Your first two minutes don't count. Your last two do.

You set off. You are running, or climbing, or pulling on the rower, and you are unmistakably working. You glance at your phone and training time is still at zero. A minute later, still zero. Somewhere in the second or third minute the app finally decides you have started, and the clock begins.

It feels like being short-changed, and the instinct behind that feeling is correct: the work really did start before the score noticed. Here is what is actually happening in those minutes, and why the end of the session quietly hands them back.

The work starts before your heart rate does

When you go from sitting to working, your muscles need ATP immediately, but the aerobic machinery that supplies it does not arrive immediately. It ramps up along an exponential curve, and the gap between what you need and what oxygen is delivering in those first seconds is called the oxygen deficit. It is not a metaphor — it is paid for out of phosphocreatine stored in the muscle.

The timescale is well characterised. In their review "A century of exercise physiology: key concepts on coupling respiratory oxygen flow to muscle energy demand during exercise", Ferretti, Fagoni, Taboni, Vinetti and di Prampero report that the time constant for muscle oxygen uptake at exercise onset sits between 20 and 25 seconds, matching the rate at which phosphocreatine falls — Binzoni and colleagues measured that fall directly by nuclear magnetic resonance and fitted a time constant of 23.4 seconds. The same review notes that above roughly 50–60% of maximal aerobic power the time constant gets longer, so the harder you start, the longer the transition takes.1

A time constant of about 23 seconds means you are around 63% of the way to steady state after 23 seconds, 86% after 46, and 95% after roughly 70 — for a moderate effort, and longer for a hard one. That is where "a minute or two" comes from.

You can see the same thing in the energy accounting. In "Bioenergetics of the VO₂ slow component between exercise intensity domains", Colosio, Caen, Bourgois, Boone and Pogliaghi had eight active men ride constant loads for 3, 6 and 9 minutes at three intensities and costed each three-minute segment in oxygen-equivalent terms. At moderate intensity the first three minutes cost 2126 ml against 2687 for the next three at the identical workload — about a fifth less. The same pattern held in the heavy domain (4278 against 5121), and in the severe domain the cost kept climbing across all nine minutes.2

So your suspicion is right. Those early minutes are real training. They are simply being funded by a system that heart rate cannot see, and the measurement catches up afterwards.

Why does this matter beyond bookkeeping? Because the size of that deficit is a fitness marker in its own right. In their review "Oxygen uptake kinetics", Poole and Jones summarise it plainly: "fast VO₂ kinetics mandates a smaller O2 deficit, less substrate-level phosphorylation and high exercise tolerance," while slow kinetics "incurs a high O2 deficit, presents a greater challenge to homeostasis and presages poor exercise tolerance."3 As you get fitter, your lag shortens — the frustrating dead time at the start of a session is something training actually shrinks.

Why ZonePoints waits anyway

Given all that, why not start the clock when you obviously started working?

Because ZonePoints only knows one thing: your heart rate. There is no start button and no timer to configure — a session begins when you cross into zone 1 and pauses when you drop below it. Everything the score claims has to be traceable to a measurement, and crediting work the sensor did not see means guessing at the size of the guess.

The alternative is worse than it sounds. A threshold that fires on intent rather than on measurement is a threshold that fires when you climb the stairs, and the whole point of putting the floor at 50% of your heart rate reserve is that ordinary movement does not reach it. That argument is laid out in full in nobody agrees what a heart rate zone is; the short version is that a session timer which starts on a brisk walk to the kitchen is measuring your day, not your training.

So the app under-counts at the start. It does so knowingly, and, as it turns out, temporarily.

Then the zone starts holding you up

The second thing you have probably noticed is that the back half of a session is easier to keep in zone than the front half. That is not your imagination either, and it has a name: cardiovascular drift.

In "Cardiovascular drift during prolonged exercise: new perspectives", Coyle and González-Alonso describe it as a progressive decline in stroke volume after 10 to 20 minutes of exercise, and argue that it is driven primarily by the rising heart rate rather than by blood being diverted to the skin as body temperature climbs.4

The size of it has been measured. In "Cardiovascular drift is related to reduced maximal oxygen uptake during heat stress", Wingo, Lafrenz, Ganio, Edwards and Cureton had nine male cyclists ride at a constant 60% of VO₂max in 35°C and measured them at 15 and 45 minutes. Heart rate rose 12%, from 151 to 169 beats per minute, while stroke volume fell 16% — at a workload that never changed. Oxygen uptake, they note, "increased only slightly over time."5

Eighteen beats over half an hour is the difference between scraping along the bottom of a zone and sitting comfortably inside it. Hold the same pace and the score climbs on its own.

But this is not a loophole, and it is worth understanding why. The same study measured VO₂max immediately after the ride and found it had fallen 19%, so the identical workload that represented 63% of maximum at 15 minutes represented 78% of maximum at 45. The authors' conclusion is that the results "support the validity of using changes in HR to reflect changes in relative metabolic intensity during prolonged exercise."5

In other words, your heart rate is not drifting away from the truth. It is tracking it. The work has genuinely become harder relative to what you are currently capable of, and a score that reads relative effort is right to say so. The minutes you bank late in a long session are earned.

Two honest caveats. That study was run at 35°C, and heat makes drift considerably larger — in a cool room on a short session you should expect much less of it. And drift is a reason the last twenty minutes of a long ride feel harder than the first twenty at the same pace, which is not the same as those minutes being free.

And the end gives the start back

Now the part that closes the loop. When you stop, your heart rate does not fall off a cliff — it decays, over minutes.

