While the prediction market mechanism can be a fun, interesting way to generate crowdsourced forecasts, there are cases when other methods of eliciting forecasts are desirable. It can seem inappropriate or insensitive to be betting on certain topics. Or participants can be overwhelmed or uninterested in the technical aspects of trading in a prediction market. Fortunately, there are other mechanisms for doing crowdsourced forecasting.
One common crowdsourced forecasting tool is known as an Opinion Pool (also known as an opinion poll or survey platform). In an opinion pool, participants do not buy or sell stocks like a prediction market. Instead, participants are asked to give a probabilistic estimate for the likelihood that an event will occur. For example, say you ran an opinion pool for the question "Will Apple buy Tesla in 2017?" Participants would respond by submitting a probability. So if I think it's unlikely that Apple will buy Tesla, I might submit 5% as my forecast.
An opinion pool platform will take these probability forecasts from a large group of users and aggregate them, generating what's known in opinion pool parlance as the consensus. The algorithm for aggregating forecasts into a consensus can be as simple as averaging all forecasts that users submit, or it can be something much more complicated. Some additional aggregation algorithms include mean, median, voting, logit, and l2e.
Go back to The Ultimate Guide to Crowdsourced Forecasting and Prediction Markets.