Student Forecasting Tournaments

Explore Decision Education concepts through a gamified learning experience

Strengthening Forecasting Skills in Students—Because Every Decision is a Forecast of the Future

We are finding ways to integrate Decision Education into classrooms across the country. One such initiative is our Student Forecasting Tournament—an exciting, gamified learning opportunity for students to practice forecasting about real-life events and improve their decision-making skills.

Students submit their own forecasts about various questions about current, real-world events. At the end of the tournament, the students with the most accurate forecasts based on their Brier score (a method of calculating forecasting accuracy, described below) win!

What is Forecasting?

Forecasting involves making predictions about a future event or outcome based on historical and current data.

Why Should We Teach Students About Forecasting?

All decisions are a forecast of the future because they involve thinking through the likelihood of certain outcomes using information that’s available. Students practice real-world skills and dispositions through forecasting, such as: exploring base rates, finding and evaluating sources for data, updating their beliefs, resisting cognitive biases, and thinking probabilistically. These are all reflected in the Decision Education K-12 Learning Domains.

Frequently Asked Questions

Using current events, students learn the skills to test, bring to life, and sharpen their decision-making skills. When they learn how to predict the outcome of future events, they learn to explore an evolving world in a structured way—developing the skills and dispositions to update their beliefs based on new information. Forecasting tournaments provide a gamified way for students to engage with Decision Education—and it’s fun!

Students answer questions about current events with various focuses—like sports, pop culture, the arts, the economy, and more! Students choose from multiple possible outcomes for each question, and they provide their forecast and rationale for their answer. An example question could be: “How much money will Inside Out 2 gross on its opening weekend?” Students would look at the various outcome options for the question—in this case: under $40 million, between $40 million and $65 million, or more than $65 million—and provide their forecast and rationale for their selection.

The Brier score was originally proposed to quantify the accuracy of weather forecasts, but it can be used to describe the accuracy of any probabilistic forecast. The Brier score indicates roughly how far away from the truth a forecast was.

The Brier score is the squared error of a probabilistic forecast. One way to calculate a Brier score is to first calculate the forecast probability, which may range between 0 (0%) and 1 (100%). Then, code reality as either 0 (if the event did not happen) or 1 (if the event did happen). For each answer option, take the difference between the forecast and the correct answer, square the differences, and add them all together.

For example: For a yes/no question where someone forecasted 70% and the event happened, the score would be (1 – 0.7)² + (0 – 0.3)² = 0.18. For a question with three possible outcomes (A, B, C) where they forecasted A = 60%, B = 10%, C = 30% and A occurred, their score would be (1 – 0.6)² + (0 – 0.1)² + (0 – 0.3)² = 0.26. The best (lowest) possible Brier score is 0, and the worst (highest) possible Brier score is 2.

Students with the lowest Brier score (i.e, the most correct predictions with the highest level of confidence) win.

 

Stay informed and join our mailing list