The SRS-to-Immersion Ratio That Most Guides Get Wrong

It's the most searched question in language learning communities for a reason: everyone wants one number. The honest answer is that the ratio shifts by level - and the research is clearer than most guides admit.

If you've spent any time in language learning communities, you've seen the argument a hundred times. One side swears by flashcards. The other side insists immersion is the only thing that builds real fluency. Both are right about half the time, and both are wrong about the ratio.

The research is actually clearer than the forums make it look. Here's what studies on SRS-to-immersion balance actually show - and why the right ratio depends on where you are, not on which camp you belong to.

Why the ratio question matters at all

Time is the binding constraint for almost every language learner. If you have 60 minutes a day, every minute you spend on flashcards is a minute you don't spend reading, listening, or speaking. That's the real trade-off, and it's why the ratio question keeps coming back.

The answer isn't about which method is better in isolation. Spaced repetition and immersion train different things. SRS builds your vocabulary store - the set of words you can recognize and, with enough repetition, actively retrieve. Immersion trains your brain to process that vocabulary at real-time speed, infer meaning from surrounding context, and operate in a language where you don't control the topic.

Neither works well without the other. A 3,000-word vocabulary that you can recite in isolation but not understand in fast native speech is useless in a conversation. Fluent-sounding conversation with a 400-word vocabulary collapses the moment the topic shifts. The learners who make the fastest progress treat SRS and immersion as a system rather than competing strategies.

The research-informed answer: roughly 20–30% SRS

For intermediate learners - the group most of the research focuses on - studies and long-term practical experience point to a consistent range: roughly 20–30% of your study time on structured spaced repetition, with the remaining 70–80% spent on comprehensible input and immersion activities.

In concrete terms, if you study 60 minutes a day, that's 12–18 minutes of SRS reviews and 42–48 minutes on reading, listening, watching shows, or speaking practice. If you study 30 minutes, it's roughly 6–10 minutes of reviews and 20–24 minutes of real-language exposure. The specific split inside that range matters less than staying inside it.

That ratio isn't arbitrary. Research on second language vocabulary acquisition has consistently found that SRS produces the biggest retention gains when the words being reviewed are also encountered in real context shortly before or after review. Pure flashcard study shows diminishing returns once vocabulary size grows, while pure immersion fails to build high-frequency vocabulary quickly enough at lower levels. The 20–30% band is where the two reinforce each other most efficiently.

Why the ratio should shift by level

The 20–30% rule is a reasonable default, but treating it as fixed is where most learners go wrong. The right ratio genuinely depends on your current level, and it should move as you progress.

At the very beginning - roughly the first 500–1,000 words - the ratio shifts toward more SRS. This is counterintuitive if you're allergic to flashcards, but the reason is simple: immersion only works when the input is comprehensible. If 90% of the words in a podcast are unknown, you're not learning from it, you're just listening to noise. At beginner level, a 40–50% SRS split is often more productive because it builds the foundation vocabulary that makes immersion start paying off.

At intermediate level - roughly 1,000 to 3,000 words - the 20–30% rule applies cleanly. Your SRS routine maintains and extends the vocabulary you're actively using, and immersion carries the bulk of the work because you can now understand enough to learn from context.

At advanced level - past 3,000–4,000 words - SRS usually drops to maintenance mode: 5–10 minutes of reviews to keep older cards from decaying, with essentially all remaining study time spent on immersion, output practice, and rare-word acquisition from real content. Adding new cards at this stage becomes optional unless you're studying specialist vocabulary for a specific field.

What counts as immersion for the purposes of this ratio

Not all immersion is equal, and the ratio assumes you're doing immersion that actually works. Passive listening to a podcast while doing chores barely counts. Active engagement with material you can mostly understand counts a lot.

Comprehensible input is the useful definition: material where you understand roughly 80–95% of what's being said. At that level, you can infer the meaning of new words from context, which is how vocabulary genuinely enters long-term memory from immersion. Below 80% comprehension, you're drowning. Above 95%, you're not learning anything new - you're just practicing retrieval of words you already know.

