, , , , , , ,

English: A panorama of a research room taken a...

The DAGGRE prediction market during break. (Photo credit: Wikipedia)

Thanks to effective participation from all of you volunteers in the DAGGRE prediction market, we were awarded the second year (Y2) option on the DAGGRE project, and we are now in the second month of Y2.

We learned many lessons in Y1 which we will apply in Y2 to improve your experience and improve our ability to operate prediction markets.

One of the lessons we learned in Y1 was that we were not always clear about the participation conditions and how those are coupled with the research year. At the outset of Y2 we want to clarify this, as well as giving you information about our aims and approach for Y2.

First of all, we are funded year-by-year, starting on 26 May each year, for up to a total of four years. So Y2 began 26 May 2012.

Each April, IARPA makes funding decisions depending on performance, available funds, and their own priorities. Each year we close out the budget and volunteer participation protocols and start again.

In Y1 we used a fixed participant compensation approach beginning in September 2011. Thus the wording “up to 7 months of compensation” applied to those volunteers who joined in September and continued to actively participate through April 2012. People who joined after September 2011 had correspondingly fewer months for which it was possible to be compensated.

We realize that we should have been clearer about the research year timeline, and that our approach to incentivizing volunteers is likely to change from year to year. We apologize to participants who were confused by this omission.

For Y2, we made several modifications to our compensation scheme. First, due to the success of the participant workshop we held last March, we decided to shift some funds from direct participant payments toward workshops. In Y2 we have budgeted for two workshops. These provide a much better way to sharpen your skills, learn about forecasting, meet forecasters and researchers, and provide feedback. Second, for budgeting purposes, we needed to control the extreme variability in monthly distributions to participants. This is why we moved to giving a fixed number of awards each month to randomly chosen participants, rather than issuing payments to everyone who met our participation threshold. Finally, we realized from our Y1 experience that we needed to allocate more funds to software development and participant support. Although this has reduced the available funds for compensating participants, we expect that you will see big benefits in more stable, bug-free software and faster response when you need to contact us.

The Y2 award approach is each month to award sixty $50 Amazon Gift Cards delivered by email as electronic codes. Using gift cards will avoid the delays some of you have experienced due to administrative difficulties in processing payments through the university’s financial systems. To keep things fair, the awards will be randomly selected from among those volunteers who actively participated in the previous month. Your chances depend on the number of active participants, and on your activity level. As you participate actively over multiple months your number of “tickets” steadily increases, and resets when you win.

We are selecting June’s recipients this week, and will announce them by username when vetted.

We apologize sincerely for the delays some of you experienced on account of our administrative problems, and we hope these delays are a thing of the past. We now reply to emails, calls, or helpdesk tickets within one business day. Our participant coordinator is finally free from Y1 financial issues and is on call to attend to your questions and issues. The electronic Amazon Gift Cards will streamline and expedite our payment process. So overall, we hope your experience this year will go much more smoothly than previously.

We want to emphasize that this is a research project. Our goal is to improve forecasting, adapting as we learn.

Frankly we see the gift cards as a secondary, fun incentive for participation. We hope that the main reasons you participate are because you want to learn more about prediction markets, because you are excited about the issues that are addressed in the online market questions, and because you wish to further our country’s ability to forecast the future of important events and therefore be better able to focus and be appropriately responsive in a world of real-time, global communication.

Learning, we hope, is a major motivation for your participation. Therefore we are improving our tutorials, planning workshops, and increasing feedback, all while pushing the research frontier.

We want participants who are passionate about Daggre. Our success depends on your enthusiasm about the material, your drive to improve your standing on the leaderboard, your efforts to improve your own performance and your interest in furthering this important area of research.

Thanks for sticking with Daggre. We look forward to an active and successful Y2!