Why streaks work, and when they stop

This site keeps a streak, so it is worth being clear about what a streak does, what it does not do, and where it turns against you.

What actually makes a habit form

The reliable ingredients are unglamorous.

A stable cue. Habits attach to context — a time, a place, a preceding action — far more readily than to intention. “After I make coffee” is a cue. “More often” is not.

Low friction. Every step between the cue and the action is a place to stop. The single most effective change most people can make is reducing the number of decisions between deciding and doing.

Repetition, over a long time. The widely quoted “21 days” has no good evidence behind it. The better-known study on this found a median of around 66 days to reach automaticity, with enormous variation — from about three weeks to well over half a year depending on the person and the behaviour.

Something that makes today’s instance easy to start. Not motivation, which is unreliable, but a small concrete first step.

What a streak specifically does

A streak is a commitment device: it manufactures a small cost for skipping that would not otherwise exist. You are not protecting your health or your reaction time on any given evening — you are protecting the number.

That works, and the reason it works is loss aversion. Losing something you have is felt more strongly than gaining something equivalent, so a 30-day streak is a real thing to lose in a way that “doing it again” is not a real thing to gain.

It is also honest about its own mechanism, which most engagement features are not. A streak does not pretend the activity is intrinsically compelling. It admits it is providing an external reason.

Where it goes wrong

Two failure modes, and the second is the serious one.

Optimising for the streak rather than the thing. Doing the minimum that counts, at 11:58pm, to protect a number. The streak is intact and the activity has become hollow.

Abandonment at the break. This is the real problem. Someone with a 40-day streak who misses a day frequently stops entirely — not because one missed day matters, but because the thing they were protecting is gone and there is nothing left to protect. A counter designed to motivate has become a reason to quit.

That second failure is a direct consequence of how well the first mechanism works. The more successfully a streak converts your motivation into protecting a number, the more completely that motivation collapses when the number does.

How this site handles it

Three deliberate choices, stated because the alternative is the design most apps ship.

A day counts if you played today or yesterday. The streak does not die the instant the clock passes midnight on someone who plays every evening and has not played yet today. It ends once a full day has genuinely been missed.

Your history survives the streak. Breaking it resets one number and loses nothing else. Every result you have recorded is still there, and the trend — which is the actually useful output — is unaffected.

There are no notifications and no emails. There is no account to send them to, and building something that manufactures obligation is not what this is for. If you miss a week, nothing chases you.

Why “21 days” refuses to die

The figure comes from a 1960 book by a plastic surgeon, Maxwell Maltz, who observed that his patients seemed to take about 21 days to adjust to a changed appearance. It was an observation about adjustment to surgery, offered as a minimum, and it was never a study of habit formation.

It spread because it is short, memorable and encouraging. The better-evidenced figure — around 66 days, with a range from about three weeks to well over half a year — is none of those things, which is presumably why it has not replaced it.

The practical difference matters. Someone told 21 days who is still finding it effortful on day 30 concludes the habit has failed. Someone told “usually a couple of months, sometimes much longer” concludes they are on schedule, and carries on.

Making the minimum genuinely minimal

The most common way a daily habit fails is that the daily version is too big.

People set the bar at the good-day version — the full round, the long session — and then miss on the first bad day, because the bar was never calibrated for bad days. The fix is to define the habit as something you would do while ill, travelling or busy, and treat anything more as a bonus rather than the target.

On this site that means one test counts. It does not complete a round and it does not extend the streak, but it records a result, keeps the history continuous, and preserves the cue. A habit that survives a bad week intact is worth more than one that produces better numbers for three weeks and then stops.

Missing a day does not undo anything

The research on this is reassuring and worth knowing before you need it.

A single missed day has no measurable effect on the eventual formation of a habit. The curve is not reset by a lapse; it dips and continues. What derails habits is not the missed day but the conclusion drawn from it — that the attempt has failed, so there is nothing left to protect.

This is exactly the failure mode a streak counter amplifies, which is why the rule that actually helps is never miss twice. It concedes the first miss in advance, so it cannot be read as evidence of failure, and it targets the second one, which is where a lapse becomes a pattern.

Building one without an app

If you want a daily habit that is not on this site, the same principles apply and no software is required.

1. Attach it to something that already happens daily. After coffee, after brushing your teeth, when you sit down at your desk. Borrow an existing cue rather than building one.
2. Make the minimum version genuinely small. One test, not eight. Five minutes, not thirty. A version you will do on a bad day is worth more than a version you will do on a good one.
3. Record it somewhere you will see. A paper calendar with a cross works as well as anything, and it has the advantage of not sending you push notifications.
4. Decide in advance what a miss means. “Never miss twice” is the useful rule, because it targets exactly the failure mode above — one miss is an accident, two is a new pattern.

What to actually track

For this site, the streak is the return mechanism and the score history is the point. A 30-day streak of rounds tells you you showed up. A fortnight of daily scores tells you what a bad night’s sleep costs you, whether your afternoon slump is real, and how much of your variation is state rather than ability.

The second of those is worth having. The first is scaffolding that helps you get it — useful, and not the thing itself.