How do you compare the same week in different years?

Some weeks come back on their own every year. The week the school term starts. The week of the holiday your family always does the same way. The week before the conference, or the week after it, when everything is running a little behind. It is natural to wonder whether this year's version was better, worse, or about the same as last year's — and the answer that arrives first is almost never a comparison. It is a feeling about how that week usually goes.

"The same week" is not the same date

The first problem is quieter than it looks. School terms do not start on the same date every year. A holiday that always falls on a weekend lands on different weekdays depending on the year. The week before a recurring deadline shifts by however many days the deadline itself shifts. So "the same week" almost never means the same calendar week number — it means the week built around a recurring anchor, wherever that anchor happens to land this time. Compare by date alone and you can end up comparing the wrong days entirely: a calm Tuesday against what was actually moving day, just because both happened to carry the same date.

Why the comparison collapses into a mood

Even once you have the right week lined up, the comparison rarely stays a comparison for long. This year's version is still being lived — tired or excited or annoyed, right now, with nothing softened yet. Last year's version has had twelve months to compress: the ordinary parts of it are gone, and what is left is a single texture, usually just "that week was rough" or "that week was nice." Put those two things side by side and you are not weighing one week against another. You are weighing something you are currently inside against a rumor of a week, and the rumor was assembled by memory with none of the detail intact.

An asymmetry that favours nobody's actual year

This has a specific, unfair shape. The week you are in gets judged at full resolution, including every small irritation that has not had time to fade. The week from last year gets judged from a calm distance, months after anything about it could still sting. That asymmetry alone is enough to make this year's version of an anchor week feel worse than last year's, almost regardless of what happened in either one — not because this year actually was worse, but because only this year is still close enough to hurt.

What a real comparison actually looks for

Once you have entries from both weeks — written at the time, not reconstructed afterward — the differences worth noticing are rarely about which week was "better." They tend to be smaller and more specific:

• A category that used to fill the week and doesn't anymore. The holiday used to be almost entirely one colour in the log and this year it isn't, or the opposite.
• Someone who used to show up in it. A name that appeared in every entry from last year's version of the week and is simply absent from this year's.
• How the week tends to end. Last year's entries from the last day of term read as relief; this year's read as something else entirely.
• The same weekday against the same weekday. If the anchor landed on a Thursday last year and a Monday this year, compare Thursday to Thursday and Monday to Monday within each week rather than day-of-week against day-of-week across them, so you are not quietly comparing a weekend to a workday.

None of this needs a duration attached to it. A line that just says what happened, filed under a category and a colour, is enough to place next to the same line from a year earlier.

What it takes to check instead of guess

The only way around the asymmetry is to stop asking memory to hold last year's week at all, and read what was actually written down instead. TimeMap keeps each day as a handful of short entries with a category colour, filed to the calendar they happened on, and lets you put a stretch of days from one year next to the matching stretch from the year before. Instead of putting this year's tiredness up against last year's memory of tiredness, you are putting a written week against a written week — the only pairing where the comparison is actually even.

You do not need to remember how this week usually goes. You need the version of it that was written down while it was still happening, sitting next to last year's, so the question can be answered by reading instead of by however this particular Tuesday happens to feel.