual. They were watching a World Cup match, which is also not unusual. The unusual part was the screen.

It was bolted to the side of a truck, and beside the live match it carried something no television broadcast has ever shown me, live betting odds, a scrolling set of prices, and a QR code promising $50 free to anyone who wanted to place a wager on the outcome. The whole rig belonged to Polymarket, a prediction market.

Nobody in that crowd knew there was a blockchain underneath it. Nobody needed to. There was no wallet to set up, no seed phrase to guard, no gas fee to puzzle over. There was a match, a number, and one tap.

I have spent years telling anyone who would sit still that crypto's real breakthrough would be payments. Standing on that corner, I understood that I had been watching the wrong screen. The technology I kept waiting to see arrive had already arrived. It had simply disappeared, folded so completely into an ordinary evening that the people using it could not have told you it was there.

So I will concede the point I resisted for a long time. The tech works.

What I want to argue in this piece is narrower and, I think, more important. A prediction market is a wager first, even when it can double as research. And calling that wager a "market" is doing quiet regulatory work that deserves to be said out loud.

What a prediction market actually is

Strip away the branding and a prediction market is a place to buy and sell contracts on whether a specific event will happen. Will Spain win the World Cup. Will a given candidate take a given state. Will a named team advance past the quarter-final.

Each question becomes a tradeable share.

The pricing is the elegant part, and the honest part. A share pays $1 if the event occurs and $0 if it does not. Between those poles it trades somewhere from $0.01 to $0.99, and that price reads directly as an implied probability. A contract changing hands at $0.62 is the crowd saying, with real money behind it, that the event is about 62% likely.

When the event resolves, winners collect $1 a share and losers collect nothing.

Resolution is where the machinery gets interesting. Someone, or something, has to declare what actually happened and pay out accordingly. On a crypto-native venue that job falls to an oracle, a mechanism for reporting a real-world outcome back to the contract so it can settle without a human clerk in the middle.

Get the oracle right and the market is trustworthy. Get it wrong, or leave it gameable, and every price above it is built on sand.

Liquidity is the last piece. A market is only as good as its ability to let you in and out near the quoted price, which is why the deep contracts, the ones with millions of dollars flowing through them, produce the sharpest numbers. The thin ones wobble on a single large order.

Take the tournament I had been watching. The final on July 19, 2026 was Spain against Argentina, and Spain won it. In the days before kickoff you could have watched the Spain share on Polymarket climb and dip as team news, injuries, and nerves washed through the crowd, each move a small referendum priced to the cent.

When the whistle went, the contract resolved, the oracle reported the result, and the shares paid out. That is the whole life cycle in one evening, a question, a price that breathes, a settlement, a payout.

None of this is new as finance. Event contracts have existed in one form or another for a very long time, and economists have written admiringly about them for decades. What is new is the plumbing, and who can reach it.

Why crypto rails suit them so well

If you were designing infrastructure for a global betting market from scratch, you might well end up building something that looks a lot like a public blockchain. The fit is almost embarrassingly good.

Consider what these contracts need. They need to be open around the clock, because a World Cup final does not wait for banking hours and neither does a crowd in a dozen time zones. They need to settle in a currency that behaves the same in Lagos, Lima, and Lisbon, which is exactly what a dollar stablecoin provides. They need to be reachable by anyone with a phone rather than gated behind a licensed brokerage account, which is what a permissionless network delivers by default.

And they benefit from letting users hold their own funds rather than parking them with an operator, so custody sits with the bettor until the moment of settlement. That is a genuine improvement on the model where a betting house holds your balance and you hope it stays solvent.

Put those properties together and you get a system that is global, always on, and cheap to settle. In my reading, this is the quiet reason prediction markets took off on crypto rails and not inside a traditional sportsbook. The rails were already built for exactly this shape of activity.

The part that unsettles me is the same part that makes it work. The better the rails get, the less the user sees them. On that street corner there was no crypto vocabulary anywhere in view. The blockchain had become as invisible as the electricity behind the screen.

That is a triumph of design. It is also, I would argue, a triumph of framing, because when the machinery vanishes so does the moment where a person might stop and ask what exactly they are doing.

The 2026 explosion

The numbers this year stopped being a curiosity and became a phenomenon. The single clearest data point sits at the top of the pile. Polymarket's World Cup winner market traded roughly $4Bn in cumulative volume by early July 2026, edging past the platform's 2024 United States presidential contract, which had itself been the largest market it had ever run (24/7 Wall St., July 2026).

A word of care on that figure, because it is easy to misread and often is. $4Bn is cumulative volume traded over the life of the market, every buy and sell added up. It is not the amount sitting live at any one moment. The active stake at any given instant is far smaller.

The headline captures churn, not a standing pile of money. That distinction matters, and I will come back to why.

The platform-level story is just as steep. Combined monthly volume across the two largest venues, Kalshi and Polymarket, climbed from under $5Bn in September 2025 to around $24Bn by April 2026 (The American Prospect, July 2026). That is not a trend line. That is a wall.

Line those numbers up, a presidential-election record matched by a single sporting market and a monthly base rate several times higher in a matter of months, and the shape is unmistakable. A year earlier, none of it would have sounded believable.

Zoom out from that one market and the sector-wide climb is just as steep. Independent trackers put combined Kalshi and Polymarket volume near $45Bn in June 2026 alone, and the Pew Research Center, documenting the same run, described the growth of recent months as dramatic (Pew Research Center, May 2026). A monthly pace in that range annualizes to well above $500Bn, many times the sector's 2025 level.

