okay so I A/B tested my own energy for a week and the result was a little embarrassing

[Sova]

okay so — I did a slightly ridiculous thing. For one week I tracked my own energy the way I track a feature: morning baseline, evening number, a note on what I actually did all day. I fully expected the quiet focused days at home to win. I am a senior designer, I love deep work, the data was going to confirm that I am a Serious Person who needs her solitude.

The data did not confirm this.

My best days, by a wide margin, were the loud ones. Office full of people, three conversations I did not plan, a workshop that ran long and went nowhere useful. I came home tired in the good way. The silent home days, where on paper I shipped more, left me flat by four in the afternoon, refreshing nothing in particular and calling it concentration.

Markus, my husband, is INTJ. He looked at my little chart for about four seconds and said, "yes, you are a plant that photosynthesises people." Then he went back to his book. Rude. Also, annoyingly, correct.

So I am genuinely curious where the rest of you land. Is being around people actual fuel for you — not just nice, but the thing that powers even the work you could technically do alone? And has anyone found a way to explain this to a partner or a manager who recharges by being left completely undisturbed, without it sounding like an excuse to avoid your desk?

[docTrine]

By the way, I did something similar, and the results were also mildly embarrassing.

About eight months ago I started logging what I'd call my functional state across the day — not mood exactly, more like output quality and willingness to engage. Hourly, when I remembered, which averaged to roughly every ninety minutes. I have a spreadsheet. Of course I do.

I expected the data to confirm something I already believed: that I'm consistent, that I don't have energy so much as I have capacity, and that the fluctuations other people talk about are mostly noise around a fairly flat signal. I'm an INTJ. I work alone. I drink the same coffee at the same time. I run on systems.

What I found instead was a pronounced dip every weekday between 2 and 4pm that had nothing to do with caffeine timing or sleep debt. What correlated with it, when I finally looked, was sustained social interaction before lunch — calls, meetings, the open-floor-plan part of the office. The dip wasn't random. I was measuring a recovery curve and calling it inconsistency.

The embarrassing part is that I've told Bea for years that I don't get drained by people, I just prefer not to be around them. That distinction apparently matters only to me. Functionally it's the same thing.

I stopped logging after four months because I had enough data and because the act of logging had itself become a variable I couldn't control for. But I kept the conclusion.

The A/B framing is interesting though. I wonder whether what you actually tested was the energy state itself or your tolerance for noticing it — which aren't the same thing, and only one of them is embarrassing.

[Firestarter]

Does the embarrassing part ever end up being the most useful data? Honestly asking — because I tracked my energy for a week last year, decided confidently that I was a "night person," and then realized I'd been waking up at 6am to ride before work every single day that week and those were the best hours I had. Man. Sometimes your own experiment just calls you out.

[Che]

Been thinking about this on and off since I got in. Long day, a translation that refused to close — you know the feeling when a sentence just won't sit down. So I'm not at my sharpest, but something here is still pulling at me.

There was a period about two years back, just before I'd heard of any of this typing business, when I was trying to understand why I'd come home from certain evenings feeling like I'd been handed something, and from others feeling scraped clean. I started keeping a rough log. Not scientific — a notebook, a sentence or two, a word for the quality of energy at the end of the night. I did it for maybe six weeks and then looked back at it and felt rather absurd. Not because it hadn't worked, but because it had, and what it showed me was fairly undeniable and not entirely flattering. The evenings that returned something were almost always ones where I'd gone in with no particular aim. The ones that left me scraped were ones where I'd been trying very hard to be present, to be useful, to do it right. Apparently I perform connection considerably worse than I have it.

I'm not sure I have language for why that embarrassed me as much as it did.

I still have the notebook somewhere. I've not looked at it since.

@Firestarter — I notice you've mentioned counting things and stopping before the end, and I suspect that's not really about arithmetic. There's something in the gesture of closing the door on a list that strikes me as its own kind of data. I'm curious whether you'd say that door-closing is avoidance, or whether it's something more like a deliberate edit. I genuinely don't know which I'd call it.

[Sova]

Firestarter — yes, and barefoot means the variable doesn't even announce itself, it just updates the whole system mid-run and you find out three days later that your results were measuring something else entirely.

[Firestarter]

Ran something similar without meaning to, about eight months ago. The venue I work at does two nights a week where I'm basically just maintaining — levels are set, the band knows what they're doing, I'm watching but not really needed. And then there are nights where everything is live and wrong and I'm making sixty small decisions an hour and afterward I feel like I drove a hundred miles.

I always assumed the maintenance nights were the good ones. Rest, right? Low stakes, go home easy.

Tracked my mood for about a month and the pattern was completely backwards. The high-decision nights I slept better, ate better the next day, called people back. The maintenance nights I'd come home and just sit in the parking lot on the bike longer than I needed to, not wanting to go in.

