Original research on markets and their participants.
Prediction markets
Browse live odds by topic
6 h 56 minPresidency: Polymarket above 90% before the AP callElection Night Odds: Polymarket Passed 90% for Trump Nearly 7 Hours Before the AP Call
On election night 2024 Polymarket's presidential market moved from 58.3% for Trump at 7 p.m. Eastern to above 90% at 10:38 p.m., 6 h 56 min before the AP call. In Michigan the lead was 12 h 19 min. Minute by minute, against every AP call.
None foundCalifornia orders against Kalshi or PolymarketIs Kalshi Legal in California? Kalshi, Polymarket and Tribal Land
California has no legal sports betting, yet Kalshi and Polymarket US take Californians' sports contracts. The Ninth Circuit's 16 September 2026 ruling covers tribal land only. What each venue does in California, and what the state has done.
2 of 3Federal appeals courts that have ruled for the statesIs Kalshi Legal in My State? Kalshi and Polymarket US, State by State
Every US state and DC, with the court rulings, lawsuits and regulator letters aimed at Kalshi and at Polymarket US, kept apart. A blank record is not permission, and polymarket.com stays closed to new US trades.
None foundTexas actions naming Kalshi or PolymarketIs Kalshi Legal in Texas? Kalshi, Polymarket and Texas Gambling Law
Texas makes betting on games and elections a crime, but no Texas authority has acted against Kalshi or Polymarket US. What the Penal Code says, what the Senate is studying, and how polymarket.com differs.
7Kalshi contract categories barred in WashingtonIs Kalshi Legal in Washington State? The Court Order, and Polymarket
A King County judge ordered Kalshi on 12 August 2026 to stop offering sports, election, politics, culture, tech and mentions contracts in Washington. What the order covers, what it leaves out, and where Polymarket US and polymarket.com stand.
0Prediction markets recognised or registered in CanadaIs Polymarket Legal in Canada? Kalshi, Polymarket and the Provinces
Kalshi's agreement restricts its event contracts in all of Canada, Polymarket limits four provinces, and Canada's securities regulators have recognised no prediction market. Province by province, with Ontario's 2025 order.
Blockedpolymarket.com in IndiaIs Polymarket Legal in India? Kalshi, Polymarket and the Online Gaming Act
India's Online Gaming Act bans online money games, and the government says polymarket.com is blocked. Kalshi lists India as restricted. What the Act and the Rules say, and what each venue does.
302Places in the directoryIs Polymarket Legal in My Country? Kalshi and Polymarket, Place by Place
A directory of 302 places with what Kalshi's member agreement and Polymarket's documentation say about each, kept apart from local law. Guides for the US, Canada, the UK and India, with more to come.
0Gambling Commission licences for Kalshi or PolymarketIs Polymarket Legal in the UK? Kalshi, Polymarket and the Gambling Commission
In Great Britain a prediction market needs a Gambling Commission licence, and neither Kalshi nor Polymarket has one. Northern Ireland has its own law. What each venue's terms say and what the regulators have said.
3Supreme Court petitions on sports event contractsIs Polymarket Legal in the US? Where Prediction Markets Stand in 2026
Polymarket US runs on a CFTC-registered exchange; polymarket.com takes no new U.S. trades. Two of three federal appeals courts have sided with the states, three Supreme Court petitions are pending and the CFTC adopted an interim rule on 19 October. Where things stand at 30 October 2026.
15 of 15ECB decisions called a week outECB Rate Decision Odds: Polymarket's Favourite a Week Out Was Right at 15 of 15 Meetings
A week before each ECB decision from October 2024 to September 2026, Polymarket's favourite was the decision the ECB took, at a median 98.2%. A month out it missed twice. Both 2026 rate rises were priced above 98% a week ahead.
24 to 69¢$10,000 on a Polymarket long shot, paid above the screen priceHow to Read Election Odds: Five Checks Before You Trust a Prediction-Market Price
A prediction-market price is what a $1 contract costs, not a vote share. On seven thin midterm contracts in September 2026 a $10,000 buy paid 24 to 69¢ above the screen price, and political long shots won more often than priced. Five checks, and how Vultax will score 13,186 midterm contracts.
24European jurisdictions with a public entry on recordRegulatory actions on Polymarket and Kalshi in Europe
European regulatory records on Polymarket and Kalshi, dated and sourced: 24 jurisdictions, seventeen blocking orders, one penalty, one licensing regime. Eleven of the seventeen blocks are not applied by Polymarket itself, and only France, Spain and Switzerland have network tests.
