Find the hours when you are actually sharp

Tap a number whenever you notice your state change, three or four times on a normal day. Do not try to log on a schedule, and do not go back and reconstruct yesterday from memory. A reading taken in the moment is the only kind worth having here.

What the numbers mean

Use the scale loosely: 1 is running on empty, 3 is the ordinary working state you could hold all afternoon, 5 is the state where hard problems feel easy. Consistency with yourself matters more than matching anyone else's definition, so pick a rough anchor on day one and keep using it.

Reading the analysis

The hourly chart only draws hours you have actually logged, and it shows the reading count next to each average so you can see which bars rest on one data point. Treat an hour with two readings as a hint, not a finding. Two weeks of steady logging usually makes the shape obvious.

Then change the schedule

The log is only useful if something moves. Once your best window is clear, put one hard task inside it tomorrow and push a recurring meeting out of it. The crossover card tells you how much of your current focus time already lands in that window.

Below ten readings the page shows the tap panel, the count boxes, today's list and a progress bar, with the Best window box left on dashes. At ten the hourly chart, the weekday chart, the fourteen-day strip and the prime-time card appear together. Ten is where a chart stops being actively misleading, not where it becomes reliable.

The window itself is picked in three steps. Every reading is averaged into the hour it was taken in. Hours logged at least twice are preferred, and if fewer than three clear that bar the page falls back to every hour with any data at all. Then it takes the three highest averages, settling ties on the reading count and after that on the earlier hour, and prints them in clock order.

A worked example

Suppose sixteen readings spread over six hours: four at 09:00 averaging 4.3, three at 10:00 also averaging 4.3, three at 11:00 averaging 3.7, three at 14:00 averaging 2.3, two at 16:00 averaging 3.0, and one lone 5 tapped at 21:00. Five hours have been logged more than once, which is enough, so the 21:00 reading is left out despite being the highest number in the log. The three best averages are 10:00, 09:00 and 11:00, the card reads 9:00 to 12:00, and the average box reads 3.7. Those two 4.3s are the reminder that the ranking uses unrounded figures while the chart prints one decimal place.

Reading it without over-reading it

The end of that label is exclusive: 9:00 to 12:00 is the three hours 9, 10 and 11, not four. When the three best hours are not next to each other the card lists them separately instead of inventing a range across the gap. The crossover figure underneath counts focus sessions by the hour they ended in rather than the hour they started, so a fifty-minute session begun at 08:40 is counted against 09:00, and seven of twenty sessions inside the window prints as 35 percent.

Chronotype is real. People differ in when their body clock puts alertness and body temperature peaks, and the synchrony effect, where performance is better at your preferred time of day, shows up repeatedly in the literature. It is also conditional: the effect is largest for strong morning and strong evening types, smaller for the majority who sit in the middle, and it varies by task. Memory and attention tasks show it more clearly than routine work.

That is exactly why this page logs instead of asking you questions. A questionnaire sorts you into a category built from group averages. Two weeks of your own readings tell you what your own days do, including the parts a chronotype label cannot know: your commute, your meetings, your sleep debt, your lunch.

Expect the post-lunch dip to show up around the early afternoon, since it is a normal feature of the daily rhythm rather than a personal failing. If your readings show one, schedule around it rather than fighting it.