Zone 2 has become the most discussed intensity in amateur endurance training, and most people are training in a zone derived from a guess about a number they never measured. The idea is sound. The arithmetic underneath it usually is not.
What zone 2 is actually trying to capture
Underneath the zone models is a physiological boundary: the highest intensity you can sustain while remaining comfortably aerobic, with lactate production and clearance in balance. Training there for extended periods drives mitochondrial and capillary adaptations that harder work does not, and it is sustainable enough to accumulate real volume.
The percentage-of-max model is an approximation of that boundary, not a definition of it. Which matters, because the approximation is built on an estimate.
The error stacks up twice
First error: your maximum heart rate. The 220 minus age formula is a population average with a standard deviation of roughly 10 to 12 beats per minute. That means a substantial fraction of people are 10 or more beats away from the estimate in one direction or the other. It was never intended for prescribing individual training.
Second error: the percentage boundaries. The relationship between percentage of maximum heart rate and the actual physiological threshold varies between individuals too. Two people with identical measured maximums can have first ventilatory thresholds several percent apart.
Stack those and a calculated zone 2 ceiling can plausibly be 10 beats wrong. That is the difference between genuinely easy aerobic work and grinding away in the moderate zone that produces fatigue without the benefits of either end.
The talk test, which is better than it sounds
Zone 2 is the highest effort at which you can speak in complete sentences without needing to breathe mid-sentence. Not single words. Not gasping between phrases. Full, unhurried sentences.
This is not a beginner’s substitute for real data. The talk test corresponds reasonably well to the first ventilatory threshold, which is the thing the zone model is estimating in the first place. It is also free, self-calibrating, and automatically accounts for heat, fatigue, altitude and illness in a way that a fixed heart rate ceiling does not.
Practical approach: find the talk-test effort, note the heart rate it corresponds to, and use that number as your ceiling from then on. Re-check it every few months, because it moves as you get fitter.
Why two devices disagree about the same run
Apple, Garmin, Polar and WHOOP all use different zone definitions. Some are percentages of maximum heart rate, some of heart rate reserve, some of lactate threshold heart rate. Some let you set the boundaries and some do not.
So the same 45 minute run genuinely can be zone 2 on your watch and zone 3 in another app, with neither being wrong. They are answering different questions. This is the same class of problem as two recovery apps disagreeing about the same night, which we took apart in how recovery scores actually work.
The mistake that matters more than the zone boundary
Getting zone 2 exactly right matters far less than not running your easy days at moderate intensity. The characteristic amateur pattern is a cluster of sessions in the middle: too hard to be recovery, too easy to drive high-intensity adaptation.
Our analysis of real training logs found that elites are not doing less hard work in absolute terms, they are doing much easier easy work. Elites train easier than you do goes through the distributions, including the ways the popular version of that claim is wrong.
To calculate your zones once you have a real maximum, use the heart rate zone calculator, and to raise the ceiling itself see how to increase your VO₂max.