Streaks are the Wrong Way to Think About Habits

By Connor Neblett on Thu Aug 08 2024

Editor’s Note: I originally wrote this in July 2021, after seeing Justin Kan (of Twitch fame) launch a new habit-tracking app on ProductHunt. I’m publishing it in 2024 because I want to be better about actually getting my ideas out there, and better late than never. I’ve decided to leave the blog post mostly unedited from that 2021 draft, so please keep that in mind.

I still think a lot about the marshmallow experiment. If you haven’t heard of it, here’s the gist: it’s an experiment in delayed gratification. A kid is given a marshmallow, and told that if they don’t eat it until the researcher returns, they will be given a second marshmallow. In the original Stanford study, the researchers kept up with these children as they matured, to see if maybe the ones who demonstrated self control (and waited for that second marshmallow) demonstrated similar attributes into adulthood. While obviously a bit contrived (and somewhat contested by reproduction attempts), there is of course truth in the idea that focusing on the thing you need to do rather than the thing you want to do can often pay dividends later.

This applies doubly to improving ourselves, where almost all improvements can compound. I think building good habits (read: improving ourselves in a certain area) is one of the most important things we can do as humans. But to be honest, I’ve been more of a “live in the moment” person for as long as I can remember. If I had been one of the kids in the marshmallow experiment, there is a high chance I would’ve eaten the marshmallow right then. In spite of that, I still try to build good habits. But I’m not great at it, which is why there have been many moments in my life where I bit off more than I can chew in an attempt to force myself into productivity, and that has rarely worked out well (and in general is a stressful way to get things done).

To this end, I’ve spent a good amount of time looking at habit-building apps. Just the other day I saw Justin Kan’s new app Kin, which is what prompted me to write this piece. Some apps try to gamify habit-building, such as Forest where building habits plants virtual trees. For a while, my favorite app in this genre was Streaks, which does pretty much what is says on the tin – it helps you keep track of streaks. It has a clean interface where you put little buttons on the main page of the app. These buttons have a little counter on them. If you’ve done the thing you wanted to do that day (meditate, exercise, not procrastinate, not doomscroll twitter), you press and hold the button and the counter increments. That’s it. The goal is to be proud of the streak that you’ve maintained, developing momentum until the habit is such a flywheel that it’s hard not to do it. At least that’s the theory. In practice, I’ve found this system isn’t quite right. Yes, when you’re on top of the streak, the momentum gained can be clear and very helpful, but when you break the streak (most streaks are broken eventually), your momentum is shattered.

A brief interlude for a concrete example with everyone’s favorite bad habit that needs breaking – boozing. Let’s forget about apps for a second and instead focus on the most widely used habit-builder people use when they want to quit alcohol: Alcoholics Anonymous (AA). One key principle of AA also involves streaks, though AA calls them “chips”. You get a chip when you’ve been sober for 30 days, then more as you progress through the months and years without drinking. But people relapse. It’s frequently not a never-ending upward trend until eventually you have your last chip on your deathbed – “sober for 60 years!“. Instead, something happens in your life (or nothing happens at all), the stars align, the bar to drinking becomes too damn low, and you go on a bender. The next morning you wake up with a killer headache, the midday sun is beating down on you, you’re in a dumpster without your pants, and your streak is broken. When it comes time for your next AA meeting you have two options. Option 1 – lie and pretend the bender never happened, convincing everyone else you’re still going strong and maybe holding on to a sliver of that momentum you had before your relapse. Or option 2 – fess up and start over from day one.

I don’t recommend the first option, as you start training yourself to get gratification from people thinking you are doing well, while not actually treating the underlying problem. This can easily spiral downward. More importantly thought, you’re lying. And lying, though something we’ve all done, is almost always counter-productive ^[if you want to be convinced of this fact, that is far beyond the scope of this article, but you might be interested in Sam Harris’ treatment of this subject (aptly titled “Lying”), which is excellent]. So let’s discard that as a non-viable option and we’re left with option two: fessing up.

Now, in a situation without any sort of social accountability, this is where the problem starts. Because now you’re back at day one. You’ve broken your streak. Yes, you woke up in a dumpster, but the parts you remember before the dumpster were actually a lot of fun. It’s been so long since you let your hair down to that extent. You felt truly free for the first time in a while. Your streak is already at zero, if you go out again tonight it will still be zero, and that’s not so bad. You can pick back up on your streak in a couple days.

Before you know it, you’re totally off the wagon.

