Estimating Central, Peripheral, and Temporal Visual Contributions to Human Decision Making in Atari Games

TL;DR

This study quantifies how peripheral vision, gaze, and past states influence human decision making in Atari games, revealing peripheral info's dominant role through a novel ablation framework.

Contribution

Introduces a controlled ablation framework to dissect the contributions of different visual information sources in human gameplay decisions.

Findings

Peripheral information has the strongest impact on decision accuracy.

Gaze information contributes minimally to decision prediction.

Different behavioral regimes are identified based on information source reliance.

Abstract

We study how different visual information sources contribute to human decision making in dynamic visual environments. Using Atari-HEAD, a large-scale Atari gameplay dataset with synchronized eye-tracking, we introduce a controlled ablation framework as a means to reverse-engineer the contribution of peripheral visual information, explicit gaze information in form of gaze maps, and past-state information from human behavior. We train action-prediction networks under six settings that selectively include or exclude these information sources. Across 20 games, peripheral information shows by far the strongest contribution, with median prediction-accuracy drops in the range of 35.27-43.90% when removed. Gaze information yields smaller drops of 2.11-2.76%, while past-state information shows a broader range of 1.52-15.51%, with the upper end likely more informative due to reduced peripheral-information leakage. To complement aggregate accuracies, we cluster states by true-action probabilities assigned by the different model configurations. This analysis identifies coarse behavioral regimes, including focus-dominated, periphery-dominated, and more contextual decision situations. These results suggest that human decision making in Atari depends strongly on information beyond the current focus of gaze, while the proposed framework provides a way to estimate such information-source contributions from behavior.