Predictit.org - Crowdsourced Political Prediction Market

Predictit.org - Crowdsourced Political Prediction Market

With the US presidential election this fall, there has been a large interest in predictit.org - this is an authorized prediction market on the topic of political questions and polling. The participants deposit funds using debit or credit cards, and after participating in the market for 30 days or more, are able to withdraw the funds if they don’t want to continue in the market.

There are a number of one-time and weekly revolving questions to estimate the result of. The questions are either yes/no or bracket format (one of 5 outcomes, etc).

The typical polling question markets generally use a third party political analysis website, realclearpolitics.com , as source of the question, and as the guage of the outcome.

Many participants utilize hedging models to minimize their own risk, and leave the potential for outsized gains on non-hedged traders. Some dedicated participants claim regular weekly profits from their participation - and the longevity depends on learning the patterns of political data that will be utilized in the resolution of the prediction question.

The popular weekly markets include:
“Obama Approval (OA)”
This week’s OA market link is here:
https://www.predictit.org/Market/2514/What-will-Obama's-RCP-average-job-approval-be-at-end-of-day-September-30
It is based on the outcome of data reported here:
http://www.realclearpolitics.com/epolls/other/president_obama_job_approval-1044.html

Today’s data is in this table:

The market is trading on the likelihood that the “RCP average” in the top row of the table, in the “Approve” column, will be a figure that falls into one of the PredictIt brackets (e.g. B1, B2, B3 …) at market closing time of 11pm ET Friday September 30th.

There are many potential outcomes, and they are driven by the new data points entering the equation used by RCP, along with RCP’s judgement of which older data points are no longer relevant, and are dropped from the equation.

The traders that do well have learned the frequency of the new data additions and have developed a sense of which older data points will be dropped.

I participate in this market to some degree, and am blogging to share what I’ve learned. Anyone else here participate? Thoughts on how to improve the process using a blockchain?