What is it about?
Human decisions often seem random. Even simple decisions like what food to order at a restaurant can be difficult to predict ahead of time. One outstanding question is where the randomness in our decisions comes from. Sometimes, our seemingly random decisions are driven by predictable external factors, like what the guests at the next table order could influence what we order. Other times, our decisions are not driven by external factors but are instead made by random thoughts within our brain. How can we tell whether a seemingly "random" decision is actually random? In this paper, we developed a novel computational method to quantify the extend to which our seemingly random decisions are driven by predictable external factors, versus being actually random and unpredictable.
Featured Image
Photo by Robert Stump on Unsplash
Why is it important?
The randomness in our decisions can actually be beneficial, effectively allowing us to explore new options by chance. For example, picking a random item out of the menu might actually lead you to your new favorite dish. Understanding the source of behavioral variability is thus key to understanding the nature of exploration. By quantifying and studying the roles of both external/predictable (e.g., my dinner decision is driven by what the guest at the next table ordered) versus internal/unpredictable (e.g., I just felt like eating pizza today) sources of randomness in decision-making, we provided new insights on how humans flexibly adapt the level of randomness in their behaviors in the service of exploration.
Perspectives
The beauty of this work lies in the method used to decompose randomness in behavior into predictable versus unpredictable components (check out the 18 supplemental figures used to validate and test the limit of this methodology). Computational modeling is both necessary and beautiful in this work.
Siyu Wang
University of Arizona
Read the Original
This page is a summary of: Separating random and deterministic sources of computational noise in explore-exploit decisions, PLoS Computational Biology, March 2026, PLOS,
DOI: 10.1371/journal.pcbi.1014026.
You can read the full text:
Contributors
The following have contributed to this page







