Why We Need Prediction Markets Now More than Ever
Shutterstock
Main page Opinion, Blockchain, Market
Hot topic
Dec. 17, 2018
63 63

Information overload is such a ubiquitous problem in the 21st century that it’s turning into white noise. Everyone knows they’re faced with more information than they could possibly keep up with, but the issue is so obvious it can seem as if it’s hardly worth mentioning anymore.

That doesn’t mean it’s not problematic. The Harvard Business Review reported that information overload costs the United States economy $900 billion a year — and that was back in 2009. The publication also reported that “knowledge workers” switch tasks every 3 minutes and that email causes stress for 40 of knowledge workers.

It’s not just productivity that’s suffering. The information needed to decipher crucial metrics such as public opinion is now scattered across the internet in the form of tweets, Facebook posts, blogs, and obscure comments in sub-sub Reddit feeds. The days of reading a Gallup poll to learn what your nation thought about something are over. Instead, companies are relying on complex algorithms to aggregate information and determine what people think or want. But considering the extraordinarily vast amount of information on the Internet, there may be no good way of knowing whether the algorithms are actually pulling the most representative set of data; considering that we only give computer programs a limited amount of time to run before we expect an output, even the most sophisticated code can’t possibly process all the information out there.

This puts society in a bit of a bind. While individuals may be accustomed to living their lives surrounded by uncertainty — I have no idea what most of the people around me are doing or thinking, and frankly I’m not that interested anyways — organizations such as governments or corporations need reliable metrics about these things if they are to have any hope of meeting the population’s needs.

Fortunately, un-processable amounts of information are not a new problem, and we already have a solution that’s stood the test of time for centuries: markets.

Markets take all the information about people’s needs, values, wants and resources — more data than a central party could ever hope to understand, especially considering the subjectivity of wants and values—and aggregates all this information into a single metric, called a price. It’s just another instance of the information overload problem, and this one was solved centuries ago, if not more.

The question, then, is whether markets offer a compelling solution for our specific information overload problem in the 21st century. I think the answer is yes, and will highlight three characteristics of markets that seem particularly relevant for the digital age: scalability, adaptability, and productive self-selection.

First, markets are easily scalable, in the sense that the amount of work needed to create and maintain a market does not increase proportionally with each additional participant in the market. Once a market has been created (or developed organically), participants can enter simply by buying or selling shares, goods, or products. If there is enough liquidity to match new entrants’ orders, then the party maintaining the market, if there is any, hardly has to put forth any marginal effort per new market participant. In the case of a centralized market, there may be some amount of administrative work involved, but digital platforms and automation are making even this work much cheaper and faster than in the past.

Contrast this with polling, another popular form of information gathering. Many polls still involve calling a large number of people, and hoping that a good chunk of them answer the phone and take the time to answer your questions. The amount of effort the pollster puts forth increases more or less linearly with the number of people polled, because each person has to be called individually. (In fact, it’s more likely that the amount of work increases exponentially for each response received, considering that the pollster likely has to call several people for every one response.) Because of this, polling is not scalable. In the digital age, where there is more information out there than we can wrap our minds around, and the amount of information is growing every single day, we need an information gathering tool that can scale quickly and easily.

Of course, there are more scalable versions of polling, such as online surveys, where the marginal effort needed to send the survey to an additional person is very small. But markets also have an advantage over these, because they are adaptable. By this I mean that their output, the price, changes in real time to reflect changing opinions. I might buy shares of public company XYZ this week, thinking that the price is going to increase on a coming round of strong earnings reports. But what if I new information becomes available and my opinion changes? Then I simply sell my shares of XYZ, and the market changes to reflect my changing opinion. Therefore, markets capture not only the nuances of each individual participant’s specific circumstances at a given time, but also the complexity of people’s constantly changing opinions. Other types of information aggregation such as polling or surveys can only capture public opinion at a particular moment in time, which is a significant shortcoming in our constantly changing information landscape.

Finally, markets allow for productive self-selection. Because markets are based around economic incentives—you generally make money if your prediction is right and lose money if it’s wrong—people have much stronger incentives to participate if they are confident in their opinions, a confidence that often comes from having the necessary information to make an informed decision. The people with the best information are the most likely to enter a market. Again, this differs significantly from other information gathering methods which depend on the ability to select a representative sample of the populations, even though the person or program selecting the sample has extremely limited information about each individual person. In a world where billions of people are connected to the Internet, and online, communities can be extraordinarily specific and granular. It’s hard to understate the value of a tool which inherently elevates the best-informed opinions.

None of this is new. We’ve been using markets to set prices for goods and services for centuries, while companies list their stocks on some of the most famous markets in the world, such as the New York Stock Exchange (NYSE). But markets can be used to gather other sorts of information as well. Some of the most famous prediction markets have been used to predict election results, but they can also be used to gather expert information about topics ranging from predicting the weather to intelligence work. It’s this ability to aggregate information on even seemingly random topics that makes prediction markets so relevant to our current digital information overload.

Which brings us to prediction markets’ latest technological update: the blockchain. Already, teams such as Augur and Gnosis are using distributed ledger technology to build prediction market platforms that are essentially uncensorable, and which use smart contracts to ensure that market participants are always paid out what they’re owed. A new project called Amoveo claims to have made significant progress in oracle design — the mechanism that allows blockchains and, by extension, smart contracts to integrate information about real-world events.

In the rest of this series, we’ll explore how blockchain technology can usher prediction markets into the 21st century through the creation of decentralized, uncensorable platforms for aggregating information on a range of important topics.

Read also:
Please describe the error
Close