On the morning of October 25th, Polymarket traders gave Flávio Bolsonaro roughly an 80% chance of becoming Brazil’s next president. Luiz Inácio Lula da Silva took the runoff that night, 51.7% to 48.3%, and many confident bettors went to bed poorer. Because Brazil bars its own citizens from these markets, foreigners trading a thin book from abroad set most of the price. They got the race wrong and paid for it in cash.
Now, 6 days later, the same platform puts Ken Paxton’s chance of beating James Talarico (CIS) in the Texas Senate race at 28.5%. The explanation making the rounds on 𝕏 is that some billionaire is buying a gloomy number to demoralize Republicans before Tuesday. I think it is wrong. The Paxton market may well be mistaken, as the Brazil market was, but there is no evidence anyone is rigging it, and the market’s design makes rigging hard if not impossible to pull off undetected. A wrong market and a rigged market are different animals, and conservatives of all people should know the difference.
Friedrich Hayek explained why in his 1945 essay “The Use of Knowledge in Society.” Every buyer and seller, he argued, folds a sliver of what he knows into a price, so the price gathers scattered knowledge no planner ever could. A price is a summary, and summaries can be wrong. A free price earns trust because anyone who thinks it wrong can profit by correcting it, and that standing invitation guards the Paxton market better than any regulator could.
The boring explanation fits the facts. In the RealClearPolling average, Talarico (CIS) sits at 48.8% and Paxton at 47%, a gap of 1.8 points. People routinely confuse a candidate’s share of the vote with his chance of winning. If polls can miss by a typical 3 points either way, a 1.8-point lead works out to roughly a 73% chance of victory, almost exactly where Polymarket sits. Kalshi, a separate federally regulated exchange with about 7 times Polymarket’s trading volume, prices Paxton at 29%, and any rigging theory has to explain that too.
A billionaire manipulating the price would leave tracks, big lopsided trades about as subtle as a bulldozer in a flower bed. There are none. The largest single bet tracked in the Texas Senate event is $58,100 against Paxton, and the largest bet for him, $35,700, came from a first-time wallet. Since September 22, trackers count 189 purchases of $5,000 or more from 85 wallets, about $2.2 million in all, spread across both sides. The biggest wallet backing Paxton holds about $122,000; the biggest against him, about $198,000. To a billionaire, those are rounding errors.
Part of the alarm comes from misread data. Pull the raw list of Paxton “No” holders off the Polygon blockchain and one address sits on roughly 903,477 shares, which looks like the mother of all anti-Paxton wagers. Nobody owns it. Polymarket’s public contract code sends tokens there whenever a trader converts a bet against one candidate into equivalent bets on the others, and no one is known to hold a key that could spend them. The address works like a shredder whose contents stay visible on the ledger, and a hasty analyst mistakes it for a whale. Set it aside, and no real wallet holds more than about 15.5% of either side, among roughly 4,300 wallets with a stake in the race.
A clever billionaire, the theory goes, would hide behind 1,000 small accounts. Opening an account on Polymarket US requires a legal name, birth date, address, phone number, and Social Security number, and the exchange can demand a government ID and a selfie. Rule 7.2(m) of its rulebook prohibits concealing or fragmenting ownership to defeat surveillance. This contract trades on the international platform, where wallets are pseudonymous, but money still has to come from somewhere and go somewhere. Polymarket says it hunts for wallets that share funding sources, trade in correlated sizes and times, and withdraw to the same destinations, and it uses Chainalysis, the blockchain forensics firm. In April, the Justice Department credited Polymarket’s cooperation when it charged a soldier with trading on classified information. A billionaire can create 1,000 wallets. He cannot create 1,000 strangers.
In October 2024, the Democrats and their willing accomplices in the drive-by media were certain a French trader called Théo was rigging Polymarket for Donald Trump. He had spread about $30 million across 4 accounts; within days, sleuths tied all 4 to one man, and Polymarket concluded he was making a very large bet. He was also right.
Critics also picture the wrong kind of market. Betting at the racetrack is pari-mutuel: every dollar goes into a common pool, the track takes its cut, and the odds are whatever the leftover money says when the gates open. A rich man who drops $1 million on a long shot changes the payout for everyone, nobody across the counter offers to bet against him at a better price, and once his ticket prints he cannot sell it. A casino sportsbook is no better, since the house sets the line on Sunday’s football game, takes the other side of your bet, and locks you in until the final whistle.