I would treat any single annual headline as an estimate, since trackers disagree on what to count, but the direction is not in dispute. In a single year this went from a thing crypto people talked about to a thing a man on a street corner taps into without knowing its name.

Money that big attracts attention, and not only from bettors. The same reporting that logged the $4Bn World Cup market noted that prominent figures, Donald Trump and Mark Zuckerberg among them, were now circling the sector, eyeing a slice of what these venues have become (24/7 Wall St., July 2026). When the powerful start wanting a piece, that is usually a sign the thing has stopped being a niche and started being an industry. It is also, reliably, the moment regulators wake up, which is exactly what happened next.

Market or gambling?

Here is where I have to be fair before I am critical, because the case for prediction markets is stronger than the skeptics usually admit, and it deserves its best form.

The steelman, a market that tells the truth

The serious argument for these venues is that they are information machines. A poll asks people what they think and hopes they answer honestly. A prediction market asks people to put money behind what they think, and money has a way of concentrating the mind. When being wrong costs you and being right pays you, you stop performing an opinion and start reporting your actual belief.

Aggregate thousands of those money-weighted beliefs into a single price and, the argument goes, you get a forecast that is hard to beat. Advocates point to a real track record of prediction markets calling elections, policy outcomes, and sports results more accurately, and earlier, than pundits or polls. The price moves the instant new information appears, because someone with an edge can profit by trading on it, and in profiting they fold their private knowledge into the public number.

A market like that is not idle chatter. It is, at its best, a live probability estimate for questions that matter, produced for free and available to everyone. That is a genuinely useful public good, and I am not going to pretend otherwise.

The rebuttal, the research is a by-product of the bet

Now the other side, which is mine. All of that can be true and the thing can still be, first and mainly, gambling.

The information is real, but it is a by-product. Nobody on that street corner tapped the QR code to contribute to the collective forecasting accuracy of humanity. They tapped it because there was $50 free and a match on, and because it felt good to have money riding on Spain.

The research value exists. It is just not why the money shows up. And a mechanism should be judged by what actually drives its volume, not by the most flattering description of its output.

This is why the word "market" bothers me. It is doing work. Call an activity a market and it inherits the whole respectable vocabulary of finance, prices, liquidity, participants, information.

Call the same activity betting and it inherits a very different set of associations and, crucially, a very different set of laws. The American Prospect made this point sharply, framing the regulatory debate as a set of word games in which the label determines the rules (The American Prospect, July 2026).

I think that framing is exactly right. The euphemism is not decoration. It is the strategy.

Recall the $4Bn figure and the care I took to label it as churn rather than a standing balance. That gap is the tell.

Real capital markets exist to allocate capital, to fund things that then produce a return. A wager on a football result allocates nothing. When it resolves, the money changes pockets and the event is simply over.

High churn and no underlying asset is the signature of a betting pool wearing a market's clothes. My view is that we should be honest about which one we are looking at, and this one is a betting pool.

What to watch from here

The interesting fight now is not technological. It is legal, and it is moving fast enough that anyone with a stake should be tracking specific signals rather than vibes.

Start with the federal rulemaking. The Commodity Futures Trading Commission put out a proposed rule running to roughly 267 pages that would carve a formal lane for sports event contracts. Under the proposal, contracts on final scores, point differentials, wins and losses, tournament advancement, and individual or season-long statistics would be permitted, while single-play bets, in-game injuries, physical altercations, officiating calls, and pre-collegiate sports would be off limits (Axios, June 2026). The line the agency is trying to draw, roughly, separates betting on an aggregate outcome from betting on a single discrete play, and where that line lands will shape the entire product category (ESPN, June 2026).

Then watch the states, because they are not waiting for Washington. Minnesota passed a first-in-the-nation ban on prediction markets, effective August 1, 2026, and Kalshi and Polymarket promptly sought a preliminary injunction to block it before United States District Judge Katherine Menendez (The American Prospect, July 2026). Kalshi is litigating parallel fights in more than a dozen states, and a live tracker of where these products are legal, contested, or banned is now a genuinely useful thing to keep bookmarked (CBS Sports, July 2026). The core tension is federal preemption versus state authority over gambling, and it is unresolved.

Finally, watch the integrity questions, because they are the ones that will age worst if ignored. Congressional Democrats have pressed the CFTC to rein in prediction-market sports betting and flagged the risk of insider trading, the worry being that someone with private knowledge of an outcome, a team's real fitness, a lineup not yet announced, can quietly trade on it (CNBC, April 2026). In a market that prides itself on absorbing information, the person with the best information may be the one you least want in the pool. If you want the fuller vocabulary behind all of this, from oracles to settlement, our glossary and the interactive DeFi Primer both lay it out plainly.

The bottom line

I keep coming back to that street corner. A screen on a truck, live odds, a QR code, $50 free, and a crowd that had no idea it was touching a blockchain and had no reason to care. For years I argued the breakthrough would look like payments. It looks like this instead, and I was wrong about the form even as I was right that the form would eventually vanish from view.

So let me be clear about where I land, because the point of writing this was not to sit on a fence. The technology genuinely arrived. That is not in doubt anymore, and pretending otherwise is just nostalgia.

What we owe each other is honesty about what it arrived as. A prediction market is a wager first. It can double as research, and sometimes the research is even good. But the volume shows up for the bet, not the forecast, and the word "market" is quietly borrowing a credibility the activity has not earned.

I will still read the odds. They are useful, and I am not too proud to use a good number wherever it comes from. I will just not pretend that reading them is the reason $4Bn moved.

Call it a market if you like. I will call it a wager, and I will keep my eyes on the screen I was ignoring for all those years, now that I finally know what is playing on it.

Further Reading