For the past few weeks I've been in a slump. My hands keep moving, but nothing lands. When I try to put that state into words, it always comes out the same way: "I haven't done anything this month."
I genuinely believed it. If someone had asked me what I'd done over the past month, I'd have said "almost nothing" without hesitating. I wasn't being modest — that was my honest read on myself.
Then I counted the records, and found out my read was badly wrong. I also found out that one thing really had gone untouched. This article is about both halves of that.
Self-assessment is a division problem — and in my slump, both numbers disappeared
Self-assessment is probably a kind of division. The numerator is what you actually did. The denominator is what you'd meant to do. Only with both in hand can you say "I got through seventy percent" or "I only managed thirty."
What happened to me in the slump wasn't that one of those numbers dropped. Both of them became unrecallable at the same time.
What I'd done wouldn't come back to me. Last week's version of me was a blur. And on top of that, what I'd planned to do wasn't clear either. I'd get as far as "I think I meant to do something about that," and stop. The ground disappeared twice over.
And yet the verdict still arrived. I never ran the division, but the answer showed up anyway: that was a failure.
That isn't a calculation. It's a mood sitting in the chair where the conclusion belongs. Moods don't need evidence. They can declare without counting, and they never argue back. Which is why, at least in my case, my self-assessment during a slump was at its most confident exactly when it was least reliable.
I counted: nine articles and about 52,000 characters in a month
If my gut can't be trusted, the only option is to look at records that live outside my head. So I counted the articles on my own site by publication date — and I set the window to match my own claim, "this past month," meaning June 22 through today.
In that month: nine articles. Roughly 52,000 Japanese characters of body text.
And four of those nine went out on June 22 alone — why to stop hunting for a way not to be nervous and instead perform on the assumption that the nerves will come, how to allocate time and avoid marking errors during the test, how to spend the night before, and what to eat the day before and the day of. Four articles for exam-takers, in a single day.
Before counting, my self-report was "almost nothing." The records said nine articles and about 52,000 characters. That gap isn't a gap in the facts. It's a gap in recall. The work existed. It had only vanished from my head.
The important part is that my memory wasn't slightly off. Zero and nine aren't a margin of error — they're different objects. At least in my case, self-reporting during a slump hadn't lost accuracy; it had stopped functioning.
But I'm not going to end on "turns out I was doing fine"
It would feel good to stop the story here. Even in a slump you're actually getting things done, so go easy on yourself. It reads nicely, too.
But it would be a lie.
On June 12 I published Starting My First Open-Source Project (Open-Source Log #1) — a public declaration that I was going to do open source for the first time in my life. I ended it like this: "I'm still getting things ready, so once it's in a shape I can release, I'll let you know here."
More than a month later, that follow-up still isn't written. There's no entry #2 of that open-source log. Because I haven't released anything. That isn't distorted memory or a bad mood. It's simply undelivered.
So the accurate conclusion is this: what I'd done and what I hadn't were mixed together, in patches. "It was all a failure" is wrong, and "actually I was doing fine" is exactly as wrong.
What breaks in a slump, I think, isn't the direction of your assessment but its resolution. It's less that you skew negative and more that something patchy gets flattened into a single color — and the flattened version happens to land on the dark side. So the fix isn't "praise yourself more." It's "make the patches visible as patches again."
What stalled was the heavy work
The patches had a pattern. Line up what kept moving against what stopped:
- Kept moving: writing and publishing articles — taking what was already in my head and getting it out in sentence form.
- Stopped: the open-source release — deciding what and how much to publish, researching a procedure I'd never followed, and exposing it to other people.
For me, the first is a road I've walked many times. The second stacks three weights at once: a decision, a first-time procedure, and public exposure. They aren't in the same class.
And a slump apparently doesn't stop everything evenly. In my case, the heavy things fell first. Familiar work kept rolling on momentum even on a bad day. Work that required a decision stalled at step one.
The nasty part is that the thing that fell is what stayed in my memory. The heavy work is usually the thing you most wanted to do. When it's stalled, it fills your whole field of view. Nine articles' worth of light work can be moving, and it still looks like total failure.
This is one person's case, so I won't state any of it as a general law — not the heavy things falling first, and not the fallen thing dominating memory. It's only what described my month. Which work is heavy and which is light flips from person to person — for plenty of people, writing is the heaviest thing there is. The point isn't the contents of my list. It's that the act of sorting by heavy and light is itself what helps.
Why memory behaves this way (as supporting context, not proof)
Psychology has long described a link between mood and recall: when your mood is low, negative events come to mind more readily. There's also the observation that people mistake "easy to recall" for "happened a lot." If failures come to mind easily, it feels like there were a lot of them.
That said, none of it proves anything about my experience. Research describes tendencies across groups; it didn't explain my June. All I can claim is one instance in which counting the records diverged sharply from my self-report. Please read this section as inference, not evidence.
In any case, you don't need the research to get to the fix. The conclusion is simple: don't let memory do the grading.
Steps: re-assess yourself on records, not memory
Here are the concrete steps. You don't need all of them. Adding even one changes the quality of your self-assessment on a bad day.
