The Sharp Migration

Rufus Peabody is one of the most successful sports bettors in America. He has made millions betting on sports over the past decade, developing quantitative models sophisticated enough to beat the sharpest books in the world. He told Bloomberg this week what he sees coming: “It really feels like everything's prediction markets, prediction markets, prediction markets.”

Peabody is not alone. Professional gamblers — the sharps — are migrating from sportsbooks to prediction markets in increasing numbers. The reasons are structural, not sentimental. Sportsbooks limit winning bettors. They restrict account access. They cap bet sizes. A sharp who develops a genuine edge faces a paradox: success leads to exile. The better you are, the less you can bet.

Prediction markets invert this constraint. On Kalshi, Polymarket, and ForecastEx, there are no account restrictions based on performance. No bet limits tied to win rate. No trader classification that separates winners from losers. A sharp can deploy capital at scale without the platform engineering them out. The market structure treats every participant identically.

The additional appeal: prediction markets are tradeable. A sportsbook bet is a static position. Once placed, it resolves at settlement. A prediction market contract can be bought, sold, hedged, and arbitraged continuously until expiration. Wall Street strategies — market making, statistical arbitrage, event-driven trading — translate directly. For a sophisticated bettor accustomed to thinking in probabilities and expected value, prediction markets are not a new product. They are the same product with better microstructure.

The Super Bowl as case study

Super Bowl LX produced a natural experiment in the two ecosystems. Traditional sportsbooks handled approximately $1.76 billion in legal wagers. Prediction markets processed roughly $1.3 billion. The gap has closed by orders of magnitude in a single year.

Sportsbooks had what industry sources called a “helluva day.” The low-scoring game favored the house. Unders hit. Spreads moved in book-friendly directions. The public lost. This is the normal Super Bowl outcome. The books are designed to win.

The prediction market side told a different story. Kenneth Walker III was trading at 8 percent for Super Bowl MVP before kickoff. He won. Anyone who bought that contract at 8 cents collected a dollar. Devin Booker, the Phoenix Suns guard, won $800,000 betting Seahawks -3.5 — presumably through a prediction market platform, since no regulated sportsbook would allow an active professional athlete to place an $800,000 Super Bowl wager without disclosure requirements. The line between sports betting and prediction markets has dissolved. The participants are the same. The venue has changed.

The Theo Effect

The catalyst for institutional attention arrived before the Super Bowl. In November 2024, a French trader known as Theo wagered approximately $45 million on Donald Trump to win the presidential election. He made $85 million in profit. His edge: private polling data. He commissioned surveys in swing states, analyzed the microdata, and concluded the public polls were underestimating Trump support. He bet the divergence.

Theo was right. And his success — documented extensively by The Wall Street Journal, Bloomberg, and The New York Times — became the proof of concept for professional-grade prediction market trading. The story demonstrated that information edge could be monetized at scale. It showed that prediction markets were liquid enough to absorb eight-figure positions. And it proved that the platforms would not limit or ban a trader for winning.

The industry calls this the “Theo Effect.” Institutional interest accelerated in the weeks after his success became public. Hedge funds began exploring prediction market allocations. Goldman Sachs executives took meetings with Kalshi and Polymarket leadership. The narrative shifted from “interesting retail product” to “legitimate alternative data source.” One trader's success changed the perception of an entire asset class.

The reflexive dynamic is visible. Theo's profit attracted attention. Attention attracted capital. Capital improved liquidity. Improved liquidity attracted more sophisticated traders. The loop feeds on itself.

668 addresses, revisited

The sharp migration raises a question the industry prefers not to answer: what happens when everyone is sharp?

As documented in “668 Addresses,” Polymarket's profit distribution is radically concentrated. Of 1.7 million trading addresses, 668 captured 71 percent of all realized gains. The bottom 70 percent of addresses lost money. The structure is not a bell curve. It is a funnel where capital enters at the bottom and exits at the top.

This distribution exists because of information asymmetry. Sophisticated traders — arbitrageurs, model-based quantitative traders, and those with genuine information edge — extract value from less sophisticated participants. The 668 addresses are not lucky. They are skilled, or connected, or both. They represent the population that prediction markets reward.

The sharp migration threatens to change the composition of participants. When professional gamblers enter, they bring skill. They bring capital. They bring models. They compete with the existing 668 for the same edge. The pool of unsophisticated capital — the liquidity that funds the profits — does not grow proportionally. The denominator stays flat. The numerator of skilled participants increases.

In financial markets, this dynamic is well understood. When alpha becomes crowded, it decays. The strategy that works when one firm runs it stops working when a hundred firms run it. Information edge has a half-life. The more participants who exploit it, the faster it disappears.

The efficiency paradox

The prediction market industry celebrates the sharp migration as validation. Professional traders entering the market means the market is serious. It means the prices are worth trusting. It means the asset class has arrived.

This is true. And it is also the mechanism by which edge disappears.

