Skip to content

Forecasting

Forecasting is the simple practice of making numerical probabilistic predictions for specific single events in the future. Being able to accurately predict the future is a superpower. Many experts are bad at it. Journalists and pundits often don’t even try (with some exceptions). Instead, they hide behind terms like “probably”, “very likely”, “possibly”, and so on. So do most of us. We leave ourselves plenty of wiggle room.

Putting well-calibrated probabilities on subjective estimates can help you with all sorts of things. All of our plans in business, politics, even our personal life, are influenced by uncertain events in the future. Some we need to avoid, for others we can prepare, and still others we welcome with open arms. Having an accurate map of these events will help you to make good decisions.

I started forecasting on Good Judgment Open (GJO) in early 2019. I have since made forecasts on over 270 questions, ranging from “Before 15 November 2019, will there be a lethal confrontation between the national military forces of Iran and Saudi Arabia either in Iran or at sea?” to “Between 2 August 2019 and 31 October 2019, will the United Auto Workers (UAW) union go on strike against General Motors (GM)?” and “Who will be the last character shown sitting on the Iron Throne in the series Game of Thrones?”. Over 230 of these have already been resolved by now (October 2020). I have consistently ranked among the top ten forecasters out of thousands (usually at least in the 99.5th percentile).