The Georgian Institute of Politics (GIP). Find out more about GIP and its activities on the website.

Taken together, the views of various experts in most cases will be better than the views of any single expert. For elections, collecting views gives us a good estimate of likely outcomes.

Identifying the most likely scenarios (while being prepared for others) helps to focus the attention on where it matters. Good estimates can help people to plan, and be prepared.

We believe that reasonable people often can have divergent views on an issue, and that everyone wins if they listen to each other. In line with that, expert views show a diversity of views.

For our polls we try to recruit a range of opinions, from across the political spectrum, local and international. We reach out to these experts through a variety of means.

We typically try to engage 30-50 experts in our polls. This gives us a good range of opinions. More participants typically do not add much diversity.

If you have a track record of thoughtful contributions, do get in touch and let us know about your work.

We have a number of ways of cleaning the estimates, based on 10+ years of experience of working with polls.

Surveys are incredibly useful tools. For predicting election outcomes, however, they have several major limitations.

One of our role models is the website FiveThirtyEight by American pollster Nate Silver. His aggregation of polls has regularly yielded accurate results. We are adapting this model to the Georgian context.

This activity is currently funded by the National Endowment for Democracy, for when your project from October 1, 2016

Get in touch and let us know. We’ll be happy to answer your questions.