How to know your fatigue baseline is ready
Do not ask Hydrogen for deep explanations too early. First check whether your baseline is stable enough to compare real periods instead of noisy days.
- Two or three normal workdays are the minimum useful baseline.
- Coverage quality matters more than asking too early.
- Comparison prompts outperform vague summary prompts once the baseline is ready.
The mistake almost everyone makes
Most people ask Hydrogen for interpretation before the Windows app has enough normal workdays to support one. They want the answer first and the data second.
That produces weak answers, because the system is being asked to compare unstable coverage rather than stable behavior. The agent does what you ask. It just cannot make up data it does not have.
A better readiness check
Start by asking whether your recent coverage is strong enough for the comparison you want. That is a question Hydrogen can answer well on day one — it is a question about the data itself, not about what the data means.
- Two or three normal workdays before your first serious comparison.
- Seven to fourteen days for stronger pattern calls.
- Treat low-confidence or interrupted days as context, not conclusions.
First questions that actually work
Once the baseline is ready, comparisons consistently outperform vague summary prompts. The questions Hydrogen handles best look like specific, time-bounded comparisons — and look much less like "how am I doing."
- Compare my last 14 days against the previous 14.
- Which low-energy window repeats most on workdays?
- What changed in my active minutes this week versus last week?
