Experimental Feature: Question Links
9
We’ve just released a new experimental feature that lets you connect forecasting questions that you think are causally related. You can now link questions to each other, noting the direction and strength of the causal impact (e.g. “large positive”, "small negative"). When questions are linked, updating one will prompt you to check whether your forecasts on the others should change too. You’ll also be notified when a question you’ve marked as causally upstream resolves. These links are **private for now**, intended to help you track dependencies in your own reasoning. They’re part of a broader effort to support [**reasoning transparency**](https://www.metaculus.com/notebooks/29025/what-metaculus-is-building-next/#reasoning-transparency-world-model-elicitation) on Metaculus, alongside existing tools like Key Factors. In future updates, we will aggregate question links across forecasters, and build a community-level network of linked questions. The goal is to make it easier to see how forecasts relate to one another and to build structured models of how we collectively reason about the future. You’ll find the new **“Add Question Link”** option under Key Factors on any given *binary* question page. More to come, soon, both for this feature and about the larger project this is part of: helping surface structured reasoning, and uncovering where our models – and assumptions – diverge. 