If you’ve never heard of freestyle slalom skating before, prepare to be amazed. In this sport, roller skaters skillfully maneuver their way around a set of tightly spaced cones at mind-boggling speeds, balancing on one skate, skating backwards, or sometimes both! If you ever want to watch some impressive feats of athleticism and artistry on YouTube, it’s worth a watch (this video and this one are good examples).

While it’s easy to appreciate the athleticism involved, there are also some impressive visual abilities behind these moves. To maintain their balance and gracefully pivot themselves at just the right moment, skaters rely on something called optic flow.

Optic flow is the pattern of motion in images that our eyes receive as we move through the world. This is something that you can easily appreciate whenever you’re in a moving car. You’ll notice that as you move, the objects around you stay still, which means that the images of the trees and lane markings appear to fly past you. Your brain then interprets this pattern of motion from your eyes to help you figure out the direction in which you’re moving and to help you avoid obstacles. We use small changes in optic flow for maintaining our balance as well, as shown in this classic video by David Lee . In other words, optic flow information isn’t unique to roller skaters (or even humans), as it’s used by many organisms, and even by robots and autonomous vehicles.

It’s clear that optic flow is an important source of visual information, but are we equally sensitive to all optic flow patterns, or are there certain visual features that need to be in these patterns of motion for us to notice them? For example, are we more sensitive to optic flow patterns that we’re used to seeing, compared to ones that we’re less familiar with? These are the questions that Li Li, Xuechun Shen, and Shuguang Kuai (pictured below) investigated in a recent paper in the Psychonomic Society journal Psychonomic Bulletin & Review.

To explore these questions some more, we know that not all optic flow patterns are exactly the same. When we move backward, we see a contracting pattern of motion, like the image below on the left. When we move forward, we see an expanding pattern, as shown on the right. Another thing to notice is that the speed of motion shown in the pictures below (shown by the lengths of the lines) is similar to what you would get if you were moving through a real three-dimensional environment. For example, the speeds are slower close to the center (the point from which the image expands).

To investigate our sensitivity to these different features, the authors of the study showed participants different optic flow patterns, as shown in the image below. In one task, the researchers added different numbers of randomly moving dots to these patterns to make them harder to detect (for example, half the dots are moving randomly in the bottom row in the image). Participants reported whether they saw any coherent motion within those moving dots.

To figure out whether we’re uniquely sensitive to motion patterns that we see in the world, in one set of conditions, the authors used “real” optic flow patterns (the two rightmost columns), which are similar in direction and speed to those that we’d get from the geometry of a real environment. In another set of conditions, participants saw “unreal” optic flow patterns (the left column), which don’t have those properties. For example, the speeds were the same throughout the pattern or randomly assigned across locations.

How did participants’ performance change across these conditions? For “real” optic flow, participants were better at noticing the motion when looking at expanding patterns compared to contracting ones. For “unreal” optic flow, the reverse was true. Participants were actually better at detecting the motion when it was contracting compared to expanding, as shown in the figure below.

In a follow-up experiment, the authors also looked at how stimulus duration affected performance when judging the center of the dots’ motion. Interestingly, for real optic flow, performance improved when the dots were shown for a longer duration (400 ms compared to 100 ms), but for unreal optic flow, performance was similar between the two durations.  These results indicate that the brain handles these two types of motion stimuli differently.

These differences in sensitivity to “real” compared to “unreal” optic flow may be a product of our visual experience with the world. We’re used to seeing certain patterns of motion as we move through our environment, and we’re used to the specific ways in which speeds vary. Not only that, but whenever we move forward through our environment, we see an expanding optic flow pattern. We’re much less accustomed to seeing contracting patterns (though you might think of some exceptions – like skating backwards in a freestyle slalom trick!). This could make us uniquely sensitive to certain optic flow patterns, particularly when the speed profiles match those that we normally see in natural, 3D environments. According to Li Li,

“These differences reveal influences of self-movement in natural environments, enabling the visual system to uniquely process stimuli with significant survival implications.”

These results also have many potential applications outside the lab, particularly for the design of games and virtual reality environments, where the goal is to create a realistic or immersive setting. By matching the optic flow to what we experience in everyday life, designers of virtual environments can potentially capitalize on this sensitivity, making them more effective. They could even make the next slalom skating video game for unathletic folks like me!

Featured Psychonomic Society article 

Li, L., Shen, X., & Kuai, S. (2025). Distinct detection and discrimination sensitivities in visual processing of real versus unreal optic flow. Psychonomic Bulletin & Review, 1-11. https://doi.org/10.3758/s13423-024-02616-y