Why we didn't build a streak counter

Almost every language app has a streak counter by now, and it’s easy to see why: it works, in the narrow sense that it gets people to open the app again tomorrow. It’s a habit loop borrowed from games, and it’s good at what it does.

What it doesn’t do is tell you, or anyone else, whether you’ve actually learned anything. A streak measures one fact: did this app get opened today. It says nothing about whether you can produce the past tense of catch without stopping to think, which is the thing that was supposed to be the point.

What we track instead

DigDug doesn’t have a streak. What it does track is closer to the actual goal: which verbs you’ve heard, how many times, and whether you can still produce them correctly after some time has passed. That’s a slower, less flashy number than a flame icon with a day count next to it, and it’s also the one that corresponds to something real.

This isn’t a small stylistic choice — it shapes what the app does and doesn’t build. No leaderboard. No push notification guilting you for missing a day. No points for opening a screen you didn’t actually use. If a feature’s main effect is to make the app itself more compelling to return to, rather than to make a verb easier to recall, it’s off the table.

The tradeoff, honestly

Streaks aren’t purely manipulative — for a lot of people, a visible chain is genuinely motivating, and losing it is a real deterrent to skipping a day. Cutting it is a real tradeoff, not a costless one, and it’s possible some people would stick with a habit longer with one than without it.

We made the call anyway, because a streak optimizes for the wrong number. If you use DigDug for five minutes on a Tuesday and don’t come back until Saturday, that Tuesday session still taught you something, and the app should say so rather than treat the gap as a failure. Progress here is supposed to describe your English, not your relationship with your phone.