Checking the stats at 3 a.m.: metrics, likes and self-worth

There's a scene I know well, and I suspect plenty of people who put things online know it too. It's late, very late. The flat is dark, the day has nothing left to give, and instead of going to sleep I open the stats app. Not because I'm expecting anything in particular. I open it the way you open the fridge when you're not hungry: just in case.

At that hour the number has barely moved since I last checked. I know that before I tap. I tap anyway. And then, depending on what I see, I either drift off calmly or lie there for ages chewing over something that, in daylight, wouldn't have cost me five minutes.

I'm not going to share my figures here. They're not needed, and besides, they're not the problem. They could be double or half and the scene would be exactly the same. What I want to talk about is what happens inside when a number starts calling the shots on your mood, why platforms are built to keep us coming back, and what's helping me make the stats inform me without ruling me.

Why we keep checking

For a long time I thought checking my stats at odd hours was a character flaw. Anxiety, insecurity, a lack of discipline. Something wrong with me. Then I read a little about how reward-based learning works and felt calmer, not because it let me off the hook, but because I understood I'm neither alone nor especially strange.

In the mid-twentieth century, the psychologist B. F. Skinner studied how animals learn when an action brings a reward. He found something that now turns up in every psychology textbook: if the reward always arrives, the behaviour is learned, but it's dropped quickly once the reward stops. If the reward arrives only sometimes, with no pattern you can predict, the behaviour becomes stubborn. It's repeated far more often and it's much harder to give up. This is called intermittent reinforcement.

In plain terms: what hooks us most isn't the thing that always pays out, it's the thing that pays out now and then, without warning. That's why slot machines work the way they do. Most of the time nothing happens. But every so often the little tune plays, and your body learns that the next pull might be the one.

Now think of an artist profile, a social feed or the counter under a video. Most times you look, nothing. A number much like yesterday's. But every so often there's a jump: someone shared a song, a lovely comment, a spike you didn't see coming. That unpredictable reward is precisely the kind that sets a habit hardest. Nobody needs to have designed it with bad intentions for it to work like this. The format does it on its own. And many apps add gestures that feel a lot like pulling a lever: swipe down to refresh, watch the little wheel spin, wait.

Knowing this doesn't cure me, but it changes how I talk to myself. When I catch myself in the small hours with my thumb on the app, I no longer think "how pathetic". I think something more useful: "this is the lever, not a need".

When the number becomes a verdict

Looking at a figure is one thing. Letting the figure tell you who you are is something else entirely.

Stats, in principle, answer small questions: how many people listened to this, from which country, in which week. Those are useful questions. The trouble starts when, without noticing, you ask them a much bigger one: am I worth anything? Is what I make worth it? Should I keep going?

A number can't answer that. But if you ask it every night, it ends up answering anyway, and almost always badly. A low day turns into "nobody cares about this". A high day turns into a relief that doesn't last and wants repeating tomorrow. Neither reading is true. Both are the number playing judge at a trial it has no business presiding over.

I have a song, "So Insignificant", with a chorus that goes "I'm of little importance, / So insignificant". I didn't write it with stats in mind, but rereading it now I recognise that thought. It's the one that shows up when your mood is already low and finds a figure that seems to confirm it. The number doesn't create that thought. It hands it a megaphone.

There are a few signs that help me notice when things have tipped from informing to ruling. They're not a test or a diagnosis, just clues:

• I check the stats without a specific question, just to look.
• What I see changes my mood for hours.
• I stop enjoying a song of mine because it "didn't work".
• I check before bed or the moment I wake up, almost without deciding to.
• I struggle to celebrate someone else's good news because I instantly compare.

If any of these is constant, it doesn't mean something serious is wrong with you. It means it deserves some attention.

Comparison, that late-night sport

In 1954 the psychologist Leon Festinger proposed that people tend to assess themselves by comparing themselves to others, especially when there's no objective measure of how they're doing. It's very human and not always harmful: comparing can help you learn or get your bearings.

The trouble is that online, comparison is rigged. You see other people's numbers but not their context. You don't know whether that song with ten times your plays landed on a huge playlist, whether there's a campaign behind it, whether its creator has been at it fifteen years or two months. You're comparing your back room with someone else's shop window. And you're doing it with a bare figure that flattens everything.

Add to that something research on social media has pointed out many times: upward comparison, with people who seem to be doing better, is the kind most likely to drag your mood down, and it's exactly the kind platforms make easiest, because what gets shown most is what's most successful.