64%Polymarket books that could fill 10,000 contractsPolymarket Order Book Depth: 10,000 Contracts Fit 64% of Books
Vultax walked the displayed asks in 7,354 hourly Polymarket order-book snapshots at three order sizes. 100 contracts fit all 2,701 books in range. 10,000 fit 64% of them, at a median 0.25¢ above the best ask, before fees.
$149,859 a dayPolymarket liquidity rewards offeredPolymarket Liquidity Rewards: $150,000 a Day, and Not Where the Trading Is
Vultax read every listed Polymarket liquidity reward late on 13 September 2026. Polymarket offered $149,859 a day across 17,866 markets, with finance and weather taking 27% of it on 6% of trading; the median rewarded market offered $3 a day in total.
59.6%Observed Polymarket trades in up-or-down marketsPolymarket's Up-or-Down Crypto Markets: 60% of Its Trades, Priced About Right, 1.6% Fee
Vultax counted 15.0 million Polymarket trades over 251 hours: 59.6% were up-or-down crypto bets, 379 wallets made most of them, and takers paid about 1.6% of notional on five-minute trades. In 2,497 markets checked against outcomes, Up won 50.2% and favourites won as often as priced.
55.3%Polymarket open interest on politics and electionsWhere the Money Waits: 55% of Polymarket's Open Interest Is on Politics
Vultax read every open Polymarket event late on 13 September 2026. Of $341.3 million of open interest across 20,851 events, 55.3% sat on politics and elections and 7.8% on sports; the ten largest events held 44.4% of it. Polymarket's taker fees by category are set out beside it.
6,776 marketsValidation sampleCalibration of Polymarket prices against 6,776 resolved markets
Original Vultax research: in an experimental test, a calibrated fair-value model beat raw Polymarket prices by ~9.6% on held-out Brier score. Where the favorite-longshot effect lives, what the model can honestly flag, and why it still is not a trading edge.
21.4%Esports share of observed Polymarket moneyPolymarket Sports and Esports: Esports Took 21.4% of the Money on the Thinnest Books
Of $727.0 million of Polymarket trading Vultax observed from 21 August to 12 September 2026, esports took 21.4%, peaking mid-morning UTC. Before start on 12–13 September, a typical League of Legends book held $413 within 2 cents; an NFL game $523,000.
77.9%Ocasio-Cortez's implied chance if nominated2028 Election Odds: Vance 20%, Democrats 64.5%, and Three Democrats Within Two Points
JD Vance leads Polymarket's 2028 presidential market at 19.9% and Democrats are 64.5% to win the White House. Alexandria Ocasio-Cortez, Gavin Newsom and Jon Ossoff are within two points of each other for the nomination, and the three markets do not add up.
62.1%Milwaukee Brewers, 2026 champion priceWorld Series 2026 Odds: Brewers Were 62% Before Game 1 and Lead 3–0
Before Game 1 the Milwaukee Brewers were 62.1% to beat the Tampa Bay Rays on Polymarket, against 2.2% on 1 February; they won the first three games. The Dodgers led the market on 264 of its 271 days and were swept. In 2025 the Dodgers fell to 10.5% in Game 7 and won.
67.1%Harry Kane, 2026 winner price a week outBallon d’Or 2026: Kane Won at 67%, and the Price Did Not Move Early
Harry Kane won the 2026 Ballon d’Or after leading Polymarket at 67.1% a week out. Unlike 2024, his price jumped only in the minute of the announcement.
51.79%Lula, runoffBrazil Election 2026 Results: Lula 51.79%, and How the Polls and Odds Did
Lula won Brazil's runoff with 51.79% of valid votes. Ten of 11 final polls, the Vultax polls model and the market leaned Flávio Bolsonaro. The model missed his share by 2.2 points, closer than nine polls; on count night the market moved 75 minutes before the count.
64.3%Deadline markets above 65¢ a month out that came truePolymarket 'By Date' Favourites a Month Out: 64% Came True, and 79% in a 2026 Recheck
Every settled Polymarket market worded 'by' or 'before' a date, January 2025 to April 2026, scored against its price 90, 30 and 7 days before the deadline and clustered by event. Markets above 65¢ a month out came true 64.3% of the time; a week out, 90.5%.
+65.8 ptsWinner's move in the 12 hours before the 2025 announcementNobel Peace Prize odds: 2026 result, saved forecast and 2025 study
Navi Pillay won the 2026 Nobel Peace Prize. Compare the official result with our frozen forecast, dated 2026 market snapshots and the 2025 study of prices before the announcement.
2.2 hoursMedian time from scheduled end to settlementSettlement times of 110,576 Polymarket markets, 15 August to 10 September 2026
Vultax measured every Polymarket market that ended between 15 August and 10 September 2026. Most settled within hours of their scheduled end, 37% closed early because the result was already known, and 89 were disputed, rare overall but most common on geopolitics and big markets.