The mistake that breaks the ratio entirely

The most common failure mode isn't getting the split wrong. It's letting SRS and immersion drift into separate worlds that never touch each other.

A learner with 2,000 words in their deck and 500 hours of anime exposure shouldn't have two independent vocabulary lists. The words they encounter in real content should feed back into their deck, and the words they review in their deck should start appearing in their real content. When that loop closes, retention rises sharply because every word gets hit from multiple angles.

Practically, that means adding new words to your deck from the immersion you're actually doing - not from generic top-1000 frequency lists. A word you heard your favorite creator say yesterday is already half-memorized before it hits your deck. A random word from a list has no context and fights for memory space against everything else.

A concrete weekly plan at the 25% split

Here's what a 60-minute-per-day routine at a 25% SRS split looks like in practice, for an intermediate learner. It's not the only way to structure this, but it's representative of what works.

Each day: 15 minutes of SRS reviews (morning, before anything else). Then 45 minutes of input, usually split between 20 minutes of reading and 25 minutes of listening - either a podcast or a show. Twice a week, replace one of the listening sessions with 25 minutes of conversation practice or language exchange.

New cards get added from the immersion, not scheduled separately. If you hit a word during listening that you had to look up, and it seems useful, it goes into the deck that evening with an example sentence taken from the actual context you heard it in. That's where SRS and immersion reinforce each other.

Weekly volume: about 7 hours. Roughly 1h45min of SRS, 5h15min of immersion. At 10–15 new words a day, that's 70–100 new words per week going into the deck with real context already attached. Six months of that rhythm produces the 1,500–2,000 word working vocabulary most learners are chasing.

When the ratio shouldn't be 20–30% at all

There are specific situations where the standard ratio is wrong and should be ignored.

If you're preparing for a specific exam with a defined vocabulary list - IELTS, TOEFL, a university entrance test - SRS should go up, often to 40–50% of study time, until you've covered the list. The goal is different: you're not building general fluency, you're memorizing a finite set of words against a deadline.

If you're at the very end of a language (C1–C2 range), SRS can drop below 10%. At that level, you're mostly maintaining what you already know and acquiring rare vocabulary through reading. Structured reviews become less valuable than volume of input.

If you're maintaining a language you already learned but don't use daily, SRS often rises temporarily - 50% or more - specifically because maintenance is exactly what spaced repetition does best. Immersion still matters, but preventing decay is the main job, and SRS does that efficiently.

How to tell if your ratio is wrong

The symptoms are specific and easy to spot if you know what to look for.

If your deck keeps growing but your ability to understand real content isn't improving, your ratio is too heavy on SRS. You're adding vocabulary faster than you're learning to use it. Cut back on new cards and spend the recovered time on input at your level.

If you're doing a lot of immersion but the same basic words keep surprising you, your ratio is too light on SRS. Your passive vocabulary is drifting and the high-frequency words aren't locking in. A few weeks of heavier review - focused specifically on words you've been hitting and missing in immersion - usually fixes this.

If your review queue is collapsing and you're skipping days, the problem isn't the ratio. It's the total volume. Drop new cards to zero until the queue is manageable, then restart at a lower daily limit. Spaced repetition only works when the spacing actually happens.

How MindDory fits the 25% model

MindDory is built around the assumption that SRS should be fast enough to leave the majority of your study time for real-language exposure. Reviews are designed to fit into 10–15 minute mobile sessions, cards come with example sentences and audio by default so the context side of the equation is covered, and the new-card limits are set conservatively so the queue stays manageable. For a more complete view of the research this ratio is drawn from, the spaced repetition effectiveness article covers the underlying studies in detail.

The ratio is the easy part to get right once you know it. The harder part - staying consistent, closing the loop between what you review and what you consume, and shifting the split as your level changes - is where the real work is.