Sharp traders improve price efficiency. When Rufus Peabody spots a mispriced contract and trades it toward fair value, the market becomes more accurate. When quantitative funds deploy models that identify arbitrage opportunities, they eliminate those opportunities. When Theo-style private polling informs eight-figure bets, the information gets impounded into prices. The market reflects what the sharps know.

This is good for information consumers. CNN, Bloomberg, and Google display prediction market prices as authoritative probabilities. Those prices are more accurate when sophisticated traders correct mispricings. The public good — accurate probability estimates — improves.

It is bad for the sharps themselves. Edge exists in inefficiency. The Kenneth Walker III contract at 8 percent was a market failure — and a profit opportunity. If the market had priced him at 25 percent, the opportunity would not have existed. When sharps migrate, they arbitrage away the mispricings that attracted them. Their presence is self-defeating.

The reflexive loop runs in both directions. Sharp entry improves efficiency, which reduces edge, which should reduce sharp interest. But the sportsbook alternative — limited accounts, capped bets, exile for winning — remains structurally hostile. Even a diminished edge in prediction markets beats no access to the sportsbook market. The migration continues because the relative opportunity persists, even as the absolute opportunity shrinks.

The industry fracture

The sharp migration is accelerating a structural divide in the gambling industry. DraftKings, FanDuel, and Fanatics have all exited the American Gaming Association — the trade group representing traditional casinos and sportsbooks — to pursue prediction market ventures. The AGA spent 2025 lobbying against prediction markets as unlicensed gambling. Its largest members concluded prediction markets were the future.

The fracture is about customer economics. Sportsbooks profit by limiting sharps. They want recreational bettors who lose predictably. Prediction markets profit by volume. They want anyone who will trade, including sharps who attract liquidity. The business models are incompatible. The same company cannot simultaneously limit winning bettors and welcome them.

DraftKings Predictions, FanDuel Predicts, and Fanatics Markets are attempts to have it both ways. Offer prediction markets to the sharps who cannot bet on the sportsbook. Offer sportsbooks to the recreational bettors who lose. Geographic separation enforces the divide: DraftKings Predictions operates in 17 states, specifically the states where DraftKings does not hold a sportsbook license. The company is segmenting its customer base by regulatory arbitrage.

The AGA members who remained — Caesars, MGM, Wynn — do not have prediction market ventures. They continue to oppose the asset class. The industry is splitting along a fault line: those who see prediction markets as a threat to their sportsbook margins, and those who see prediction markets as an escape from geographic licensing constraints.

The equilibrium question

The sharp migration poses a question prediction markets cannot answer internally: where does the liquidity come from in equilibrium?

In the current state, the 668 addresses profit from the 1.2 million addresses that lose. The success stories — NPR profiles, Bloomberg features, Theo documentaries — attract new retail participants. Those participants become the liquidity pool. The loop closes.

If the sharp migration continues, the composition shifts. Sophisticated traders on one side, more sophisticated traders on the other. The retail participants who fund the game either learn (and become sophisticated themselves) or leave (and stop providing liquidity). Either way, the profit opportunity compresses.

Institutional capital does not solve this problem. Hedge funds do not enter markets to lose money. They enter to capture edge. If Goldman Sachs launches an Event-Linked Notes product, it will be structured to profit from prediction market exposure — likely by providing liquidity and earning bid-ask spreads, or by using prediction market probabilities as inputs to other trades. Institutions are not the counterparty. They are additional competition for edge.

The endgame may be a market that is highly efficient, highly liquid, and minimally profitable for participants. A utility for probability discovery, not a venue for wealth extraction. The sharps will have improved the market by entering it. They will have destroyed their edge in the process.

The reflexive conclusion

The sharp migration is the prediction market industry's validation and its eventual limitation.

When Rufus Peabody says everything is prediction markets, he is describing a transition. Professional gamblers are leaving sportsbooks because prediction markets offer better structure. That migration improves prediction market quality. The prices become more accurate. The liquidity deepens. The institutional case strengthens.

And in the same motion, the migration erodes the conditions that made it attractive. The mispricings get arbitraged. The soft money gets educated or exits. The edge that drew the sharps evaporates as they arrive. The system is reflexive: the participation changes the market, and the changed market changes who wants to participate.

The prediction market industry is entering its professionalization phase. The winners will be the platforms that can sustain volume as edge compresses — the ones that attract capital for reasons beyond profit opportunity. Hedging. Information consumption. Entertainment. The sharps will remain. But they will be trading for thinner margins in more efficient markets, competing with each other rather than harvesting from the uninformed.

This is what it means for a market to mature. The sharps arrive. The inefficiencies disappear. The product becomes a utility. The gold rush ends, and the infrastructure remains.

The question for the 668 addresses — and for every sharp migrating into prediction markets today — is whether they are early enough to profit from the transition, or late enough to become part of the efficient equilibrium they helped create.

dawson@monetizeopinion.com