There is a particular face people make when they tell you about their two-hundred-day Duolingo streak. It is the face of someone who has accomplished something measurable, who can prove it, and who is not entirely sure what it adds up to. The streak is real. They have shown up every day for two hundred days. The question is what showing up has done for their command of the language, and the honest answer, in most cases, is less than the size of the number suggests.
The problem with streaks is not that they are evil. The problem is that they substitute a proxy for the thing the proxy was meant to track. The thing meant to be tracked is the slow accretion of competence in a foreign language. The proxy is whether or not you completed a lesson today. These two things look related and in fact diverge sharply as soon as the streak takes on a life of its own, which it will, because that is how the design works.
You can see the divergence by paying attention to your own behaviour on the days when the streak is in danger. If you have an evening with no time and the streak is at risk, you will not do the half hour of difficult, useful practice you wanted to do — there isn’t time. You will do the minimum the app demands, which is some perfunctory matching exercise that requires no thought, and you will go to bed with the streak intact and nothing learned. On a day when you have plenty of time and have already completed your lesson, you have no reason to do another, because the streak is satisfied. The system, in other words, manufactures floors but no ceilings: it ensures you do something every day, but it actively discourages you from doing more on any given day. For most learners this is exactly backwards. The right shape of practice for an intermediate adult is not daily but rhythmic — three or four sessions a week of an hour or two each, interspersed with rest. The streak punishes the long session and rewards the trivial one.
There is also the matter of what kinds of practice the streak format will accept. Almost without exception, streak-based apps reward closed-form exercises: matching, multiple choice, fill in the blank. These are pedagogically the cheapest form of practice because they can be machine-graded, and the cheapest is what scales. They are not the practice that builds fluency. The practice that builds fluency — reading whole articles, watching films, having actual conversations, writing things in your target language — does not register in the app’s accounting. So a learner who quits the app to read a book has, by the app’s lights, broken their streak. The thing they ought to be doing destroys the score they have been measured by.
This is not an argument against tools. Spaced repetition systems, vocabulary apps, structured grammar drills — these have their place, and the apps that provide them are well made. The argument is against the gamification layer that sits on top of them and reorients the user from the goal of learning to the goal of preserving the streak. Once that reorientation happens, and it tends to happen within the first few weeks, the user is no longer learning the language. They are managing a number.
The defenders of streaks will say, fairly, that without the streak many people would not show up at all. This is probably true, and for a particular kind of learner — the one whose alternative to a streak is nothing — even the diminished form of practice the streak produces is better than zero. But it is worth being clear that the streak is a habit-formation crutch, not a learning tool. It produces consistency at the cost of intensity. For someone who already has the motivation to study and would do so without external prodding, the streak is almost purely a downside, because it warps the shape of their practice toward the format the app can measure.
The cure is straightforward and unglamorous. Decide how many hours a week you can give to the language and protect them on the calendar like you would protect any other appointment. Do not give yourself credit for opening the app for two minutes on a busy day. Do not feel proud of a year-long streak that consisted of two minutes a day. Track, if you must track anything, the work you have done in hours, or the books you have finished, or the conversations you have had. These are slower to accumulate and harder to brag about, which is part of why they correlate better with progress. The whole point of learning a language is that the measure of success is not the number of days you have logged in to an app. It is whether you can, in fact, use the language. The streak forgets this. You shouldn’t.
