Jeff Probst says prediction markets like Kalshi and Polymarket are incentivizing Survivor contestants to spoil outcomes for betting profits. The platforms are now considering anti-spoiler measures, but the damage reveals how prediction markets can corrupt the things they're betting on.

This is prediction markets eating themselves.

Here's how it works: Prediction markets allow betting on Survivor outcomes. Who wins immunity? Who gets voted off? Who wins the season? The markets require information to function. And the people with the best information are contestants.

So contestants - or people close to them - have a financial incentive to leak spoilers. Bet on the correct outcome, then make sure that information gets to market makers. Profit.

For a reality show, this is existential. The entire product is suspense. If outcomes leak ahead of broadcast, viewing drops. Ad revenue drops. The show dies.

Probst, who has hosted Survivor for over 40 seasons, told Variety that prediction markets are "incentivizing people to lie, cheat and steal" - ironically, the show's tagline. But this isn't part of the game. It's external corruption of the game itself.

Kalshi is reportedly considering measures to prevent spoiler-based trading. But that's nearly impossible to enforce. How do you prove someone's bet is based on inside information vs. analysis? How do you distinguish a lucky guess from a leak?

You can't. Which means prediction markets will always incentivize insider trading when they allow betting on events with non-public information.

Survivor spoilers are trivial in the grand scheme. But the same mechanism threatens elections, sports, and any other event where insider information has value.

We've already seen this with Polymarket during elections. Traders with access to internal polling or campaign information could profit by betting before that information becomes public. The line between "prediction" and "insider trading" blurs.

Regulated markets solve this with insider trading laws. If you have material non-public information, you can't trade on it. Crypto prediction markets like Polymarket operate in regulatory gray zones where those rules don't clearly apply.

Kalshi is regulated by the CFTC, but even they're struggling with how to prevent information-based manipulation on entertainment markets.

The deeper issue is that prediction markets change the incentives around the events they're predicting. When there's money in outcomes, people with influence over those outcomes will be tempted to manipulate them.

In sports, this is match-fixing. In elections, this is polling manipulation. In reality TV, this is spoiler leaking. Same problem, different contexts.

Prediction market advocates argue these platforms aggregate information efficiently and produce accurate forecasts. That's true when participants are betting on public information and analysis. It breaks down when participants can profit from controlling or leaking non-public information.

Survivor isn't going to kill prediction markets. But it's a canary in the coal mine. If these platforms can't prevent spoiler trading on a reality TV show, how are they going to handle prediction markets on elections, regulatory decisions, or corporate earnings?

The technology is impressive. The question is whether anyone needs it.

That's my catchphrase, and it applies here perfectly. Prediction markets are technically interesting. But if they corrupt the events they're predicting, they're net negative for society.

Jeff Probst is fighting for his show's survival. The rest of us should be watching how this plays out, because the same dynamics will apply to things that matter more than reality TV.