the schedule

A plan that survives contact with a real week

Immersion methods die from inconsistency, not from lack of intensity. So this plan is built around a job, a roster and ordinary tiredness, with a floor for bad days and a ceiling for good ones.

the principle

Consistency beats intensity, and this is now measurable

Cepeda's meta-analysis of 317 experiments showed the spacing effect isn't just "reviewing helps". The same material studied in spread-out episodes beats massed study for long-term retention, and the optimal gap grows with how long you want to keep the memory. A language you want for life wants daily contact, not weekly marathons.

Hartshorne's 670,000-person dataset found the same thing at the scale of a whole life. After age of first exposure, the variable that predicted ultimate attainment best was years of immersive exposure. Not course grades, not anything resembling talent. Time in the language, accumulated. Everything below is a machine for accumulating it.


the standard day

About two hours, assembled from scraps

sleep
SRS
passive audio
active
work / life
passive
SRS
sleep and rest passive immersion, free minutes active immersion, focused SRS reviews work and life
BlockTimeWhat happensWhy
Morning reviews15–25 minAnki with or before breakfast. Ten to twenty new cards at most. Reviews always beat new cards.Spacing: morning reviews catch yesterday's cards at the useful point on the forgetting curve.
Commute and chores30–60 minPassive audio. Yesterday's episode re-run, podcasts, radio. No pressure to catch everything.Sample size. These minutes are free and they still count as exposure.
Active immersion45–60 minOne episode or one chapter with target-language subs or a dictionary open. Mine five to fifteen sentences as you go.Comprehensible input at full attention. The highest-yield hour of the day, so protect it.
Evening reviews10–15 minClear the day's queue on the couch or in bed.A second spaced episode for today's cards.
Wind-down20–30 minEasy target-language content instead of English: a vlog, a comic, a re-watch. Nothing new required.Low filter, absorbing message. It replaces scrolling you were going to do anyway.
FLOOR — for the wrecked days

Fifteen minutes, non-negotiable

Twelve-hour shift, sick, zero motivation: do only the Anki reviews. Fifteen minutes keeps the spacing schedule alive and keeps the identity ("someone who does Mandarin every day") intact. The only real rule is never break the chain twice.

STANDARD

The table above

Roughly ninety minutes to two hours of deliberate work plus whatever passive audio the day naturally holds. This is the cruising speed that compounds over a year.

CEILING — for days off

Four to six hours, cinema style

A film with target subs, a long reading session, an hour with a tutor, deep mining. Don't try to live here. The standard day is the system; the big day is a bonus.


the long arc

What a year actually looks like

MONTHS 0–2

Foundations

Tones and pinyin, or the alphabet. First three to five hundred words through a beginner deck. Beginner comprehensible-input videos. Phone flipped, passive audio running from day one. No speaking pressure. By the end you can hear where words begin and end, read the basics, and survive a hello-and-how-are-you exchange.

MONTHS 2–6

Building the engine

Kids' shows, learner podcasts, graded readers at level one or two. Sentence mining becomes automatic at ten to fifteen cards a day. First tutor sessions around month four, awkward at first and then oddly addictive. Expect roughly a thousand to fifteen hundred words, seventy to eighty percent comprehension of easy content, and simple spoken exchanges.

MONTHS 6–12

The intermediate climb

Transition to real shows, re-watches first. Web novels, slower podcasts. The deck passes three thousand cards. Tutor twice a week. This stretch feels the slowest, and that feeling is the mid-frequency vocabulary wall from Nation's cost curve, not evidence you're doing it wrong. Expect comfortable easy-native content, two and a half to four thousand words, and real conversations that take effort.

YEAR 2 ONWARD

The language stops being a subject

You watch, read and play in the language because you want the content. Mining tapers to genuinely rare words. Conversations run past half an hour without anyone switching to English. Refold's rough published timeline for conversational comfort is about eighteen months to two years at consistent daily hours. Treat that as the honest horizon.


staying alive

Consistency engineering

Anchor to habits you already have

Attach study to things you do regardless: Anki with morning coffee, audio when the car starts, one episode after dinner. The existing habit carries the new one, so no extra willpower is required. This is exactly what a roster-nudge app does for shift workers.

Track two numbers, not ten

Minutes of immersion per day, active plus passive, and reviews cleared. Not words learned, not streaks, not app experience points. Those measure activity. Hours of comprehensible exposure is the variable the research says predicts outcomes.

Make quitting annoying

Preload the next episode. Keep the deck synced to your phone. Keep a "when in doubt, watch this" list of content you know you can understand. On a tired day the decision needs to have been made for you in advance.

Plan the relapse

You will miss days. Everyone does. The Canadian immersion programs lost between 43 and 97 percent of their students over a decade and the survivors still got fluent. The only rule that matters is never miss twice in a row, and the floor-day exists so a bad day can't become a bad week.