1. Count records, not self-reports
Don't ask yourself "how was this month?" Open something external — a calendar, a submission history, an app log, the page count in your notebook — and count by date and by unit. Why it works: A mood hands you only a conclusion; a record hands you evidence. The numerator that had gone missing comes back.
2. Decide what you'll count on a good day
Articles published, problems solved, pages covered, checkmarks on a calendar. Don't use subjective scales like "effort level." Use things where one unit is obviously one unit. Why it works: If you have to figure out what to count on a bad day, that decision alone is heavy enough to stop you. Get the deciding done while you have energy.
3. Bound the period with actual dates
Don't review with phrases like "lately" or "this past month." Cut it as "June 22 through July 22," with a dated start and a dated end. Why it works: A vague period lets your worst few days stand in for the whole thing. Dated boundaries put the bad days and the good days on the same field.
4. Sort stalled work into heavy and light
Write out everything that's stalled and mark each item: does it require a decision? is it a first time? will other people see it? The ones with the most marks are the heavy work. Why it works: "I'm stalled" turns into "the heavy work is stalled," which moves the fix from personality reform to task design. You can't rewrite your personality, but you can split a task.
5. In a slump, deliberately shift toward the light work
Don't shove the heavy stuff through on willpower. Set aside days for low-decision repetition. For studying, rather than opening a new unit, cycle through things like vocabulary and idioms where nothing but "answer it" is required. If you're using the English problem set on passed.jp, one low-decision way to cycle is this: the moment you finish a set, use the list screen to check your own answers against the correct ones, and go back into the explanation only for the ones marked wrong. Why it works: Light work still runs in a slump. And once it leaves a record, it becomes the numerator the next time you count.
6. Keep your goals outside your head
Write down what you meant to do, with a date. Tell someone, put it in a notebook, declare it publicly. The format doesn't matter. Why it works: If the denominator lives outside you, a slump can't erase it. I could name "the open-source release is undelivered" not because I remembered it, but because it was written in a June 12 article. Without that, it would have stayed an unactionable fog: "I feel like I left something unfinished."
7. Admit what's undelivered — with the scope kept narrow
Write "I haven't released the open-source project for over a month." Don't write "I'm useless." Why it works: A narrowly scoped miss tells you your next move (what to clear first). A miss expanded to your whole character decides nothing — and it drags everything that was working down with it.
8. If you can, have someone other than you do the counting
A friend, a family member, an AI. Ask: "From the records, count what I did between June 22 and July 22." Why it works: While you count, your mood keeps interrupting with "it's obviously going to be a small number." When someone else counts, there's no gap for that voice to get in.
What I've forbidden myself
From here on, these aren't instructions for you — they're rules I've set for myself. If you're caught in the same trap, feel free to copy them.
- Don't issue an overall verdict while still in the slump. A score produced by a broken grader doesn't get treated as a score. Counting is fine. Concluding waits until the counting is done.
- Don't conclude "I did nothing" without looking at the records. It's the mistake I make most, and the easiest one to kill. It depends on how your records are kept, but in my case the counting took only a few minutes.
- Don't swing the other way and let "actually I was doing fine" excuse everything. That's self-defense, and no next move comes out of it. After counting, always name one thing that's undelivered.
- Don't review with the period left vague. When I open a self-assessment with "lately, constantly…," it usually turns out to be an exaggeration.
- Don't attribute the miss to my character. If I stop at "I'm weak-willed," the only available fix is "try harder." Attribute it to the weight of the task and the fixes become: split it, reorder it, move it to a good day.
- Don't charge at the heaviest task on a bad day. You reach for the heavy thing to make up ground, stall again, and add another failure to the record. In my case, leaving the heavy things alone while I was low produced better results.
- Don't over-engineer the tracking. The moment tracking becomes the job, it stops. One box a day was enough for me.
To anyone who feels "it's all been a waste" this summer
Summer is a season where this happens easily. There's more time, so the plans get bigger, and bigger plans have vaguer denominators. Add the heat wearing you down — in my case, my recall seemed to degrade along with it. The numerator won't come back, the denominator was never sharp, and all that's left is the feeling that everything is a waste.
Two things are enough.
First, count. Count the number of problems you've solved since July began, by date. Which way it comes out, you won't know until you've counted. You might be doing more than you thought, or exactly as little as you feared. Either is fine. What matters is that the records answer the question instead of your mood.
Then name exactly one thing you genuinely haven't touched. One — not "everything." Do this whether the count came out good or bad. Skip it and the counting becomes nothing but reassurance.
For me, that one thing is the open-source release. Over a month, and it still isn't done. So I'll put it here at the end of this article: the next thing I need to write is entry #2 of that log.
One more note on scope. Rebuilding with systems rather than willpower when your progress flatlines is covered over in When Your Studying Stalls: Restart With Systems, Not Willpower. That piece is about fixing how you move; this one is about your assessment of yourself becoming unreliable. In order: pin down where you actually are by counting, then read that one. I've tried fixing a plan while my own ruler was still bent, and I came away with the feeling of having fixed it and nothing else.
