Shindo Before anything else: nothing here measures your heart#
Worth stating up front, because it changes how to read the rest. This guide is not about capturing heartbeats — Shindo does not record workouts and does not read a sensor. The beats came from whatever watch, strap or ring you already wear, and they arrive through Apple Health. Everything below is arithmetic performed on data that has already been collected.
That matters more than it sounds. Setting zones is not a measurement problem, it is an inference problem: you have months of heart rate against pace sitting in a database, and the question is what boundaries that history supports. The answer does not require new equipment. It requires not throwing the history away in favour of a formula.
Where 220 minus your age comes from#
It is a linear fit published in the 1970s, drawn from a set of studies not designed to produce a prediction equation. It is a decent description of a population average and a poor description of any individual: the standard deviation around it is roughly 10 to 12 beats, meaning a substantial share of people sit 20 beats or more from their predicted maximum.
Twenty beats is not a rounding error. Applied to a five-zone chart it moves every boundary by around 10 to 15 beats, which is the entire width of a zone. The athlete whose true maximum is higher than predicted spends their easy days at a genuinely moderate intensity; the athlete whose maximum is lower spends their hard days unable to reach a target that does not exist and concludes they are unfit.
Later formulas — 208 minus 0.7 times age, and others — fit population data better and do nothing about the individual scatter, which is the actual problem. No equation whose only input is your age can tell you about your heart.
Percentage of what?#
Three anchors are in common use, and they are not interchangeable.
Percentage of maximum. The default on most watches. It requires knowing your true maximum, which requires a maximal effort you have probably never performed, and which declines slowly with age in a way the formula only crudely tracks.
Percentage of heart-rate reserve. Maximum minus resting, scaled, plus resting. Better than raw percentage of maximum, because it accounts for the fact that a fit person’s resting rate is much lower. It still needs the maximum.
Percentage of threshold heart rate. Anchored to your lactate threshold rather than your ceiling. This is the one worth using: the threshold is what training intensities are actually organised around, you can find it in the field, and it responds to training — which means re-testing it tells you something, whereas re-testing your maximum does not.
Finding your threshold in the field#
The standard protocol is a 30-minute solo time trial at the hardest pace you can hold evenly, on flat ground, in cool conditions, after two easy days. Discard the first ten minutes while heart rate climbs, and take the average heart rate over the final twenty. That figure is a workable estimate of threshold heart rate.
Two practical notes. Run it alone — a group turns it into a race and ruins the pacing. And do not do it in the heat, because the drift will inflate the reading and every zone you derive from it. A 20-minute version works if 30 is unrealistic; take the average of the last 15 and expect the result to sit one or two beats high.
A zone set anchored to threshold#
Once you have a threshold heart rate, the zones follow. These bands are the common convention and there is nothing sacred about the exact percentages:
| Zone | % of threshold HR | What it is for |
|---|---|---|
| 1 — recovery | Below 81% | Genuine recovery. Should feel almost pointless |
| 2 — aerobic | 81–89% | The bulk of endurance volume |
| 3 — tempo | 90–93% | Moderate. Useful in small doses, easy to live in by accident |
| 4 — threshold | 94–99% | Around the boundary you measured |
| 5 — above threshold | 100%+ | Intervals. Duration limited, heart rate largely irrelevant |
Re-test every eight to twelve weeks during a block, and after any long layoff. Threshold heart rate moves less than threshold pace does — a well-trained athlete usually gets faster at the same heart rate rather than reaching a higher one, which is itself the clearest evidence the training is working.
There is a related question of where the aerobic boundary sits, which is a different threshold from this one and needs a different method: zone 2 without a lab test covers it.
What zones cannot tell you#
Cardiac lag. Heart rate takes 30 to 120 seconds to respond to a change in effort. In intervals shorter than about three minutes it is describing the previous repetition, which makes it useless as a control signal there. Use pace or power for short work and heart rate for long steady work.
Drift. Over a long effort at constant output, heart rate climbs — from heat, dehydration and accumulated fatigue. Holding a heart-rate target on a three-hour ride therefore means gradually slowing down, which may or may not be what you wanted.
Everything non-training. Caffeine, poor sleep, a stressful morning, illness and altitude all shift heart rate at a given effort by several beats. A day when the numbers look strange is usually a day about you, not about your zones.
When there is not enough history#
An honest answer here is short: if you have never performed a maximal or threshold effort and have a few weeks of easy running, nothing can locate your threshold. The available options are to invent a number from your age, or to leave it blank.
Leaving it blank is the better option, and it is worth insisting on. A fabricated boundary looks like information, and every zone, every intensity target and every compliance judgement built on top of it inherits the error silently. A blank at least tells you what you do not yet know.
In the meantime, breathing works from the first session: you can hold full sentences below the aerobic boundary, and you cannot near threshold. It is less precise than a number and considerably more accurate than the wrong one.
Where heart rate is simply the wrong tool#
For short intervals, for anything with pace changes under a couple of minutes, for hill repeats, for the opening kilometre of anything, and for cross-comparison between people. It is at its best exactly where it is least exciting: long steady efforts, and tracking what happens to your pace at a fixed heart rate across a season. That second one is one of the cheapest and most informative long-run measurements available, and it needs no test at all.
Shindo derives thresholds and zones from your own training history rather than from 220 minus your age, modelling critical power and W′ for cycling and critical speed and D′ for running and swimming from the best efforts in your data. It records nothing itself — every heartbeat arrives through Apple Health from whatever you already wear. When there is not enough history to model a threshold, the value is left blank rather than filled with a population estimate. Threshold and zone modelling is part of the paid tier; readiness and its drivers, training load and your weekly plan over the most recent 60 days of history are not. The methodology sets out what is modelled and from what.
Read next