Zhong Chinese now runs on FSRS

When I first built Zhong Chinese, I implemented it using the — legacy — SM-2 scheduler that other language learning apps like Anki use. The main reason was simplicity; SM-2 is a far more pedictable and simpler scheduler than the more modern FSRS. However, from a very early point of development, it became obvious to me that using FSRS would be a much more efficient algorithm to use, and it would save Zhong users potentially dozens of hours of learning repeating flashcards.

Therefore, Zhong Chinese has moved from the legacy SM-2 algorithm to FSRS.

Why We Switched

SM-2 has been the industry standard since 1987. It works, but it treats every learner the same — fixed intervals, static multipliers, no adaptation to individual memory patterns.

FSRS is different. It models three variables per card: Difficulty, Stability, and Retrievability. The result is scheduling that adapts to your forgetting curve, not a generic average from four decades ago.

What This Means for Learning Chinese

Less time reviewing. FSRS achieves equivalent retention with roughly 20-30% fewer reviews. For a curriculum with 5,000+ vocabulary items, that’s dozens of hours saved over the course of your studies.

No more review snowball. The algorithm is better at predicting when you’re about to forget something, which means fewer “lapsed” cards that reset to zero and pile up in your queue.

It gets smarter over time. After a brief calibration period, FSRS learns your personal retention patterns. The longer you use it, the more efficient it becomes. Nothing Changes on Your End

You don’t need to do anything. The algorithm runs quietly in the background, scheduling your reviews at the optimal moment. Open the app, do your cards, trust the system.

I built Zhong Chinese for serious learners who want to actually acquire Chinese — not play at learning it. Moving to FSRS is part of that commitment: better science, less wasted effort, more characters in long-term memory.

That’s it. Back to studying.