21 of 22Month-end checks the week-out favourite wonBest AI Model Odds: Polymarket's Favourite a Week Out Won 21 of 22 Month-End Checks
Polymarket asks every month which company will top the LMArena text leaderboard. A week before each check from December 2024 to September 2026, the favourite won 21 times in 22. The one miss: Gemini 2.5 Pro, released six days before the March 2025 check.
−20 to +40 sIllustration: a one-minute sampler around a call at 12:00:20Did a market move before the election call? First measure the clock
A price can appear to beat an election call when a minute of missing observations hides the order. Here is how to read the clock before claiming a lead.
50% / 50%Datafolha valid-vote sharesBrazil Election Polls vs Polymarket Odds: What the Numbers Mean
Datafolha’s 20–22 October poll and a dated market snapshot answer different questions. Read vote shares, win probabilities and the model’s limits.
20:00 UTCOfficial results beginBrazil Election Results 2026: When Counting Starts and Where to Check
Brazil's runoff is on 25 October. Official results begin after 17:00 Brasília. Check UTC and local times, the TSE portal and partial-count labels.
79.3%Races where the favourite already trades at 90¢ or morePolymarket 2026 midterm scorecard: 13,186 contracts frozen before the vote
Every Polymarket contract about the 2026 US midterms, frozen on 17 October: 13,186 contracts in 1,352 events, a hashed manifest, two price checkpoints and a Brier rule declared before the vote. In 402 of 507 races the favourite already trades at 90¢ or more.
87¢Typical bid sum across every outcomePolymarket efficiency: two tests, updated monthly
On PredictIt, the bids on every outcome of a busy market added up to 110 to 125¢. On Polymarket the typical event adds up to 87¢, and long shots win more often than their price says. Two tests, rerun every month.
$8Platform charge on $1,000 of traded valueGate Polymarket Fees: The Extra $8 on a $1,000 Trade
A worked guide to Gate's platform charge and the separate Polymarket protocol fee, including maker orders, fixed spending and a round-trip example.
31%The favourite range won, two days before the closeElon Musk Tweet Count on Polymarket: 101 Rounds Checked, the Favourite Wins 1 in 3
We checked 101 rounds of Polymarket's bet on how many times Elon Musk posts on X in a week. Two days before a round closes, the crowd's favourite range wins 31% of the time, close to its 34% price. When the crowd is wrong, Musk has usually posted more than it expected, not fewer.
84.5%Sports and esports share of Polymarket's dollars, Sunday 27 September 2026Is Polymarket Just Sports Betting? 84.5% of Its Money on an NFL Sunday
On Sunday 27 September 2026 sports and esports took 84.5% of the dollars traded on Polymarket and 83.8% of the contracts. Over sixteen months the dollar share ran from 31% to a low of 22% and up to 46%, and NFL Sundays sat about twenty points above weekdays.
2 of 3Federal appeals courts that have ruled for the statesPrediction Market Lawsuit Tracker: 50 Cases, Checked Against the Court Record
Every court case Vultax follows over sports and event contracts, from Supreme Court petitions to class actions and two insider-trading prosecutions: 50 in all, checked against the court's own documents or docket on 27 September 2026 and updated on 19 and 25 October.
$46.6MTaker fees, January to April 2026How Polymarket Makes Money: $46.6 Million in Taker Fees in Four Months
Polymarket charged no trading fees in 2025. From 1 January to 28 April 2026 takers on its international exchange paid an estimated $46.6 million, $28.4 million of it in April, two thirds on crypto markets. Fees by month and category, the schedule behind them and its other revenue.
309 of 1,000Most-traded open markets that charged takers nothingPolymarket API in Python: 3 Requests, No API Key Needed
A standard-library Python script reads one Polymarket market, the order book for one outcome and the fee that market charges, in three public requests with no key. Read this way, 309 of the 1,000 most-traded open markets charged takers nothing on 20 September 2026.
87.5%September hike, saved midpointSeptember Fed Odds: 87.5% Hike Price, but an Incomplete Scorecard
Polymarket priced the 16 September hike at 87.5% at two saved checkpoints. One missing midpoint makes the full five-outcome score unavailable. October captures and their quote coverage are preserved separately.
Polymarket traders
Open Polymarket trader analytics
−5.40¢Corrected tight screen, matched comparisonA long-shot screen on historical Polymarket fills: base rates and a correction
The long-shot screen tested on historical Polymarket fills, with a correction to the tight screen: 1,316 bets and a matched gap of −5.40 cents per share. Earlier sensitivity tables are retained as historical results.
