Forecasting, Science, and Epistemology

4 min readMay 7, 2022


By Gaia Dempsey, CEO of Metaculus

At Metaculus, our mission is to build epistemic infrastructure that enables the global community to model, understand, predict, and navigate the world’s most important and complex challenges.

Why do we focus on building “epistemic” infrastructure? To answer that question, we’ll dive into what epistemology is, and why it matters in forecasting and in the world.

Epistemology is a branch of philosophy that deals primarily with human knowledge and beliefs. It investigates questions about how we come to believe things are true, how we use reasoning to justify our beliefs, and how it becomes possible to hold beliefs that are rational.

Whether we realize it or not, we are epistemology machines. We are constantly evaluating information about the world and determining what we consider likely or unlikely to be true. When we consider the evidence for the efficacy of a new vaccine, weigh the promises made by politicians, or decide how much to trust the latest headlines from our favorite (or least favorite) newspaper, we are processing new data, adding it to our current model of the world, and updating our beliefs in an attempt to align them with what is actually true. This is part of the normal functioning of the human mind.

Epistemic Uncertainty. Source:

But in spite of the sophistication of the finely tuned evolutionary apparatus that is the human brain, and the best efforts we put forward to try to understand whether it’s safe to get vaccinated, or who we should vote for, or whether we agree with the conclusions of a persuasively argued op-ed, sometimes we come to hold beliefs that have no basis in reality. Through faulty reasoning, or falling victim to effective propaganda, or just simple misinterpretation of the facts, our model of the world can easily be distorted by biases, blind spots, or worse, entire mirages.

Flawed Data. Source:

Paying attention to epistemic processes is not only useful in helping us figure out how much trust to place in information we get from the world. Intentionally improving our own epistemics also helps us to investigate — and mitigate — the ways in which different cognitive biases can color the way we see reality.

Fortunately, there is a system to check for and correct biases in our beliefs: this is the grand purpose of the scientific endeavor. That we should test the validity of our beliefs by making predictions about the world and then checking to see whether they are right or wrong is the fundamental insight at the heart of the scientific method. It is predictive accuracy that is the gold standard for justifying our trust in a particular belief, model, or theory about the world.

Prediction. Source:

Validating theories according to their predictive accuracy has been extraordinarily successful in advancing our understanding of the physical world. Thanks to thousands of scientists working over many decades, we have seen enormous leaps forward in our understanding of physics, biology, medicine, space, computation, neuroscience, and many other fields. Commercializing, productizing, and otherwise exploiting the insights generated by centuries of scientific exploration has left us standing at the precipice of the impossibly sharp peaks of progress that industrialized societies occupy today.


But, can we bring good epistemics into other important fields of human endeavor, transferring knowledge from the natural sciences to the realms of policymaking, economics, and global media?

That is exactly what we aim to do, and is the inspiration for our mission.

Today, we enact our mission mainly in three ways:

  1. Providing Forecasts as a Public Service: We freely provide aggregate forecasts on a range of important topics, maintaining a track record of accuracy for each individual and for the community as a whole.
  2. Fostering a Global Forecasting Community: We provide a forum for an active, global forecasting community to gather, learn, practice, and grow.
  3. Supporting Forecasting Research: We help advance research and innovation in the field of forecasting science by providing access to data and collaborating with researchers in the field.

And we’re just getting started. Whether you’re interested in understanding where the world is going, or making quantitative forecasts yourself, we invite you to join us in our journey at