Forecasting is integral to decision making. All decisions implicitly involve a view of the future, something that is getting progressively difficult to do with any level of certainty.
The good news is that humans can build good habits for better forecasting. As Tetlock and Gardner detail in Superforecasting, forecasting can be learned. They offer simple tips for better forecasting habits, for example:
- Make some “best guesses” and assign a probability (or range) to each.
- What information is needed? Is it better to waste time predicting the unpredictable or fail to predict the predictable?
- Compete against your own ideas: list the things that would make you believe the opposite side.
- “Gisting” —have everyone write a short form description of cause and effect and use it identify clashing causal forces.
- Use numbers not words. “Maybe” is useless. “Probably” is confusing. Ask for a probability instead.
The best forecasters have good habits and develop reliable intuitions and have a well-calibrated confidence in their sense of what’s known versus unknown. But how can you put this in practice in group decision-making?
The optimal place to start practicing these tips is in developing deviations from a base-case forecast. But rather than predicting high and low scenarios by reasoning forward from cause to effect, reason backwards and diagnose alternative causes for the effect. Humans need causality to spot errors; reasoning forward is error-prone because we fail to see alternative causes. Reasoning backwards is far less error prone and is perhaps uniquely human.
Next time you are presented with a base case forecast, ask “what would it take for this outcome to be x% better, y% worse?” The process of working backwards from effect to causes, related to the similar process of a pre-mortem, is a highly effective technique for making better decisions.
This article is one in a series of hacks, tips and tricks for making better decisions.