In my case there's one more wrinkle. My music is made in a particular way: I write the lyrics, compose and produce, and the music and vocal performance are made with artificial intelligence. Some people don't get it and some people are annoyed by it. If I compare myself with an artist who fills venues, I lose on any chart. But that comparison makes no sense. We don't do the same thing, we're not after the same thing, we didn't start from the same place.

The only comparison that makes sense is with myself, and not with my numbers but with my work: whether this lyric is better written than last month's, whether this video tells what I wanted it to tell more clearly.

What a number can't measure

In "El segundo exacto" there's a question that still strikes me as the best answer to this whole obsession: «¿Será que podrá la inteligencia artificial escoger el segundo exacto donde rompiste a llorar?», roughly, "Will artificial intelligence be able to pick the exact second you burst into tears?". I wrote it as a reply to people who think AI does everything by itself. But it works just as well for stats.

No dashboard knows the second someone was moved by a song. It doesn't know whether a person heard it once on the metro and laughed, or played it three times in a row because it spoke to something they were going through. It doesn't know whether someone read the whole lyric, sent it to a friend, downloaded it to have on their phone offline. It measures what it can measure, which is far less than what matters.

I've already written in "What Spotify doesn't tell you about an artist" about what gets left out of a profile. Here I want to add something else: what gets left out isn't only information about the work. It's information about you, too. Stats don't know how many hours you spent tuning a line, how many versions you threw away, how hard it was to sit down and write on a bad day. And those are precisely the things that depend on you.

Habits so the stats inform without ruling

I don't believe in heroic fixes, the "delete every app and go live on a mountain" kind. I need to know how my music is doing, and anyone who puts work out there has every right to want to know. What helps me is putting the data in its place. These are the things I try to do, with more or less success depending on the week:

• A set time to look. I choose a specific moment, mid-morning and not every day, to check stats. Outside that window, no. Checking on a schedule turns an impulse into a task, and tasks weigh less.
• Never in bed. Not when I lie down, not when I wake up. At those hours your mood is more fragile and any number reads worse. Charging the phone outside the bedroom, or at least out of reach, helps more than any resolution.
• Notifications off. I switch off alerts for stats, followers and reactions. If something matters, I'll see it in my slot. Most phones also let you schedule rest or do-not-disturb modes and see how much time you spend in each app. That figure, funnily enough, is one of the few worth looking at.
• A question before opening. I go in with a question: which countries is this song being played in? Have new people reached the website this week? Once I have the answer, I close it. No question, no opening.
• Trends, not single days. One day says nothing. A month says something. Comparing longer periods takes the drama out of the ups and downs, which are normal in any project.
• Hide what you can. Some networks let you hide like counts on your own posts and other people's. It doesn't change what happens, but it changes what you see, and that's already a lot.
• Measure what does depend on me. I keep a different tally: lyrics finished, hours of writing, songs revised, ideas jotted down. It's a statistic that grows if I work and doesn't depend on anyone pressing a button.

And one more, the hardest for me: when I see something good, to really celebrate it, without rushing to check whether it happens again tomorrow. Relief that wants repeating isn't joy. It's the lever again.

When a schedule isn't enough

All of the above are habits, and habits help, but they have limits. If the distress around numbers is part of something bigger, a schedule won't fix it.

In other posts on this blog I've said that I live with depression, anxiety and social phobia, and that a good psychologist is helping me. I'm not bringing it up to talk about myself again, but because I know from experience that when your mood is very low, any figure becomes ammunition against yourself. In those moments the problem isn't the app. It's what the app finds inside you.

If for weeks you've felt worthless, nothing excites you, you can't sleep, or your head keeps telling you that you don't matter, it isn't about willpower or using your phone less. It's something that deserves professional attention. You can start with your GP or a licensed psychologist. And you can start with something smaller, too: telling one person you trust.

Music keeps me company a lot, and writing songs helps me put my feelings in order. But it doesn't replace therapy. Not mine, not anyone's.

If you'd like to read more about this, in "Mental health without taboo" I've gathered all the posts along with a map of where to find help.

If you're having a hard time today:

• In Spain, 024 is the suicide prevention line: free, confidential and available 24 hours a day.
• If there's immediate danger, call 112.
• In Latin America and the rest of Europe, call your country's emergency number.
• Talk to someone you trust and reach out for professional help. Asking for help works.

The stats will still be there tomorrow, at a decent hour. You deserve to be there too.