Down $1.72M on Polymarket. Still holding roughly $100K on the Bears. Tiger200’s open position shows 281,499 Bears shares against the Eagles, bought at an average of 35¢. If the Bears win, those shares pay roughly $281,500. If they lose, they pay zero. The same profile lists a $555.9K biggest win and $45.1M in total volume. Even that huge win hasn’t kept the overall account profitable. A six-figure whale bet gets attention. The history behind it deserves a look too.

THE BLOCK: The U.S. Court of Appeals for the Sixth Circuit ruled against Kalshi in its fight with Ohio and Tennessee over sports-event contracts. The court found that Kalshi failed to show the contracts qualify as "swaps" under the Commodity Exchange Act and, even if they did, the CEA does not preempt state gambling laws.

Damn Kalshi temp markets just smoked official weather forecasts in six of seven US cities with ten percent less error

Kalshi’s temperature markets beat the strongest individual public weather forecast in six of seven US cities, according to a new study. Researcher Alexander Crosier compared forecasts drawn from market prices with public weather models across 7,590 city-days. One hour into trading, the market’s forecast error was about 10% lower than the National Blend of Models. Its lead continued overnight into the day it was predicting. Traders could already read those public forecasts. The finding suggests prices were adding information beyond what any single forecast offered. That makes weather markets worth watching even without placing a bet. The September 21 preprint covers data through August 12. In the chart, lower error means a better forecast: purple is Kalshi, green is the public model.

Alright, gang, here’s the latest bullshit from Kalshi. Apparently, Kalshi has been sending emails to tribal leaders across Indian Country asking for meetings at G2E. Of course, I got my hands on the email. So, who do you think they’re sending to Las Vegas? The A Team? B Team? D Team? Gotta be the best, right? Nope. They’re sending the coffee boy. Meet Nicholas Hull, the Intern. I think this might be his first big boy job. We’ve asked several tribes to take the meetings so we can compare deals. We want to know everything We’re also asking tribes to make appointments and then no-show. Let's have fun with this.

Polymarket can load while trading has problems. Its status page shows 100% website uptime for September so far, versus 98.91% for the prediction trading API. Yesterday’s degraded-trading incident ran for 24 minutes. Overnight maintenance also warned that placing orders, cancelling them, and getting fills could be temporarily unavailable. All systems are operational now. But if your plan is “I’ll sell when the news hits,” being able to open the app isn’t enough. Your exit still depends on the trading system working.

China’s biggest rival in the Trump–Putin meeting market is no meeting. Polymarket currently puts China at 70%, no meeting before year end at 22%, and the United States at just 3.4%. The week’s chart shows China pulling further ahead while the no meeting price falls. Over $1.55M has traded across the event. For traders, those are two different bets: expecting another Trump–Putin meeting doesn’t automatically make China the right position. They could meet somewhere else and China YES would still lose. And buying China NO doesn’t require picking another country. It also wins if they don’t meet before the deadline. Snapshot: September 25.

$234k has already traded on which company finishes first on the agent arena leaderboard this month anthropic is priced at 95% openai is at 6% looks like a novelty market but i think this is the early version of benchmark derivatives model rankings are starting to determine enterprise spend, developer adoption and eventually billions in revenue teams will want to trade or hedge model releases, benchmark changes, inference prices and outages the opportunity isn’t guessing the best model every month it’s owning the trusted evaluations and resolution data every market settles against benchmarks become indexes indexes become markets

🆕 New project added to the prediction markets wiki Seer @SeerMarkets is a permissionless prediction-market protocol on Gnosis and Ethereum. Trade binary, categorical, scalar, and futarchy markets with Kleros-powered dispute resolution. ⚡️

JUST IN: @Kalshi hits $7B in Weekly Perps Volume

95% of Stock Token holders in Zerion’s Robinhood Chain sample also held speculative tokens. So the people buying tokenized stocks weren’t sitting in a separate corner of the chain. Most were already active in more speculative markets, too. Stock Token adoption also grew fast in the sample, from 2.4% of addresses in week one to 8.9% by week six. NVIDIA’s Stock Token was the fourth most-held asset, and five of the top 15 were Stock Tokens. The interesting part for builders is the overlap. These wallets are already using different parts of the ecosystem, so apps that combine stocks, swaps, and lending could actually have users waiting for them. Worth keeping the limits in mind, though: Zerion looked at addresses active through its own products during the chain’s first two months. So this doesn’t prove speculation led people into Stock Tokens. But it does show the same users are touching both sides of the chain.

