2026 Forecasting Contest
A forecasting contest for 2026, featuring UK politics and global events
This is the Edrith annual Forecasting Contest, featuring UK politics and economy, global politics and world events; and technology, the arts and miscellany.
Last year’s contest had 374 entrants - the largest UK-focused forecasting competition of its type I’m aware of. It was featured in the i newspaper, attracted some highly respected names from politics, journalism, think-tankery and academia - and was won by an operations manager at TfL, showing that you don’t need to work in politics to do well! Let’s make this year’s even bigger and better.
Read on to find out more and enter.
The contest encourages and rewards people for not just getting things right, but for quantifying how likely something is to happen, so that you can see how well-calibrated you are - and ideally get better at making predictions.
Enter the 2026 Forecasting Contest here.
The contest consists of 50 questions, with 25 on UK politics and economy1; 15 on world politics and global events; and 10 on technology, entertainment and miscellany.
For each event, you have to say the probability (represented as a percentage between 0 and 100) of it happening. The aim is to predict how likely each event is to happen. The goal is not to predict everything perfectly, but to be well calibrated; in other words, if you predict 10 events have an 80% chance of happening, and 8 do, that is a good result.
For each question, you simply have to enter a number between 0 and 100 - you can take as much or as little time as you want.
You do not have to answer every question.
To enter the contest, click on the link here.
I have also recorded my own answers and registered them with a friend to ensure I don't change them. I'll publish these on this site, as well as the average result for each score, once the contest has closed.
Scoring will be by Brier Scoring. You don’t need to understand this to take part - the key thing to know is that it is a scoring system that rewards you for guessing what you genuinely think: In other words, if you think something is 70% likely to happen, write 70 - it’s as simple as that.
A few tips:
Don’t be overconfident. Brier scoring can punish you brutally for unwarranted overconfidence.
Don’t be afraid to miss out questions. In one year the person who came third missed out ten questions, showing a high score is possible even if you don’t answer every question.
Don’t guess what you want to happen, guess what you think will happen.
To get a sense of what happened last year, have a look at the results of the last forecasting contest.
This year’s results will be published on 1 January 2027. The prize is honour and glory.
The contest is open until 11:59pm UK time on Sunday 18th January. Please do share this link with others and encourage them to enter, even though strictly speaking this is not in your interests and will decrease your chance of winning.
Due to the advent of five party politics I have expanded this section by five questions, to enable more space for questions about the Greens, Reform and Liberal Democrats.

