What you can actually ask

These are real answers. They come from one real garmlink account with about three months of history: 67 sessions across swimming, cycling and running, 94 days of daily metrics, 51 nights of sleep and 58 sessions with lap-by-lap detail.

Figures are rounded and dates removed. Nothing here identifies a place, a route or a device.

Training load

Compare my last three months of training. What changed?

The volume barely moved — roughly 20 hours a month throughout — but the shape of it changed completely. The first month was about 290 km, mostly on the bike, at an average heart rate around 145. The second month had more sessions but far fewer kilometres, and the average heart rate dropped to about 130. The third dropped again, to about 118. So: same time invested, progressively easier work.

How does my time split between swimming, cycling and running? Has the balance shifted?

Completely. In the first month cycling took nearly 12 hours across 9 rides, with 4 swims for under 2 hours. A month later that inverted: 20 swims for close to 13 hours, and a single ride under half an hour. Running stayed roughly steady at 4 to 7 hours a month, which makes it the only discipline you kept constant.

Are my easy sessions actually easy?

Not really. Across the 58 sessions with zone data, you logged about 8 hours in zone 1 and 7 in zone 2, against 10 in zone 3, 8 in zone 4 and nearly 2 in zone 5. That is 56 % of your time at zone 3 or above. The classic polarised model would have most of it in zones 1 and 2 — you are spending a lot of time in the middle.

Recovery

Did I actually recover from my heaviest week, or did I just stop training?

You did, and the data shows it clearly. Your heaviest week was about 11.5 hours. The week that followed dropped to under an hour, and your readiness score went from the mid-30s to around 60, resting heart rate came down 2 beats, HRV held steady and sleep stayed at about 7 hours. That is recovery.

Compare it with the other heavy block earlier on. Two demanding weeks back to back, sleep down at 5.8 hours, and HRV fell to its lowest of the whole period while readiness sat in the low 30s. It took a near-rest week before HRV came back up. Same load, very different outcome — the difference was the sleep.

How was I sleeping before my best sessions?

Barely differently, and that is the finding. Nights before a session of two hours or more averaged 6.3 hours of sleep with 75 minutes of deep sleep. Every other night averaged 6.6 hours with 77 minutes. You are not sleeping any more ahead of a hard day than you do the rest of the time.

How many nights under 7 hours, and do they follow my hardest days?

31 of 51 recorded nights were under 7 hours — a clear majority. But they do not follow your hard days: average training load on the day before a short night was essentially the same as before a long one. Your short nights are not caused by training.

Does my readiness score actually predict my good sessions?

It looks more like a rear-view mirror than a forecast. On days when readiness read below 40 you trained about 1.3 hours; on days it read above 60, about 0.2 hours. In other words the score is low on the days you train, and high on the days you rest — it reflects the load rather than anticipating it.

One session in detail

Tell me about my longest ride.

About 4 hours 20 for 56 km, average heart rate 148, in 30 °C with 52 % humidity — it felt like 31 °C.

The lap breakdown tells the real story. The first seven 5 km laps ran between 15 and 27 km/h with heart rate from 107 up to 150. Then the last four collapsed to 7 to 9 km/h while heart rate climbed to 163. Speed halved and heart rate rose: that is a long sustained climb in the back half, not fatigue.

On that ride, did my heart rate drift at constant effort?

On the flat portion, hardly at all. Comparable laps early on ran at 26.7 km/h for 139 bpm and 26.0 km/h for 141 bpm — two beats over an hour of riding, which is negligible. The rise afterwards is explained by the gradient, not by cardiac drift.

Progress and conditions

Am I faster for the same effort than I was two months ago?

Yes, measurably. Your runs averaged about 6 min 45 per kilometre at 146 bpm early on. A month later they averaged about 6 min 00 per kilometre at 147 bpm. Same heart rate, 45 seconds per kilometre quicker.

Does the heat change my heart rate at equal effort?

Your data cannot answer that cleanly, and it is worth saying so. Rides above 25 °C averaged 149 bpm at 24.7 km/h; the cooler ones averaged 132 bpm at 17.4 km/h. But the hot rides were also the faster ones, so speed and temperature move together and the effect of heat alone cannot be separated. It would take rides at comparable speed in both conditions.

Given my last seven days, what should I do today?

Your last seven days total about 3.6 hours — well under half your usual week. Sleep has averaged 7 hours, which is above your own baseline, and readiness has averaged around 55, in the middle of your range. You are rested and undertrained rather than tired. A longer or harder session today is well within what your numbers support.

Every one of these is answered by garmlink reading your own Garmin history through your assistant. Nothing is typed in by hand. Connect your AI.