The stock market comes closer but has a weakness a manipulator can exploit. A company issues a fixed number of shares, and once a buyer gobbles up enough, nobody can conjure more. That is how a squeeze briefly made Volkswagen the most valuable company on earth in October 2008. A stock also never expires and has no fixed final value, so a distorted price can drift for years.
Polymarket is different. Picture the currency booth at an airport, posting one price to buy euros and a slightly higher price to sell them. Now imagine hundreds of competing booths, each posting its own prices with its own capital, all shown on one board. That board is the order book, and the booths are market makers. On Saturday, the best offers for Paxton “Yes” stood at 28¢ to buy and 29¢ to sell, a 1¢ spread, and the 28.5% headline is the midpoint.
Shares in a prediction market have no fixed float, and that fact alone should end most conspiracy talk. Since exactly one of the two will pay $1 when the race is called, anyone can deposit $1 and receive a “Yes” share and a “No” share on Paxton. Say a billionaire bids “Paxton loses” up to 85¢. A market maker who thinks that price is absurd deposits $1, sells the “No” share to the billionaire for 85¢, and keeps the “Yes” share at an effective cost of 15¢. If the maker believes Paxton’s true chance is 30%, he has just bought a 30¢ asset for 15¢ and can repeat the trade as long as the billionaire keeps buying. Nobody corners a supply that grows every time he pushes on it.
These contracts, unlike a racing ticket or sportsbook slip, can be bought or sold at any moment until the race is called, which turns the manipulator’s job from a single expensive shove into a siege. Holding Paxton at 15¢ for weeks means buying against every trader who spots a bargain, on a public blockchain that timestamps each trade and records each wallet. At 29¢, 100 Paxton shares cost $29 and pay $100 if he wins. A trader who honestly thinks the race is a coin flip sees $21 of expected profit, and the lower a manipulator drives the price, the fatter that profit gets. He would be paying strangers to bet against him, in public, against a deadline, and on November 3 every contract settles at $1 or nothing.
In the Republican primary market, the Anti-Corruption Data Collective, no cheerleader for prediction markets, documented a $1,240 trade that lifted Paxton’s price by 11¢, and the higher price held for weeks. That happened, but months ago in a thin market, when a trader with pocket money could move the needle because few others were watching. The same report found about $24,000 in February trades lifting Wesley Hunt’s price and roughly $38,830 pushing Paxton’s down, then watched both moves reverse once polling showed Paxton leading Hunt 38 to 17. Thin markets wobble, and more traders steady them.
Washington’s answer should be sunlight. The CFTC ought to require regulated exchanges to publish trade-level data with a short delay, along with holder-concentration figures that exclude protocol addresses like that burn wallet, so anyone alleging manipulation can test the charge against real price movements. Prosecutors should keep pursuing genuine fraud, as they did in April, and lawmakers should resist the blue-state itch to sue these markets out of existence. Republicans have homework too. We laughed at the left in 2024 for blaming a market it didn’t like, and we should not borrow its excuse in 2026.
Texas voters can read 28.5% the way they read a weather forecast. A 30% chance of rain means you carry an umbrella, not that the clouds are conspiring against you. FiveThirtyEight gave Donald Trump about a 29% chance on the eve of the 2016 election, and we all remember how that ended. Paxton’s odds are a snapshot of polls and opinion this weekend, definately not a billionaire’s decree. Early voting closed Friday; Election Day is Tuesday. The only number that matters will be tallied by county clerks from Texarkana to El Paso.
The larger quarrel is about free institutions. For decades the Democrat Party has blamed billionaires for every result it dislikes, the Kochs one year and Elon Musk the next, because a conspiracy spares you the labor of persuading anyone. Conservatives hold a harder view. Free people trading openly under clear rules produce knowledge no committee could assemble, and sometimes that knowledge is wrong, the way a jury is sometimes wrong, and we keep both anyway because each answers to evidence. A Republican who cries “rigged” every time a price disappoints him has quietly adopted the other side’s theory of the world. Polymarket is guessing about Texas, and on Tuesday Texans get to answer. I’m betting the market is wrong and that Ken Paxton will be our next U.S. Senator.
If you enjoy my work, subscribe x.com/amuse/creator-subscriptions/subscribe
Alexander Muse is a Fellow at the John Milton Freedom Foundation and publishes political analysis at amuseonx.com. Readers who identify errors are invited to contact me directly. Each op-ed is edited for grammar and clarity using AI in partnership with Grammarly. Data provided in a sponsored partnership with Polymarket.
Your keyboard to God’s ear. 🙏🏻