Zones, and why mine were wrong

May 20, 2026

For two years I trained off heart-rate zones a Garmin watch calculated from 220 minus my age. The watch put my Zone 2 ceiling at 152 bpm. I rode and ran below that ceiling, called it aerobic, and waited to get faster. I got slightly faster. Not at the rate the volume should have produced.

The problem was that 220 minus age is a population-average estimator with a standard deviation of roughly ±10 bpm. My actual max HR, measured on a hill repeat session with a chest strap to exhaustion, is 198. The watch had me at 200, close enough. But the formula it used to derive zones from that max underestimated my lactate threshold by 14 bpm.

The right way to set zones is a field test. I used the simplest one: a 30-minute time trial on a flat course, all-out, with HR averaged over the last 20 minutes. That number is a reasonable proxy for lactate threshold HR (LTHR). Mine came back at 178. Aerobic threshold sits roughly 30 bpm below that. So my true Z2 ceiling is 148, not 152, and the top of Z2 isn’t a soft cap — it’s where lactate starts to rise out of steady state. Four bpm matters more than it sounds, because the curve isn’t linear near threshold.

I retested on the bike separately. Cycling LTHR was 168, ten bpm below running, which is the expected gap (smaller muscle mass recruited at the same cardiac output). Running LTHR of 178, bike LTHR of 168, swim I don’t bother with HR for because the dive reflex makes it noisy.

Now I train in five zones derived from those two numbers:

The 80/20 polarized prescription says 80% of weekly training time below Z2 ceiling, 20% above Z4 floor, almost nothing in Z3. Z3 is the trap. It feels productive, fatigues you like threshold work, and produces the adaptations of neither. I now spend more time genuinely easy than I would have predicted I could tolerate, and the hard days are harder.

Retest every eight weeks. Fitness shifts the curve.