Remove sports from July’s prediction-market volume, and $53 billion becomes roughly $11 billion. That’s the split in Pew Research Center’s new report, published September 23. Sports accounted for: • $31.4B on Kalshi • $10.6B on Polymarket, including its US platform Together, that’s about 79% of their combined volume. Politics, meanwhile, contributed less than $900M across both platforms that month. The World Cup helped drive the surge. But combined volume still reached $47B in August after the tournament ended. One detail worth keeping straight: Pew measures contracts at their $1 face value. These figures measure notional trading volume, not $53B of fresh cash deposited. My takeaway: restrictions on sports contracts would hit the category responsible for most of this activity. The industry’s growth story is heavily tied to keeping those markets available.

Polymarket Perps has a massive wash trading problem. 17 connected wallets trading $10,510,000 strictly with EACH OTHER in less than 96 hours. This is just a tiny snapshot. There is significantly more. How the $10.5M Wash Trading Loop Worked: One wallet funds 17 sub-accounts A single funding source (a whale entity, probably Wintermute or Jump - linked to a Hyperliquid account aliased "ABC") funneled over $10M into an Arbitrum treasury wallet. From there, the money was split across 17 secondary accounts on Polymarket Perps. WASH: Instead of trading with actual market participants, these 17 accounts were paired up on the order books. Account A placed an order, and Account B instantly filled it. RISK: Seconds later, they ran the exact reverse trade. For example, in the $VVV market, two accounts swapped contracts and reversed the same quantity at the same price just 9.36 seconds later. - Net position change: Zero. - Market risk taken: Zero. COSTS: Because of maker fee rebates and dirt-cheap taker fees, running a $10,000 trade back and forth cost only $1.50 in net fees. Pumping eight figures of fake volume cost pennies on the dollar. SCALE: Over just four days (Sept 18–22), this cluster ran this loop 22,210 times, fabricating $10.51 million in volume. In markets like $VVV, this single group accounted for nearly 29% of all trading activity. You thought Kalshi was the only one manufacturing fake volume? Think again. This is only a snapshot from a few days of captured activity. There is significantly more in the public data. You can check the public receipts and reproduce the results here: LINK ⏬⏬⏬ ar.mrx.im/9rQUKrTOHf4-m7xkHq… More soon.

Polymarket had HYPE’s Binance listing at roughly 10% yesterday. Today it’s above 99%. Binance announced spot trading for September 24, sending the year-end listing market almost straight up. YES shares priced around 10¢ are now near $1. That’s roughly a 10x change in the contract’s value, before fees. Traders who backed the listing didn’t need to predict how HYPE would trade after it. The listing itself was the bet.

22% probability on a US diesel export ban before October 31 is an obvious mispricing Politico already reported that prep work for ban is underway Even though the White House denied it, a move like this to slash domestic fuel costs right before the election is what Trump needs Diesel prices have literally doubled since March and with zero end in sight for the war in Iran, how else is he supposed to secure votes in November? For September 30 - No, six days is nothing to ram through a policy like this But October 31 - Yes, it’s safer and Trump basically has to take this shot to revive approval ratings sitting at career lows

Perps aren’t always about getting lucky or doing analysis There are other strategies too, like funding farming. Funding rates often differ across perp platforms, which creates opportunities to profit by opening opposite positions For example, on @perpdexlist I found funding farming setups on Polymarket Perps with up to 543% APR, even after filtering out platforms I don’t know Just to be clear, so high APR usually doesn’t last long. With setups like this, you need to open and close positions very often and actively manage them. Some people even do this full time But there are also setups that last much longer, don’t need constant management and still offer good returns If you’re interested, like this post so I know I should share more

$4,455 on the line to make $45. The trade lost. BoneOhio bought 4,500 UP shares at an average of 99¢ on Polymarket’s September 22 BTC market, covering 5:05–5:10 pm ET. It resolved DOWN. The position shows a $4,455 loss. At that entry price and size, it takes 99 winning trades to cover one full loss, before fees.