The number climbing on the screen looks like a measurement. It is a guess. The machine is running a formula built from lab averages, and it is feeding that formula whatever inputs it has, which is usually very little. On most machines that means time, speed, and resistance setting. If you skipped the setup screen, it does not even know your weight, so it assumes a default that is often a hundred and fifty or a hundred and seventy pounds. Everything after that is arithmetic on a stranger's body. Two people can run the same thirty minutes on the same machine and burn very different amounts, and the display will hand them the same number.
Consider what the machine cannot know. It does not know your body composition, and muscle and fat burn at different rates at rest and in motion. It does not know your fitness level, and a trained person moves the same distance using less energy than an untrained one. It does not know your stride, your posture, or how much of your weight you are parking on the handrails. It does not know your age or sex unless you told it, and both change the baseline. It is estimating the energy cost of a task, then assuming you are doing that task the way the average test subject did. Lean on the rails and the machine still credits you for lifting your full weight.
The error is not random. It leans high, which is the direction that hurts. Ellipticals are the worst offenders, and studies comparing machine displays to metabolic carts have found them overstating by a wide margin, sometimes by a third or more. Treadmills tend to be closer, because walking or running at a known speed and grade is a physics problem with a decent answer. Stationary bikes with real power meters are the most honest of the group, since watts can be measured rather than inferred. Rowing machines land somewhere in the middle, depending on how well the drag factor is set. As a rough rule, the more the machine measures and the less it guesses, the closer it lands.
There is a second problem underneath the first. The machine reports gross calories, meaning everything your body used during that window. But your body was going to spend energy anyway, just sitting in a chair, breathing and keeping your heart going. That baseline is roughly one calorie per minute for many adults, so a forty minute session includes about forty calories you would have burned doing nothing. The number you actually added by exercising is the net, and it is smaller than what the display shows. Nobody advertises the net, because the gross number feels better. On a long session the gap between the two can run past a hundred calories on its own.
Heart rate does not rescue the estimate either. Wrist based sensors read a light signal through skin, and they drift during grip changes, cold hands, or fast movement. Chest straps are far more accurate for heart rate itself, but heart rate is still a weak proxy for calories. Caffeine, heat, poor sleep, stress, and dehydration all push heart rate up without pushing energy use up in step. A hard day and a hot room can produce the same reading for very different work. So a high heart rate tells you the session felt demanding, which is worth knowing, and not much about the fuel. Watches that report a daily burn are stacking the same weak guess on top of a step count.
This matters most for people who eat back what they burned. A three hundred calorie overstatement is easy on an elliptical, and a snack chosen to match it is easy too. Do that five times a week and the deficit you thought you built is gone, which is exactly the pattern behind the frustrated question of why training hard is not changing the scale. The exercise still did its job for your heart, your joints, your mood, and your sleep. The accounting is the part that failed, and it failed on both ends, since food labels carry their own margin of error.
Use the machine for what it is good at instead. Track time, distance, watts, pace, and resistance, because those are measured rather than modeled, and they tell you whether you are getting fitter. Compare this week's session to last month's at the same effort and watch the numbers move. Judge your weight trend by a weekly average rather than any single morning, since daily readings swing on salt and water. Handle the eating side with food, not with a display, and set intake based on results over three or four weeks. The calorie counter is not useless as a relative measure of your own effort over time. As an absolute number to plan meals around, it was never built for that.
