Most apps measure your progress with a streak: a number that only tells you how many days in a row you opened the app. SayLocal answers a harder, more useful question instead. How ready are you actually, and how long until you reach the next level? Here is how that estimate is built, and why we would rather show you an honest number than a flattering one.
A streak is not progress
A streak measures attendance, not ability. You can hold a 200-day streak and still freeze at a ramen counter, because the streak never knew what you could say. It is a motivation trick, and a fine one, but it is not a measurement. If you are about to get on a plane, you do not need to know that you showed up. You need to know whether you can handle the conversation.
Step one: where you are
The forecast starts from how many words you actually know, drawn from your lexicon model rather than from how many cards you have seen. It maps that count onto the CEFR scale using rough vocabulary thresholds: about 800 known words for A1, 1,500 for A2, 3,000 for B1, 5,000 for B2, and 8,000 for C1. Those numbers are not arbitrary: research on how much vocabulary it takes to actually read and listen puts the bar around 8,000 word families for comfortable, unassisted reading.[1] Vocabulary size is not the whole of fluency, but it is the single best-measured predictor of where a learner sits, which is why it anchors the estimate.
Step two: how fast you are moving
A level is only half the answer. The other half is your pace, and we measure yours rather than assume it. The model looks at how many new words you have genuinely added over the last two weeks and how well you are retaining them, using the share of your reviews you recall successfully. Until there is enough of your own data, it falls back to a careful population estimate, then switches to your real numbers as they come in. Pace times distance gives a days-to-next-level figure that is yours, not a brochure average.
We would rather tell you that you are nineteen days from A2 and have you believe it than show you a green ring that means nothing.
Step three: stay honest about uncertainty
An estimate from two weeks of data is not a promise, and we do not dress it up as one. Early on, the forecast is wide and we say so; as you accumulate reviews, it tightens around your real behaviour. If your pace drops, the date moves out. That is the point. A forecast you can trust has to be allowed to deliver bad news, or it is just another streak with extra decimals.
Readiness is the trip-shaped version
For a specific trip, the same honesty gets reshaped into a readiness score by category: survival, restaurants, dating and texting, listening speed, pronunciation, cultural naturalness. Each one climbs as you finish the missions that feed it, so you can see exactly where you are strong and where you are still exposed before you land, instead of a single number that hides the gaps.
What this means for you
You get a real read on where you stand and a date you can plan around, both built from what you can actually do and updated as you do more. It will sometimes tell you that you are further out than you hoped. That is the version worth trusting, and it is the one that gets you off the plane able to speak.
References
- 1. Nation, I. S. P. (2006). How large a vocabulary is needed for reading and listening?. The Canadian Modern Language Review, 63(1), 59–82. Link