The clinical literature fixes how slow "normal" is. In "Heart-rate recovery immediately after exercise as a predictor of mortality", Cole, Blackstone, Pashkow, Snader and Lauer followed 2428 adults referred for diagnostic exercise testing and defined heart rate recovery as the fall from peak to one minute after stopping. Their threshold for an abnormal result was a drop of 12 beats per minute or less, and 639 of them — 26% — came in under it. That group was twice as likely to die over the following six years, after adjustment for workload, perfusion defects and standard cardiac risk factors.6

That is a clinical sample rather than a group of healthy athletes, but the threshold is the useful part: a first-minute fall of a dozen beats is slow enough to be a warning sign, so a normal one is some tens of beats — nowhere near enough to drop you out of a zone the instant you stop. If you finish a hard effort 30 or 40 beats above your zone 1 threshold and then ease down instead of stopping dead, you spend the next few minutes still above it — and ZonePoints is still counting, because the only thing it ever asked was whether you are above the line.

That tail is scored honestly, too. Each span of time is credited at the heart rate measured at the end of it, so a falling rate during a cool-down is scored at the lower figure, not the one you came in at. You do not bank hard-effort points while coasting.

One thing this is not: repayment of a debt. The tidy story — deficit at the start, repaid at the end — is exactly the model exercise physiology abandoned. In "Metabolic bases of excess post-exercise oxygen consumption: a review", Gaesser and Brooks worked through the evidence against Hill's 1920s "oxygen debt" hypothesis, showed that lactate removal and the slow phase of post-exercise oxygen consumption follow different kinetics, and recommended dropping "debt" for the neutral term EPOC precisely "to avoid implication of causality".7

So the symmetry here is bookkeeping, not biology. The score under-counts a minute or two at the beginning and over-counts a couple at the end, and the two happen to be of similar size. That is a fair ledger, not a physiological law — and it is why the missing minutes at the start are not worth arguing about.

What to actually do about it

Two practical things follow, and both are the opposite of what the frustration tempts you into.

Warm up before the part you care about. If the transition costs you a minute or two, spend it deliberately rather than resenting it. In "Warm up II: performance changes following active warm up and how to structure the warm up", Bishop's conclusion is that an active warm-up improves performance lasting five minutes or more when it leaves you starting "in a relatively non-fatigued state, but with an elevated baseline oxygen consumption" — while being candid that well-controlled warm-up studies are scarcer than the consensus around them suggests.8 The mechanism is better established than the performance effect: in "How Priming Exercise Affects Oxygen Uptake Kinetics", Goulding, Burnley and Wüst review why a prior bout speeds up the oxygen uptake response, concluding that enhanced intracellular oxygen utilisation is the likely central cause rather than raised muscle temperature or lactic acidosis.9 Start the session that counts already warm and you cross the threshold sooner.

Don't stop dead. Your heart rate is going to take minutes to come down whatever you do with them, so spend them moving. Easing down rather than standing still holds you above the threshold for longer, and every one of those minutes scores.

The takeaway

The clock starting late is a real cost, and it is a real measurement problem rather than a design preference: heart rate is an honest signal and a slow one, and a score built on it has to wait for it.

What makes that acceptable is that the error does not accumulate in one direction. You lose a minute or so at the front, drift hands you easier zone time through the middle of a long session, and the decay at the end returns a comparable minute or two at the back. Train for long enough and the ledger balances itself — and the lag at the start gets shorter as you get fitter, which is the one part of this you can actually do something about.

1. Ferretti G, Fagoni N, Taboni A, Vinetti G, di Prampero PE. "A century of exercise physiology: key concepts on coupling respiratory oxygen flow to muscle energy demand during exercise." European Journal of Applied Physiology. 2022;122(6):1317–1365. The 23.4 s phosphocreatine time constant is from Binzoni et al. (1992), reported and endorsed in that review.
2. Colosio AL, Caen K, Bourgois JG, Boone J, Pogliaghi S. "Bioenergetics of the VO₂ slow component between exercise intensity domains." Pflügers Archiv — European Journal of Physiology. 2020;472(10):1447–1456.
3. Poole DC, Jones AM. "Oxygen uptake kinetics." Comprehensive Physiology. 2012;2(2):933–996.
4. Coyle EF, González-Alonso J. "Cardiovascular drift during prolonged exercise: new perspectives." Exercise and Sport Sciences Reviews. 2001;29(2):88–92.
5. Wingo JE, Lafrenz AJ, Ganio MS, Edwards GL, Cureton KJ. "Cardiovascular drift is related to reduced maximal oxygen uptake during heat stress." Medicine & Science in Sports & Exercise. 2005;37(2):248–255.
6. Cole CR, Blackstone EH, Pashkow FJ, Snader CE, Lauer MS. "Heart-rate recovery immediately after exercise as a predictor of mortality." New England Journal of Medicine. 1999;341(18):1351–1357.
7. Gaesser GA, Brooks GA. "Metabolic bases of excess post-exercise oxygen consumption: a review." Medicine & Science in Sports & Exercise. 1984;16(1):29–43.
8. Bishop D. "Warm up II: performance changes following active warm up and how to structure the warm up." Sports Medicine. 2003;33(7):483–498.
9. Goulding RP, Burnley M, Wüst RCI. "How Priming Exercise Affects Oxygen Uptake Kinetics: From Underpinning Mechanisms to Endurance Performance." Sports Medicine. 2023;53(5):959–976.