How to Rate Your Mood 1 to 10 Without the Number Drifting

2026-09-08 · Yesterself

A mood rating is a measuring instrument you build yourself. Most people build it once, casually, on a Tuesday, and never calibrate it again.

You write a 6 in March. You write a 6 in November. The two sixes sit in the same column and look comparable. They probably are not. March's six meant tired but basically fine. November's six means got through it. The number moved underneath you, and nothing in the record says so.

That is the real problem with rating your mood 1 to 10, and it has a fix that takes about twenty minutes once. You define the numbers in words — specific, boring, observable words — and keep the definitions where you can see them.

The number drifts, and nothing warns you

Drift happens because a bare number has no unit. A 6 out of 10 is 6 relative to what? Relative to how you have felt lately, which is the very thing you are trying to measure. When a rough stretch runs long, the scale slides to fit it, and a day that would have been a 4 in spring gets a 6 in winter because it beat the last five.

This is not a character flaw. It is what happens to any scale without a fixed reference. A thermometer works because 0 and 100 are pinned to something outside it. Your scale needs the same trick: each number pinned to a fact about the day rather than a feeling about it.

Anchor each number in something you can see

Define each point by behavior, not by adjective. "Sad," "okay," and "good" drift. "Ate standing up" does not.

Compare two definitions of the same rating:

Next November the first gives you nothing. The second is checkable: you either answered messages or you didn't. A year later you can reconstruct the day well enough to compare it against today.

Good anchors share a few traits:

One warning: do not anchor on productivity alone. If every number is about output, you have built a work tracker with a mood label, and it will tell you rest days are 2s.

Why five points may beat ten

Ten points feel more precise. Whether you use all ten, your own log can settle. Count the distinct numbers in your last few months of ratings. If almost everything sits in a narrow band, with the extremes reserved for a catastrophe or a wedding that never arrives, the spare rungs are decoration. And if you have started writing 6.5, the scale has too many rungs and not enough definitions.

In 1956, the psychologist George Miller published a paper on the limits of judgment along a single dimension: sorting things on one scale — pitches, loudnesses, points on a line — people reliably manage only around seven categories before the distinctions stop holding. He wrote about tones, not moods, so treat it as a hint rather than a rule. But it points where your own count probably will.

Five points give enough resolution to see a trend and few enough rungs to define each in a sentence you will remember. Three is too coarse to show movement. Ten asks you to tell a 6 from a 7 on a scale you built in five minutes.

If you have a 10-point history, map it rather than abandon it: 1–2 becomes your 1, 3–4 your 2, and so on. Write the mapping down with your anchors.

Writing your five anchors

Do this once, with two weeks of real days in front of you.

  1. Pull up your last fourteen days. Any journal, calendar, or camera roll works — actual days, not remembered ones.
  2. Find the most ordinary one. Not good, not bad. That day is your 3. Write two or three things that were true of it in plain behavior.
  3. Find your worst day in that stretch and your best. Those become 1 and 5. Do not save the ends for extremes you have never lived; a scale whose ends you never touch is a three-point scale in costume.
  4. Fill in 2 and 4 as the halfway descriptions between them.
  5. Read the five back in order. If two could describe the same day, rewrite one. Each level needs at least one fact the neighboring levels do not have.

A finished set might look like this — yours will not, and shouldn't:

Keep those five lines on the first page of whatever you rate in. Rating without rereading them is how drift returns.

Yesterself is a journaling app built on one loop: each morning it hands you a letter drawn from your own past writing, and your evening reply becomes tomorrow's mail. A number written that way comes back attached to the day that produced it, which is the context an old rating needs. It is not out yet; there is a waiting list rather than a download.

Yesterself is the journal that writes back. Every morning it sends you a letter from your own past writing, and your reply becomes tomorrow's. End-to-end encrypted, no account, no email — your archive is yours to export, free, forever.

Join the waiting list

Keep the conditions steady

Anchors fix the scale. Habit fixes the conditions around it. Rate at roughly the same hour each day, whichever you settle on: a scale rated at 9 p.m. on good days and at noon on bad ones is measuring the clock as much as the mood. Rate the day as a whole, not the current minute — if the morning was a 2 and the evening a 4, write one number and put the split in the words underneath.

If you miss a day, leave the gap. A blank is honest information. Backfilling a week from memory produces a smooth, flattering line that never happened, and there is no chain here worth keeping unbroken.

The number is an index, not the entry

A rating is a handle for finding things later, not a record of the day. Give each number a few words of company — what happened, how you slept, who you saw. A single sentence is plenty; the one line a day approach pairs well with a number for this reason.

Over months, read the numbers and the lines together and look for shapes rather than verdicts: which weekday sits lowest, what a 2 tends to follow, whether the 4s cluster after bad nights of sleep. Resist reading any single rating as a judgment on yourself. One low number is a Tuesday, not a diagnosis.

Re-anchoring without losing the past

Your anchors will eventually stop fitting. A new job, a new city, a new baby, and "went outside for no errand" suddenly describes every day or none.

When that happens, do not quietly edit the old definitions. Write a new set, date it, and keep the old set directly above it. Your log then has a visible seam: ratings before it mean one thing, ratings after another, and you can still read across the join. Re-anchoring once or twice a year is maintenance, not failure.

Not medical advice

None of this is medical advice, and a mood log is not a diagnostic tool. Tracking numbers can make a conversation with a doctor or therapist more concrete, but it does not treat or manage anything on its own. If your ratings sit low for weeks, or you are worried about yourself, talk to a qualified professional — bring the log if it helps.

Common questions

Is a 1 to 10 mood scale actually bad?

It is not bad, just harder to keep stable. Most people end up using only four or five of the ten points anyway, so defining five levels well usually gives you the same information with less drift.

Should I rate my mood in the morning or the evening?

Evening tends to work better, because you are describing a day that has already happened instead of predicting one. The more important thing is consistency: rate at roughly the same hour every day so you are not measuring the clock.

What if my day had a great morning and a terrible evening?

Write one number for the day as a whole and put the split into the words underneath, such as "3, good until about 4pm." The sentence carries the shape of the day better than any single digit can.

What happens if I miss several days of ratings?

Leave the gap rather than filling it in from memory. A missing day is honest information, and reconstructed ratings tend to smooth your log into a line that never happened.

How do I keep ratings comparable a year from now?

Anchor each number in observable behavior rather than adjectives, and keep the anchor list somewhere you reread before rating. When the anchors stop fitting your life, write a dated new set and keep the old one visible above it.