On June 13, Polymarket removed the taker delay for all World Cup matches. Here's what a goal looks like without it for the ARG win market at the 85' mark: - 13,848 shares were resting within 5¢ of the ask when the ball crossed - 90% of all cancels occurred within 65ms of each other - First post-goal taker order lands 63ms behind the cancels and picks off less than $800 - Total sniped on a 3x book move: <$1,000 (600+ total active maker addresses, 11 adversely filled)

Timely listen: my 2023 conversation with Neal Koblitz, co-creator of ECC, on his path into cryptography, early Ethereum advisory work, and the geopolitics of math across the US, Russia, and Vietnam. podcasts.apple.com/us/podcas…

How AI-related prediction markets reacted to Dario's memo:

Over the past month, Rothera’s trade count doubled while volume stayed flat, cutting average trade size in half. This more closely tracks behavior on Polymarket than Kalshi. On Polymarket, trade count outpaces volume, while on Kalshi, trade count and volume has been scaling almost 1:1.

1 in 3 @world_xyz wallets has a positive realized PnL. The median wallet has a PnL of -$2.00

In less than 4 seconds, 44 unique makers on Polymarket lost $19k to 6 takers. First trade: 1.44 seconds after LeBron's tweet Crossed 50¢: 2.26 seconds Reached 99¢: 3.98 seconds First-come, first-served markets create zero-sum latency races that degrade long-term liquidity. This market structure is inherently taker-favored: makers must remove stale quotes before anyone else in the world can take them. Because they can be picked off during every major repricing event, they are forced to quote wider and provide less depth. Exchanges want volume and to lead price discovery. But knowing LeBron chose Philadelphia four seconds earlier provides zero value to most users and imposes negative externalities on the traders creating the prices that observers consume. Different market structures (batch auctions, maker-prioritized cancels) trade off marginally slower re-pricing during rare news events with deeper liquidity throughout the rest of the market.

Market makers earned $790K while takers lost $925K in Polymarket’s LeBron Next Team markets. The $135K difference is Polymarket's fee revenue, excluding rebates. Polymarket also spent ~$65K on liquidity incentives to makers. Makers can profit even in markets where many think adverse selection should make liquidity provision unprofitable.

You can easily tell who is a real user trading on the platform based on on-chain data if a user's transaction includes relay.link