Duolingo builds habit. Habit is not the same as memory
Many people keep a streak for months or years and still do not feel real vocabulary growth. The problem is not laziness. The problem is mechanics: games, lessons, and daily streaks do not guarantee that specific words return on time and stay in memory.
Problem 1: a linear lesson path
Today lesson 1, tomorrow lesson 2, then lesson 3. But memory is not linear: while you move forward, old words start to fade. CORE95 does not follow a lesson road. It follows the memory of each word: what starts fading returns sooner, what holds well moves further away.
Problem 2: weak knowledge checks
Multiple choice often checks recognition, not active knowledge. If the answer is already on the screen, the brain can guess or recognize it. To make a word yours, you need to retrieve it yourself: type it, recall it, make mistakes, and repeat it. That is why CORE95 focuses on typed answers.
Problem 3: the user level is weakly reflected
You can be around B1/B2 and still regularly see very simple words like boy, apple, or girl. The issue is not the simple word itself. The issue is spending attention in the wrong place: showing what you already know while the words that need reinforcement may not return at the right moment. CORE95 should show words by level and by your memory, not just by a generic path.
What this feels like in practice
The frustrating feeling is that the app seems active, but attention goes past the goal. You keep meeting words you have known for years, while yesterday’s serious new words do not become recallable without a hint. There is movement through lessons, but not enough feeling that the vocabulary is now yours.
Problem 4: streak replaces result
A streak motivates you to open the app. But it does not show how many words reached long-term memory. A better metric is different: how many words can you recall without a hint after a week, a month, and a year?
Problem 5: internet dependency
For some users this is a separate risk: if the app depends on cloud state and streaks, unstable internet breaks the experience. CORE95 is built as offline-first: training continues on the device without internet, and sync catches up later.
Important note: Duolingo also talks about spaced repetition
Duolingo publicly says it uses spaced repetition and models such as half-life regression. So the fair critique is not “Duolingo has no repetitions at all”. The issue is different: in the user experience it does not feel like a transparent full vocabulary repetition system where each word has a personal schedule, strict typed checking, and clear vocabulary progress.
What should exist instead
A real vocabulary progress system should answer simple questions: which words do I already know, which words am I forgetting, what must I repeat today, why did this word appear now, and how far am I from a clear result? Without that transparency, learning easily becomes activity instead of measurable progress.
Final takeaway
Duolingo is good as a habit game. But if the goal is to quickly learn the frequency core of a language, you need different mechanics: strict checking, personal intervals, offline-first, and progress by specific words. The product model rewards daily return to the app more than quickly closing a concrete goal: learn the vocabulary core.
