Fogbreaker
Streaks work, right up until the moment they stop, and then they tend to take the habit with them. Almost everyone who has run a walking streak knows the shape of it: forty days, then a chest infection or a delayed flight or a day that simply got away, and then not five days at forty-one but nothing at all for three weeks.
That pattern is usually explained as a motivation failure. It is better explained as a design failure, and the numbers to see it with have been available since 2010.
What the habit research actually found
In 2010 Phillippa Lally, Cornelia van Jaarsveld, Henry Potts and Jane Wardle published How are habits formed: Modelling habit formation in the real world in the European Journal of Social Psychology. Ninety-six people picked a daily eating, drinking or activity behaviour, did it in the same context each day for twelve weeks, and rated how automatic it felt.
Two findings came out of it. One is famous and one is not.
The famous one: the median time to reach the automaticity plateau was 66 days, with an enormous individual range of 18 to 254 days. This is the finding that killed the "21 days to form a habit" folklore, and it is quoted everywhere.
The one that matters more here: missing a single opportunity did not materially affect the habit formation process. One skipped day left the curve essentially where it was, and the automaticity score recovered. What did hurt was missing repeatedly, several times a week, or in runs.
So the habit literature and the streak counter disagree about a specific day in a specific way, and the disagreement is total.
The punishment curve is upside down
Line the two up.
| Days missed | Cost to the habit | Cost to your streak |
|---|---|---|
| 1 | Close to nothing | Everything. Reset to zero |
| 2 to 3 in a row | Real, measurable damage | Nothing more. Already at zero |
| Repeatedly, weekly | Prevents automaticity forming | Nothing more. Already at zero |
The streak charges its entire penalty at the point where the actual cost is smallest, and then charges nothing at all across the range where the real damage happens. By the time you are in the genuinely dangerous territory, day two and day three, the mechanic has already spent its leverage and has nothing left to say to you.
Worse, it has actively inverted the incentive. On the evening of a missed day, the streak's message is that today was a write-off and tomorrow is worth the same as any other day, which is precisely the moment you most need to be told the opposite.
This is a well-documented shape in self-regulation research more broadly. Polivy and Herman described it in dieters as the abstinence violation effect, informally the "what the hell" effect: once a self-imposed rule is broken, the breach itself becomes the reason to abandon the whole project, and the collapse is much larger than the original lapse. A counter that goes from 40 to 0 is an unusually vivid way of announcing that a rule has been broken.
Which is why the freeze exists
The app industry worked this out empirically. Duolingo's streak freeze, copied widely since, exists because the alternative was watching long streaks end and users leave with them. A freeze is not a kindness. It is an admission that the raw mechanic mispriced a single missed day.
Ours does not have one, and we should say so directly, because this article would be dishonest otherwise. Fogbreaker's streak breaks the moment you miss a day. No freeze, no rest day, no grace period. It is the unforgiving version, and everything above applies to it exactly as it applies to everyone else's.
If a hard reset motivates you, that is a genuine preference and our version serves it. If you know you are the kind of person who abandons a project after breaking it, take that seriously as information about yourself, and either use an app with a freeze or use the rules below.
Rules that survive a missed day
None of these require an app to cooperate.
Set the bar at the level you can hit on your worst day, not your average one. A streak's floor should be something achievable at 22:40 in the rain with a headache. Ten minutes. Once round the block. If the floor is 10,000 steps, you are not running a walking streak, you are running a fitness streak that will break the first time life is normal.
Decide your restart rule before you need it. Write it down now: "if I miss a day, I walk the next day, minimum ten minutes, no discussion." The decision is worthless made on the evening of the miss, because that is exactly when the what-the-hell reasoning is loudest. Made in advance it is just a rule you follow.
Count something that does not reset. This is the part most people miss. Keep a cumulative number alongside the consecutive one: total days walked this year, total distance, total streets. A missed day takes the consecutive counter to zero and leaves the cumulative one untouched, which is a much more honest description of what actually happened to your fitness and your habit.
Treat two consecutive days as the real alarm. That is where the Lally data says the damage starts. One is noise. Two is a signal, and it deserves a genuine response rather than the shrug that day one deserves.
Expect it to take longer than you were told. The median was 66 days and the range went to 254. If you are on day 30 and it still feels like a decision every time, you are not failing, you are on the curve.
What a map counts instead
The reason we have thought about this at all is that Fogbreaker keeps two kinds of score, and they behave very differently when you miss a day.
The streak is a streak. It resets, with all the flaws above.
The map does not. It starts under fog, clears where you walk outdoors, and nothing you have uncovered ever comes back. Miss a week, miss a month, come back in March, and every street you cleared is still cleared. There is no decay mechanic and no penalty for absence, because the map is a record of where you have been rather than an assessment of how you are doing lately. Same with XP and levels, which only go up.
That is not us claiming to have solved streak psychology. It is a structural difference worth knowing about: a cumulative record cannot punish you for a gap, and a consecutive counter cannot do anything else. If you use ours, our honest advice is to watch the map and let the streak be a bonus, which is the opposite of how the number is usually presented.
Two limits, as ever. Fog clears from GPS movement outdoors, so an indoor walk keeps a step streak alive and changes nothing on the map. And a map gets less motivating as it fills, whereas a streak gets more motivating the longer it runs, so they fail in opposite directions and neither is sufficient alone.
If you are building the habit from scratch rather than repairing one, the case for walking every day is the place to start, and the hardest version of the problem, fitting it around a desk job, is its own article. The wider set of apps that turn this into a game is in our walking games roundup.
You missed a day. The research says it cost you almost nothing. The counter is the only thing claiming otherwise, and it is wrong.
