johnleslie_pm
Show HN: I built a free AI startup validator – it scored a dog walking app 4/10
Show HN: Free Google Places API alternative using OpenStreetMap (no API key)
Are prediction markets well-calibrated? Analysis of 7,661 Polymarket markets
The interesting design question is how platforms should handle resolution disputes like this. Polymarket uses UMA's optimistic oracle, which allows anyone to challenge a resolution within a dispute window. But that only works when someone is incentivized to challenge. If the manipulator holds both sides of the market at the right ratio, they can profit even if the resolution is later corrected. Serious prediction markets will eventually need to adopt multi-source oracle designs the way DeFi lending protocols moved to Chainlink-style aggregated price feeds after the flash loan attacks of 2020.
The comparison to stock tickers is apt. Bloomberg terminals started as specialized tools for bond traders. Once financial data appeared on every TV screen, it changed how the public understood markets. Prediction markets are following the same path, just compressed into months instead of decades.
The tension is that news organizations now have a financial incentive to cover events that move prediction markets. This creates a feedback loop: dramatic events move markets, market movements become news, news coverage moves markets further. We saw this with the Hormuz situation this month, where prediction market odds were cited in coverage of the shipping blockade, which then moved the odds.
Whether this is good or bad depends on whether you think the signal (real-time probability estimates) outweighs the noise (gamification of serious events). The Forbes incident with the mass shooting market suggests we have not figured out where the line is yet.
Kalshi has about 89% market share among regulated US platforms. They have a lot to lose if Congress decides prediction markets need the same treatment as securities. Suspending candidates who bet on their own races is a cheap way to demonstrate they take integrity seriously.
The interesting question is whether self-regulation is enough. These three candidates were caught because the bets were on their own races, which is easy to detect. The harder problem is friends, family, and staffers placing bets based on internal campaign data. That pattern is nearly impossible to catch without the kind of surveillance infrastructure that the SEC uses for stock trading.