A few weeks ago, one of the athletes in our Training Groups opened his dashboard, looked at his aerobic threshold, and decided the number was too low. He had a feel for where his easy running sits, built over a lot of miles and a lot of years, and the figure on the screen did not match it. So he questioned it.
That small moment, an athlete trusting his own experience enough to push back on a piece of software, is one of the most useful things that happens in this work. I want to walk through what came of it, because the whole sequence says something about how we build, and about how training itself should work.
What he was looking at
Aerobic threshold (AeT) is, first and foremost, a physiological state, it is the point where the balance between fuel sources shifts in a meaningful way. Below it, fat oxidation dominates and allows you to run for hours, hold a conversation, and recover quickly. Above it, the cost rises, fatigue accumulates faster, and the run stops building the deep aerobic base that mountain endurance rests on. Finding that point has traditionally meant a formal field test or a lab visit. Continuous AeT is our attempt to read it a different way, from the ordinary runs an athlete already does, week after week, without asking them to stop and test.
Continuous AeT proposed his AeT was 119 beats per minute. He thought it was closer to 131. That is a wide gap, and it is not cosmetic. At 119, nearly all of his running would fall above his threshold and the prescription would be to slow down and build base for months. The gap between his aerobic and anaerobic thresholds would suggest that his aerobic base had fallen significantly behind his top end fitness. At 131 he would look well balanced and physiologically ready to take on some higher intensity work. The same athlete, two very different training years, depending on one number.
The part I find encouraging
Here is what I keep coming back to. Will, our Product Manager building the dashboard, did not defend the reading. He sent it to me and asked me to look at the athlete’s data the way a coach would, by hand. An engineer asking the coaching side to check the engineering is exactly the loop we want, and it is rarer than it should be.
So I pulled his training file and worked through it the way a coach does.
How the reading is made
The dashboard and my personal analysis rest on the same idea: heart rate to pace decoupling. The principle is simple. Hold a steady, easy effort. If your heart rate climbed through the run while your pace stays the same, your body is working harder to hold the line, and you were above your aerobic threshold. If your heart rate stays consistent, you were at or below it. We look for the highest heart rate at which that drift stays under about five percent.
The trouble is that real training doesn’t happen under controlled conditions. This athlete runs in three conditions that all can distort the picture. He runs in heat, where heart rate drifts upward no matter how easy the effort. He runs on rolling terrain, where every hill changes the relationship between pace and heart rate unless you correct for grade. And many of his runs are short, where heart rate is still catching up to effort for the first part of the run. His summer runs showed nine to ten percent drift at a heart rate of 118, which looks like he was working far too hard, when really it was the temperature and humidity impacting his heart rate. W. His perceived effort in those sessions told the same story. The runs felt easy. That correlation between how hard a run feels and what the heart rate data shows is one of the most reliable cross-checks available, and it is exactly the kind of context that a reading built purely on heart rate and pace cannot see.
Once I adjusted his pace for the hills, set the hot and very short runs aside, and leaned on his cool, longer, steady efforts, the picture settled. One of the cleanest data points was a treadmill run, which removes terrain and the noise of GPS pace entirely. At a fixed pace for 49minutes, his heart rate barely moved at 125. An 80 minute run in cool weather at a heart rate of 126 held consistently. His pace only began to slip when his heart rate reached 130 and 131. I checked the result three independent ways and they agreed. My read is an aerobic threshold of about 128, in a band of 127 to 130.
So the athlete’s sense of his AeT was closer to right than the first reading, and the reading was low by roughly nine beats. His instinct was sound, even if his own guess of 131 sat a touch high.
Why a number can be wrong and the system still be right
This is by design, and it is worth being up front about it. Continuous AeT proposes a number, but the athlete has the final say after reviewing their dashboard. An athlete’s accumulated sense of their own effort is data too, and a good system is built to receive it rather than override it.
A first estimate from everyday runs is a starting point, not a verdict. What this case study demonstrates is concrete. A reading built on ordinary runs has to account for environmental factors, for terrain, and for how long a run is, or it will read low for exactly the athletes who train hard on hills in the summer. None of this is surprising to a coach, since coaches cross-check pace and heart rate data against terrain, environmental conditions, session length, and other variables that can impact the data. This case is a helpful reminder that cross-checks belong to the athlete too. If a number on your dashboard does not match your experience, that instinct is worth trusting. The dashboard is a starting point for a conversation, not the final word on your physiology.
Why this lives in Training Groups
Training Groups were built for athletes who train with structure and community, but without a coach reviewing their individual data. Continuous AeT exists to give those athletes access to a tool that helps them better understand their physiology. This tool provides a regular, informed read on whether their aerobic training is landing where it should. An athlete who pays attention and speaks up is not a complication in that process, they are a key part of it.
If you train with us in a Training Group, your dashboard is part of this work, and so are you. When a number looks wrong, trust that instinct. The experience you bring from your own training, accumulated over miles and years, is not something a dashboard can replicate. It is something you bring to it. That, in the end, is what the training experience is designed to be: not a verdict on your fitness, but a conversation between the data and the person doing the work.