And of course this doesn’t just apply to drinking. Trying to eat healthy? Well you just had a Big Mac yesterday so now your streak is broken. It’s been so long since you had a chocolate cake, maybe you’ll have one tonight and pick up the healthy eating thing tomorrow. Trying to exercise? You didn’t exercise yesterday and now your streak is broken. You have sooo much work you need to do today and your streak is already broken, let’s exercise tomorrow.

You get the picture.

This is the downside of streaks. While they’re working, you’re picking up positive momentum and that helps. But you’re also picking up what I’m calling “hedonistic backlog” at the same time. All that gratification you’re delaying is gaining a bit of momentum too, and when your positive momentum is shattered, the backlog comes back with a vengeance.

So how do you ameliorate this?

Well, let’s go back to AA for a second. The important thing AA has which my Streaks app does not (and which I conveniently ignored for a second) is the social aspect – shared responsibility and a safety net of encouragement in the form of meetings and sponsors. When you break your streak and you are honest with others about it (as you should be), they have the opportunity to help you bolster that lost momentum. You go to a meeting and say “I fucked up and had a bender” and everyone says “that sucks but we know you can do better”. Enough encouragement and you will hopefully have enough momentum to overcome your hedonistic backlog and push through, starting your new streak strong.

But it’s not always that easy. For starters, most bad habits you’re trying to break (or good habits you’re trying to build) don’t have a huge social support group like AA (there are AA chapters everywhere). On top of that, AA or similar orgs might have certain ideas you might not really vibe with that make it hard to get “stuck in” with the program.

To me, this is the fundamental impetus behind Justin Kan’s aforementioned new app, Kin. The UX is very similar to the aforementioned Streaks app – buttons on a screen that keep track of your streaks. The key difference is they’ve also built a social feature on top of that streak counting, so you can use your friends to hold you accountable and keep you on track.

Now while I think this is a step in the right direction, it’s still not ideal. You need to have friends, and not just any friends but good friends that will actually hold you accountable, even when the things you’re trying to achieve aren’t necessarily things they want for themselves (e.g. friends that will encourage your sobriety even if they’re still on the frequent-party-circuit). Unfortunately not everyone has that type of friends. Then there are other types of habits that might be too embarrassing for you to share in this way, such as a drug addiction people don’t know about or a rampant porn addiction. Remember that a big reason you needed the social safety net in the first place is due to a fundamental issue with streaks and what happens when you break them.

So is there a better alternative to streaks when you want to build or break habits? Possibly!

One idea I’ve been testing for myself over the past couple months (and so far have had had success with): instead of tracking a streak, track aggregates. Don’t focus on how many days in a row you’ve done something, focus on how many days you did it that week, or how many days you did it in a month. To continue the alcoholism example, you can start by tracking the number of days in a week where you didn’t drink, for example 4/7 (57%) days. The next week, you just try do better than last week, until eventually you’re not drinking 7/7 days a week. Sometimes you’ll fail and go back to 6/7, but that is just motivation to do better. Once you are consistently hitting that 100% score for the week, you can start focusing on the month. Hopefully you will be hitting 100% months in no time. The benefit of this is that you get the same positive momentum you’d get from streaks (“I hit 100% sober days again this month!”) but if/when you fail, the loss in momentum is much less severe. Instead of going from 950 days in a row of no drinking to a big fat 0, you’re going from a 100% success rate last month to a 97% success rate this month. The “oh well, streak broken let’s gooooooo” mentality disappears because if you drink again tonight your percentage goes down to 94%, then down to 90%, and so on and so forth. That positive momentum (or at least our aversion to losing) is still helping you bolster the good habit. Sure, it’s still a bit of a dip, but it’s not so bad and next month you’ll get it back up to 100%.

The important thing here is to not make the timescales too large. Too large, and I think you’ll start to see the same problem as streaks. “I had a perfect year last year, but I’ve already drunk once this year and I’m only 20 days in, what’s the point?” is not a thought you want to be having. The goal is to keep the timeframe small enough so that it never feels onerous to do better next and not procrastinate on your goals.

Anyway that’s what I’ve been trying for a bit and I’m seeing better results so far than when I last used the Streaks app, so maybe give it a try yourself. And if some industrious person with more time on their hands than me wants to turn this idea into an app, I would happily buy/use it!

Editor’s Addendum: Someone actually made an app in this vein in the time since I wrote this article. It’s called Polar Habits and it looks quite nice, although I haven’t tried it myself. It’s a slightly different implementation of the idea (it tracks “momentum” for you instead of looking at aggregates over small timeframes), but it is definitely trying to tackle the problem I’ve outlined here, so maybe give it a go if this article resonated with you.