A drill app lives or dies by the quality of its reps. Some kanji drawing apps give you reps that genuinely build the hand; others give you busywork that feels like practice but barely moves your writing. Here is how to tell them apart, so you spend your minutes on the reps that count.
A rep is only useful if it is production
The single test for any drill is whether each rep makes you produce the character from memory. If the app shows you the character and asks you to trace it, or shows it and asks whether you knew it, those are not writing reps; they are recognition or copying. As we keep coming back to, recognition is not recall, the idea explained in recognition versus recall. A drill of a thousand traced characters builds less than a drill of fifty produced from memory, because only production involves retrieval.
What a good drill app does
Four features separate a real drawing drill from busywork:
- It prompts, then hides. You see the meaning and reading; the character stays hidden until you attempt it, so every rep is recall.
- It gives a guide on demand. A faint stroke-order guide you reveal only when stuck, so you build the sequence rather than ride it. See our kanji stroke order app guide.
- It spaces the reps. Weak characters return sooner and known ones drift out, the spacing effect, so your drill is always pointed at what you are about to forget rather than what you already own.
- It keeps reps short and frictionless. A drill you can do in a couple of minutes, with one tap to start, is a drill you will actually repeat.
Which reps actually count
Not every drill rep is worth the same. Only production builds writing:
| Drill rep | What you do | Builds writing? |
|---|---|---|
| Trace a shown character | Copy the lines | Barely |
| Tap a multiple-choice answer | Recognize | No |
| Write from memory, then check | Retrieve and produce | Yes |
| Rewrite a missed one correctly | Fix the error | Yes |
A good drill app fills your session with the bottom two rows and keeps the top two to a minimum.
Why spacing matters in a drill
A naive drill app makes you grind the same list every day. That wastes most of your effort on characters you already know and under-practises the ones you are losing. A good drill app schedules each character so the reps land where memory is weakest, which is the whole point of spaced repetition for kanji writing. The result is that a few minutes a day keeps a large set writable, instead of a long daily slog that still leaves gaps. The reason this works is the same retrieval effect behind all good practice, the testing effect: you learn most from recalling something just before you would have forgotten it.
Avoid drill busywork
A few warning signs that a drill app is giving you busywork rather than practice: it makes you trace characters with a model visible the whole time, it grinds a fixed list regardless of what you know, or it pads sessions with recognition tap-throughs to feel substantial. None of those build handwriting efficiently. The drill you want is almost boring in its simplicity: prompt, blank space, attempt, check, repeat, with spacing deciding the order. That is what the best kanji handwriting practice looks like, and it is what app versus paper comes down to as well.
What a good drill feels like
A good drawing drill should feel like a short, focused test you pass a little more easily each day. You produce a handful of characters from memory, get a few wrong, correct them, and move on. Over weeks, the wrong ones become right, the easy ones disappear into long intervals, and your writable set grows quietly. Kanji Write Practice is built as exactly this kind of drill, free in early access, with the prompts, the canvas, the on-demand guide, and the spaced schedule doing the work for you. For the wider method, see drawing kanji from memory.
A worked drill session
Here is what two minutes of a good drill actually feels like. The app shows you a meaning and reading, study, learning, and the character is hidden. You write it from memory, getting the left side right but fumbling a stroke on the right. You reveal the guide, see the slip, and write the whole character again correctly. Next prompt: a character you have seen a few times, which now comes out cleanly without the guide, so you mark it known and it drifts to a longer interval. Next: a brand-new character, which you cannot produce at all, so you reveal the guide, study the parts, and write it twice. In two minutes you have done maybe eight characters, each one a real attempt, with the schedule deciding the mix of new, weak, and nearly-known. Nothing was traced, nothing was busywork, and the characters you saw were exactly the ones your memory needed. That is the difference between a drill that builds the hand and one that just fills time: every rep was production, and every rep was chosen by what you were about to forget.
