Why Your Watch Thinks Leg Day Was Recovery

You finish a session that genuinely emptied you. Heavy squats, a few hard sets of Romanian deadlifts, walking lunges until the stairs became a problem. You check your watch and it has decided this was a light day. Low strain. Modest calorie burn. Some devices will go further and suggest that, given how easy today was, tomorrow looks like a good day to train hard.

This is one of the most common complaints from anyone who lifts and does cardio, and the usual explanation you find online is only half of it. There are actually two different problems here, stacked. Only one of them can be solved by buying better hardware.

Problem one: the sensor is struggling

Optical wrist sensors work by shining light into the skin and reading the reflection as blood moves underneath. That works well when your wrist is reasonably still, held at a consistent tension, and moving in a rhythmic way. Running is close to ideal for it.

Lifting is close to the worst case. Gripping a bar hard contracts the forearm muscles right where the sensor is sitting. The wrist flexes and extends through the movement. The watch shifts on the strap. On some exercises the wrist is bearing load directly.

The result is a heart rate trace with gaps, spikes and plateaus that do not correspond to anything your heart actually did. It sometimes latches onto the cadence of the movement instead of your pulse. So before we even ask whether heart rate is the right signal, the heart rate you have is frequently not accurate.

A chest strap fixes this part. It measures electrical activity rather than light through skin, and it does not care what your forearms are doing. If you want trustworthy heart rate during lifting, that is the answer.

But fixing the measurement exposes the bigger problem underneath.

Problem two: heart rate is the wrong question

Suppose your heart rate during that squat session was measured perfectly. It would still make the session look easy, because that is a fair description of what your cardiovascular system did.

A heavy set of five takes perhaps twenty seconds. Your heart rate climbs during it, peaks somewhere in the couple of minutes after, and comes back down during the rest interval. Across an hour, the area under that curve is modest — genuinely, honestly modest. Compared to an hour of steady running, your heart did less work.

What that number cannot express is everything else the session cost. High mechanical tension through the muscle. Neuromuscular fatigue that will still be present in two days. Connective tissue and joint loading. The reason you struggle on stairs on Wednesday is not cardiovascular, and no summary of your heart's behaviour is going to contain it.

So your watch is not malfunctioning. It is accurately answering a cardiovascular question. Lifting is not primarily a cardiovascular event, which means the answer is close to irrelevant no matter how precisely it is measured.

This is the same blind spot that makes sled pushes and heavy carries register as moderate in a Hyrox block, and the reason a heart-rate-based load figure is one method among three rather than the obvious default.

Why this matters beyond the annoyance

If it were only a wrong number on a screen, it would not be worth an article. The problem is what gets built on top of it.

Your watch does not just record heart rate. It computes recovery scores, readiness, strain, training status and suggested workouts, and all of it flows from the same underlying picture. If two lifting sessions a week are effectively invisible to that picture, then every derived number is confidently describing a smaller version of your training than the one you are actually doing.

The practical shape of this: your load looks flat and comfortable through a block where you are adding gym work. Your watch tells you that you are fresh and could handle more. Meanwhile the accumulation is real and lands on the same legs you run on. The chart is not warning you because, as far as it is concerned, nothing much happened.

That is why the fix is not a better heart rate number. It is a load figure that contains both halves of your week in the first place.

What actually works

• Do not try to solve it with heart rate. Even perfect heart rate data will not represent a strength session properly. This is the point most articles miss when they stop at "get a chest strap".
• Score lifting on its own terms. Tonnage — sets times reps times weight — is objective, comes straight out of any gym logging app, and moves when your training moves. Session RPE works too and takes ten seconds.
• Combine at the load level, not the heart rate level. Compute cardio load from heart rate, compute strength load from tonnage or RPE, then add those two figures. The method is in this post, and it can live in a spreadsheet.
• Keep using your watch for what it is genuinely excellent at. Cardio. It is very good at cardio. The mistake is treating a cardiovascular instrument as a complete account of training.
• Distrust readiness scores during heavy lifting blocks specifically. Not because they are badly built, but because they are working from an input that is missing a large part of what you did.

The short version

Your watch is not broken and it is not underestimating you out of stubbornness. It is measuring the thing it was designed to measure, and that thing does not capture what a barbell does to you.

The gap only closes when strength work is counted as load in its own right rather than being inferred, badly, from your pulse.

Disclosure

That gap is why I built LOAD. It takes cardio load from heart rate via Strava or Polar, strength load from tonnage via Hevy or manual entry, and adds them into a single weekly number, so a heavy gym week shows up as a heavy week. It is free for individual athletes, and connecting Strava backfills 12 weeks so there is real history from day one.

The spreadsheet version works too, and costs nothing. The important part is that lifting stops being invisible.

The methodology and citations are public at loadtraining.app/faq.