Skip to content
against.

What the science says about habit-tracking and recovery.

The science of habit-tracking is older and quieter than the gamification industry suggests. Here is what works and what doesn't.

The science of habit-tracking is older and quieter than the gamification industry suggests. Here is what works, what doesn’t, and why against. measures only what it does.

What self-monitoring actually does

Self-monitoring (the practice of recording your own behavior) has a reasonably strong evidence base in behavioral psychology. The mechanism is simple: awareness changes behavior. When you track something, you notice it more. When you notice it more, you’re more likely to make deliberate choices rather than automatic ones.

This applies to eating, exercise, spending, and it applies to compulsive behaviors. The act of logging a relapse, particularly if you also note the trigger and the context, creates a moment of reflection that the behavior itself doesn’t. Over time, the log becomes a mirror: patterns emerge that you can’t see when you’re inside any single episode.

The effect is strongest when the tracking is honest, consistent, and private. Honest means recording the difficult events, not just the clean days. Consistent means doing it close to the event, not reconstructing a week later. Private means writing for yourself, not for an audience.

These three conditions are interdependent. If the log isn’t private, it’s hard to be fully honest. If it’s not consistent, the patterns are noise. If it’s not honest, the consistency is theater.

Why streaks help, and where they hurt

Streak counting (tracking consecutive days without a target behavior) is genuinely useful for some people in some phases of recovery. The psychological mechanism is real: a visible run of clean days creates a mild loss-aversion effect. Breaking a streak has a small cost. That cost can, at the margin, shift behavior.

Where streaks break down is equally documented. A person who has a brief relapse after a long clean run can experience a disproportionate sense of failure (“I’ve destroyed everything”) that is both inaccurate and motivationally counterproductive. In behavioral terms, this is sometimes called the abstinence violation effect: the perceived magnitude of a slip relative to the streak inflates the emotional response, which can itself trigger further relapse.

This is the paradox of streak-based tracking: the longer the streak, the higher the stakes of breaking it. For people in recovery from compulsive behavior, that dynamic can be destabilizing.

against. counts days, but it isn’t built around the streak as a primary metric. The log matters more than the number. What you wrote about Tuesday matters more than whether Tuesday was day 47 or day 1.

Why we don’t gamify

Gamification in health apps usually means points, badges, levels, and leaderboards. These mechanics borrow from game design and apply them to behavior change. The theory is that they increase engagement. The evidence is more complicated.

Short-term, gamification tends to work: people engage more with an app when it rewards them visibly. Longer-term, the evidence for sustained behavior change is weak. The engagement that gamification drives is engagement with the gamification mechanics, not necessarily with the underlying behavior. When the rewards plateau, so does the engagement.

For compulsive behavior specifically, there’s a deeper problem. Recovery from compulsive behavior involves developing a healthier relationship with reward systems: delayed gratification, intrinsic motivation, tolerance for discomfort. Layering an additional dopamine-reward loop on top of that process, in the form of badges and points, is working against the goal rather than toward it.

against. doesn’t have points. It doesn’t have badges. It doesn’t rank you against other users. The only feedback is what you put in and what the log reflects back.

What the literature says about abandonment

The abandonment rate for health-tracking apps is high. Most studies find that engagement drops sharply after the first few weeks, with a large proportion of users stopping entirely by month three.

The reasons are multiple: apps that are too complex require too much per session; apps that are too simple offer nothing after the initial setup; apps built around streaks increase the abandonment rate after a relapse because the user doesn’t want to face the counter reset.

The intervention that correlates most consistently with sustained use is low friction combined with something useful to return to. Low friction means a logging session that takes under a minute. Something useful means the log is legible: you can look back and see something real. That’s the design target for against. Fast enough that you’ll do it in the moment, transparent enough that it’s worth coming back to.

The goal is not to make an app you use forever. It’s to make one that serves the period when you need it.

Further reading

See also: Methodology · For you.