+1.43¢Extra paid 30 seconds after a top trader's buyPolymarket Copy Trading: Following 30 Seconds Late Cost 1.43¢ a Share, Before Fees
What copying a Polymarket trader costs, measured and worked. Behind buys by the 1,000 most profitable wallets, a follower 30 seconds late paid 1.43¢ a share more in event markets; fees and fill put a copier of a $50 crypto trade down about 6% before the market moves. Four tools compared.
43.9%Realised gains taken by the top 1% of winnersBest Polymarket Traders: 36 Wallets Took 44% of Gains; Winners Repeat 41–51% of the Time
Who makes money on Polymarket, whether winners keep winning, and how five leaderboards count. The top 1% of winning wallets took 43.9% of realised gains; across all wallets, a winning month was followed by another 41–51% of the time, and 124 of 137 among the most active.
4 of 22Tools that publish a latency figureBest Polymarket Tools and Whale Trackers in 2026: 22 Compared, 4 Publish Their Latency
22 Polymarket tools compared on latency, P&L method, history, liquidity context and price, including eight whale trackers. Vendor checks: 5 September 2026; Polywhaler corrected 26 September. Four publish latency figures; every category has a free tier.
32 / 37KPMG-client cycles, selected groupWhere unusual Polymarket wins concentrate
The same published accounts beat entry-price expectations on one auditor’s clients, with no excess elsewhere. A broader wallet screen raises a different question.
1.6%Wallets active in every hour of the clock1.6% of Polymarket Wallets Never Sleep. They Made 58% of All Trades.
Hourly activity for 130,346 Polymarket wallets over 12 days in August and September 2026: 2,080 wallets were active in every hour of the clock and filled 58.4% of 9.82 million trades. The ten busiest wallets alone filled one trade in five. Half of all wallets traded on a single day.
Crypto markets
Open the BTC venue desk
5.8 bpsTypical gap between the highest and lowest venueWhy Bitcoin's Price Differs Between Exchanges: 18 Venues Checked Every Minute
Vultax read Bitcoin's order books on 18 exchanges every minute for 28 hours. Half stayed within a basis point, the highest and lowest ordinary venues sat a median 5.8 bps apart, and Upbit ran 1.31% above. In four weeks of readings no gap survived fees; here is what a gap has to clear.
288–590 msFeed latency at the 95th percentileOrder-book feed latency and coverage on 16 crypto exchanges, September 2026
Order-book feed latency on 16 crypto exchanges: medians from 35.0 ms (Coinbase) to 111.5 ms (BTSE), a 95th percentile of 288–590 ms on every venue, and data in every minute of the day. Plus the coverage check that catches a feed that stays connected and stores almost nothing.
$920.2MLargest spoofing penaltyOrder-Book Spoofing: What It Looks Like in Live Data
Spoofing is defined by intent, which no market-data feed can observe. Here are the order-book patterns that stand in for it, and where they fail.
115,006Alerts logged115,006 Large-Trade Alerts: Where and When Crypto Whales Trade, by Venue and Hour
Every large-trade alert Vultax logged on six exchanges from 18 June to 3 September 2026: HTX carried 56% of large-trade notional, flow stayed close to balanced, 13:00–16:00 UTC held 23.6% of alerts, and weekends ran 23–30% quieter.
51%–95%Published fake-volume estimatesCrypto Wash Trading and Real Liquidity: Fake-Volume Estimates Run From 51% to 95%
Published estimates put fake or non-economic bitcoin volume at 51% to 95%, and above 70% on unregulated venues. Why the studies disagree, why depth and slippage judge a venue better than volume (Binance: 64.3% of volume, 30.7% of depth, per Kaiko), and what to check yourself.
The archive
All editions
What the Market Expected: 153 Historical Economic-Event Auctions
Original research and auction data by Refet S. Gürkaynak and Justin Wolfers
Explore 153 historical economic-release auctions: probability ranges, announced outcomes, source lineage and credits to Gürkaynak and Wolfers.
What the Ocean Forecast Remembers: The ENSOFORECAST Archive
Original model by Stephen E. Zebiak and Mark A. Cane
Archive: Lamont-Doherty Earth Observatory / Columbia IRI Data Library
Archive transferred by Aaron Kaplan, IRI
Explore historical ENSOFORECAST maps with original credits, traceable fields, and clear distinctions between hindcasts and unknown issue times.
More from Insights
Photographs come from Wikimedia Commons under free licences. Each piece names the photographer and the licence under its picture.
Continue your research in Vultax.
Explore crypto and prediction markets, trader analytics and event monitors. Public research views are free to browse.
New studies, by email.
Measured, sourced, and sent once per study. No digests, no promotions.