Why do we make predictions? From career changes to family planning, predictions help us to grapple with the uncertainty of what can and will happen in the future. At the individual level, they are fundamental to the way we learn to see and interpret the world. As we encounter stimuli such as images and sounds our brains are constantly creating micro-predictions based on prior encounters and contextual information. These predictions are important in helping us to notice when there is something unexpected in our environment, so that we adapt our responses accordingly. And so we don’t step out onto the road when we see a car approaching.

When it comes to longer-term predictions about events that affect entire communities or even nations, governments also turn to predictions to manage uncertainty and, most importantly, take action. As COVID-19 continues to disrupt the world, we see forecasts and predictions being used to model the spread of the disease, understand the potential impact of policy interventions and even anticipate the next large scale outbreak.

The coronavirus pandemic is one reminder that the world can suddenly change, but just twelve months ago, the UK was facing an altogether different source of uncertainty. Fresh from the extension of Article 50 at the end of March 2019, parliamentary debates had become increasingly fraught and two new political parties were gaining momentum in the run-up to the UK's European parliamentary elections. Looking back, Brexit may be but a distant memory, but it helped to define many events in 2019, from politics to the economy.

Who could have predicted that by December 2019 the UK would have seen yet another extension to Article 50 and a large Conservative majority in a General Election? One surprising answer is that over the course of 2019, a global crowd of more than 7,500 non-experts forecast these outcomes...as well as many others. Between them, they submitted more than 33,000 predictions in response to 19 questions about Brexit, technology, health and more. This crowd of forecasters were part of a year-long predictions challenge we hosted in partnership with BBC Future.

This three-part series concludes our experiment with crowd predictions and presents the final results. In the first part, we cover the headline findings and what they could mean for the future of decision-making. We give a full question-by-question overview of the crowd consensus results in the second part of our series, as well as what we can learn from where the crowd got it wrong. The third and final part presents the lessons learned and gives our top tips on setting questions for crowd predictions.