#预测市场 Seeing this article, the first thought that flashed through my mind was—I've seen this scene before.

In 1905, The Washington Post discussed the possibility of manipulation in the betting market. Over a hundred years later, we are still debating the risks of prediction market manipulation, just with a different name and a different technological shell. The mysterious surge in Romney's stock price in 2012, and the huge fluctuations in Trump's prices on Polymarket in 2024, each time trigger the same questions—are these real signals or is someone pulling strings behind the scenes?

Let me be honest, market manipulation is actually much more difficult than people imagine. Rhode and Strumpf's research on electronic markets in Iowa proved this—traders attempting to push prices higher end up losing everything because arbitrageurs swarm in immediately, pulling the distorted prices back to reality. Markets with high liquidity are like alert predators; any crude manipulation attempt will be quickly identified and countered.

But this time is different. When CNN begins to broadcast prediction market prices in real-time, and these numbers spread through countless social media accounts, the informational function of the market and its political leverage undergo a qualitative change. A seemingly minor price fluctuation, once amplified and interpreted by news organizations, can cast a huge shadow in the public psyche—even if this fluctuation itself doesn't change the election outcome, it alters people's perception of democratic fairness. That is the real threat.

I recall the strange phenomena before the 2016 Brexit referendum. Polls showed a lead for remaining in the EU, but the turnout for Brexit supporters was unusually high, while support for remaining was surprisingly low. Some believe this was a self-contradictory signal—when people think the outcome is decided, they relax their vigilance. If prediction markets are used as such tools, the issue isn't whether they can predict accurately, but whether they can be used as a covert means of manipulating public opinion.

The good news is that solutions are already on the table. Setting liquidity thresholds, publishing transparency indicators, and regulatory agencies responding quickly to abnormal trades—all can significantly increase the cost of manipulation. The key is that broadcasters and platforms need to truly implement these suggestions, rather than blindly reporting on volatile markets with low liquidity just to attract attention.

History tells us that technology and institutions always advance through mutual struggle. Prediction markets themselves are not inherently bad—especially in the AI era, when polls are becoming increasingly unreliable, this mechanism of aggregating dispersed information becomes even more valuable. We just need enough maturity to protect their informational value and to build defenses against abuse. This contest is far from over, and the next few years will be very interesting.