"Look where you're going, not at your feet" is the kind of trail-running advice that gets repeated so often it stops sounding like a claim. It's the same shape as "even effort, not even pace," and the newer idea — popular in minimalist and barefoot circles — that going shoe-light lets your feet "talk" to the ground and read the trail better. Three pieces of received wisdom, all about the same activity, all treated as roughly equally solid. They are not. One of them is a documented sensorimotor mechanism with real numbers behind it. One is half-true in a specific, checkable way. One doesn't currently have the evidence to back it up — and going looking for that evidence, rather than assuming it exists, turned out to matter.

Your eyes are running a two-step-ahead scan, and the timing barely changes with the terrain. In 2018, a team led by Jonathan Matthis put mobile eye-tracking and full-body motion capture on people walking real outdoor terrain — not a lab treadmill, not a video simulation — and measured exactly where their gaze landed relative to where their feet were about to go. The finding: gaze fixates on a rolling window roughly 1.5 seconds ahead of the current footfall, and that time window holds constant whether the ground is flat, moderately broken, or genuinely rough. What changes with terrain difficulty isn't how far ahead in time you're scanning — it's how much space that time window covers, which compresses automatically as footing gets harder and speed drops. A 2020 study on runners specifically (not walkers) on an urban obstacle course found the same anticipatory pattern: people scan ahead and commit, they don't dart their eyes reactively at the last second. "Look ahead, not at your feet" turns out to be a real, measured sensorimotor strategy, not just something a running-store employee tells you.

There's a wrinkle worth keeping, because it's the kind of thing that's easy to blur past. A different 2024 study, about people choosing between two photographed path options before they start walking, gets grouped under the same "gaze and terrain" heading — and it measures something else entirely. That study found gaze and route choice were driven more by how confident people felt about their own ability to handle a given terrain type than by how much visual information the path actually offered. That's a real finding, but it's about a decision made standing still, not about where your eyes go while you're already moving through the woods. Treating the two as interchangeable — "here's the research on gaze and terrain" — would quietly merge a study about confidence and choice with a study about real-time visual scanning. They're not the same question, even though Google will hand you both under similar search terms.

"Even effort, not even pace" is right about half the time, and the right half isn't the half people usually mean. Every ultrarunner has heard some version of it: don't chase a flat pace number, hold your effort steady and let the terrain dictate the speed. Three separate, large datasets — 13,829 Ultra-Trail du Mont Blanc results across twelve years, eighteen years of checkpoint splits from the Western States 100, and a smaller study of elite finishers at the World Mountain and Trail Running Championships — all point the same direction: runners who show less unforced pace variability finish faster. Not flatter pacing in some abstract sense — specifically, less surging, less blowing up, less of the up-and-down that isn't demanded by the actual grade. That's the real, well-supported half of the claim.

The half that doesn't hold up as well is "effort." Only one study in this research actually measured felt effort (not just speed) continuously through a real mountain ultramarathon — a 106 km race tracked by Hervé Kerhervé, Guillaume Millet, and Cyril Solomon in 2015. Even there, perceived effort rose over the course of the race while heart rate fell — which is not a flat line by either measure. The "hold your effort constant" half of the advice is repeated constantly and tested rarely. A 2023 paper by Antoine Bordas and Eric Fruchart offers a useful reframe instead of a flat answer: pacing isn't a physiological autopilot, it's a series of active, in-the-moment decisions, and the brain seems to use a different kind of judgment call for "hold this pace" than it does for "back off now." That's a genuinely interesting parallel to the gaze finding above — both pacing and terrain-reading turn out to be continuously re-evaluated, not something that just happens to you while you run.

And there's a real, if narrower, link between those two: a study led out of the same research group behind the gaze-timing finding measured gaze and stride together — cadence, stride length, pace — on people walking real streets with a mix of pavement, cobblestone, and dirt, and found the two "tightly coupled": your eyes and your feet are adjusting to terrain together, moment to moment, not as two separate systems that happen to share a body. That's a genuinely new thread for this curriculum, and it came from trying one different search term after two earlier passes had each concluded, separately, that gaze research and pacing research "don't talk to each other." They mostly don't — except where they demonstrably do. Worth flagging plainly: that coupling study is about walking on city streets, not running on unpredictable trail terrain over hours, so the extension to actual trail running is reasonable, not proven.

The barefoot/minimalist hunch — that going shoe-light lets your feet read the trail better — doesn't have the evidence behind it yet, and it's worth saying that plainly instead of assuming it. The most directly relevant study found, from 2024, put people in ultra-minimalist shoes on surfaces built to mimic rugged natural terrain and measured a real proprioceptive and postural response — your foot genuinely registers more information through a thin sole. But that study measures balance and postural activation, not whether that extra information changes where you look, how you place your next footstep, or how you pace yourself. No study found tests footwear against gaze or pacing decisions the way one 2021 study tested vision against foot placement — deliberately blurring one eye to disrupt depth perception and measuring how that changed people's real-time foot-placement safety margins on rough ground. That vision study exists. Its footwear equivalent doesn't, at least not in what two independently-worded searches turned up. That's not a small gap; it's the specific missing piece that would actually test the hunch, and it isn't there. Separately, minimalist shoes probably do change your running mechanics somewhat — shorter stride, less time on the ground, more forefoot landing — but even the classic 2013 meta-analysis on that question flags its own underlying studies as low quality. Two different "probably true, not proven" claims, stacked, are not the same as one demonstrated finding.

So: one skill, or three? Split answer, and the split is the honest finding, not a dodge. Gaze and stride are coupled moment to moment — real, measured, but so far only shown in walking, not trail running. Terrain reading and strategic effort pacing over the length of a long run are two things that look related and haven't actually been tested against each other. Footwear and either of the above is a plausible story with a real gap where its evidence should be. None of that makes any of it fake — it means the parts of "trail running feel" that are genuinely proven are more specific, and more interesting, than the whole bundled folk version. The part that would settle the footwear question — a study that manipulates plantar feedback the way researchers already manipulate vision, and measures what happens to a runner's eyes and pace on real rough terrain — doesn't exist yet. Somebody should run it.