Prediction markets as information processors or opinion measures

What does a market price tell us? A says it combines many people's information quickly and efficiently. Rather than taking everyone's stated opinion at face value, a market asks people to put money behind it. Someone can loudly praise a company, but being willing to buy its stock at the current price gives that judgment a different kind of test.

The writer emphasizes the incentives: markets reward being right rather than sounding persuasive, being aggressive, or having impressive credentials. That is the favorable position to track. It doesn't yet mean that every price must be correct or that markets solve every kind of problem.

The Iowa election market gives us a concrete example. Contracts pay a dollar if their candidate wins. A price of 53 cents is interpreted as a 53 percent chance of victory. That is a probability, not a guarantee. Suppose you think the candidate's real chance is 70 percent. A 53-cent contract would look attractive to you even though you still think the candidate could lose. The purchase expresses your judgment that the price understates the chance; it doesn't require certainty or promise risk-free profit.

A then explains what efficiency means here: the market rapidly incorporates what informed people know. Imagine news that damages a candidate's chances. Traders who learn it have a reason to sell before the price falls, and their trades help bring the information into the price.

The fictional-stock experiment shows that process without requiring a public announcement. Two of twelve traders learn the stock's value. Suppose, just to make the mechanism concrete, that it is $5. They have a reason to buy below $5 and sell above it. Other participants see the price move and adjust their behavior, even though the insiders cannot speak to them. Information that began with two people affects the whole market through trading.

Main point of A: Markets efficiently combine participants' information and rapidly transmit informed judgments through prices.

B opens with a challenge: markets aren't infallible. Its election example describes an eventual winner whose contracts traded below an opponent's for most of the preceding year. B uses that outcome to question the market's predictive ability.

We should notice a possible gap in that challenge. A low-probability event can occur. If something has a 10 percent chance, its happening doesn't alone prove the estimate was wrong; ten percent isn't zero. Nor does A claim that traders can incorporate information before anyone has it. Those distinctions help us evaluate B's criticism without losing track of what B argues.

The racetrack comparison explains changing expectations. A horse that appears unhealthy offers a larger potential payoff because victory looks less likely. When it looks healthier, the odds adjust. B sees the same process in markets: prices reflect people's views at the time.

The final paragraph draws B's conclusion. Prediction markets are essentially opinion polls with a financial entry fee. B concedes that paying encourages more serious responses, but denies that seriousness confers special foresight. People can still be wrong. A useful response is that an informed minority might trade against a mistaken majority, changing the price; a market need not function like a simple head count. That is the information-processing possibility B's description downplays.

Main point of B: Prediction markets mainly measure current opinion, and their financial incentives do not give their forecasts special reliability.

Both writers recognize that money and new information affect markets. Their disagreement is about what the resulting price achieves: A emphasizes the rapid spread of knowledge; B reduces it to a report of current belief.