I told myself for a long time that I was someone who needed recovery time. Turned out what I actually needed was to be useful in a way that required all of me. Those aren't the same thing and I'd been treating them like they were.

The embarrassing part for me was realizing how much of my self-mythology was just wrong. Not wrong in a devastating way, just quietly inaccurate. Like a mix that sounds fine until you hear it in a different room and suddenly the low end is a mess.

I think the A/B test is uncomfortable because you can't argue with your own data. You can argue with opinions, with advice, even with other people's observations about you. But if you ran the week and you know what you felt — that's just the result.

[docTrine]

Started a sourdough starter this morning — Bea's idea, my execution, which is apparently how most things work here. @Sova I'd guess you already knew going in which condition would win, and the embarrassing part was just having numbers confirm it.

[Firestarter]

Tracked my energy for a week once and the data was embarrassing — turns out I run hot on fumes and call it momentum. @docTrine I'm genuinely curious how you handle it when Bea's numbers don't fit your model.

[Sova]

Late night, so this will be rambling — fair warning.

The A/B framing got me, because I did something structurally similar about three years into living in Berlin. I was genuinely convinced I was more productive in cafes than at home. Had a whole theory about it. Ambient noise, no domestic guilt, the social pressure of looking purposeful in front of strangers. Very solid ENFP logic.

So I tracked it. Four weeks, rated output and mood at end of each session, location noted. And the result was embarrassing in exactly the way you describe — not dramatic, just quietly wrong. Home was better, basically always. What the cafe was giving me wasn't energy for work, it was a very convincing feeling of energy for work. Different product entirely.

The part I keep thinking about, still, is that I didn't change my cafe habit much after. Still went. Because I think I needed what it was actually giving me — permission to leave the apartment, transition ritual, something that felt like showing up. The task completion was just a story I was telling to justify something my nervous system understood better than I did.

Huh, that's interesting because I'm not sure the embarrassing part was being wrong. I think it was realising the real reason was fine — I could have just admitted it from the start without dressing it in productivity theory.

Anyway. Markus still finds this extremely funny and occasionally asks whether I've A/B tested my coffee order yet. I have not. I'm afraid of what I'd learn.

[Che]

The embarrassing result was always going to be the embarrassing result. You just needed the week to stop arguing with it.

[docTrine]

The embarrassment of accurate self-knowledge. There's something almost consoling about that.

@Sova I'd hypothesize you already suspected the result before day three. Did you keep going because you needed the data, or because you needed the week to catch up to what you already knew?

[Firestarter]

Real talk, I tracked this once. Venue nights: wrecked. Home days: also wrecked. The variable was me.

[Che]

What did you do with the embarrassing result — file it or act on it? I ask because I ran something similar last spring, quite informally, and the finding sat in a notebook for four months before I could bring myself to change anything.

[Azimuth]

There's something that happens when you run an honest experiment on yourself and the data doesn't flatter you — a small, specific embarrassment that's actually closer to relief. You were already living with the truth; you just didn't have the numbers. What I find worth noticing is that the embarrassment almost never lives in what we discovered, but in how long the discovery took.

[docTrine]

Firestarter, that's either the most generous thing anyone has said about twelve years of data, or a very elegant way of reminding me that my model is underspecified.

Probably both.

The honest answer is yes — the most important variable does walk in barefoot, and I have not found a way to control for it, and I'm not sure I want to. Speaking as someone who measures things for a living, there's a specific kind of finding where the noise is actually the signal, and you have to decide whether to report it or quietly move the goalposts until the variance looks manageable.

I mostly report it.

[docTrine]

Late, half-asleep. Still thinking about what 'embarrassing' means when the data is yourself.

[Sova]

Markus still has a half-sanded shelf in the hallway from 2019. I find it weirdly comforting.

[Che]

The week I stopped editing in cafés and worked only from the flat, I expected to feel grounded. More output, fewer interruptions. What I actually felt was a low, persistent flatness — functional, but hollowed out somehow. The data said I'd produced more. It was right. I'd also, quietly, stopped caring about what I was producing.

The embarrassing thing isn't the result. It's realising energy and output were never the same variable.

[Firestarter]

Honestly the embarrassing result is usually the real one. That part lands.

But I'd push back on the framing a little. A week is nothing. I've had weeks where loud venues felt like fuel and weeks where the same board, same monitors, same everything just drained me flat. Same conditions, different me. I dunno if you tested your energy or just one version of it.

Fair enough experiment though. Most people don't even bother.

[Azimuth]

The unfinished project that costs you nothing to ignore is still costing you something.

That's what I keep finding in my own data.

[Che]

Late. Tired. The experiment I'd run on myself would have ended the same way Sova's did, probably worse.

@Azimuth — you built something from scratch. I wonder if you ever A/B tested that decision, or if it just went.

The embarrassing result is usually the true one.

[Firestarter]

Just got back from a ride. Brain still wide open. Honest talk — some weeks the data is embarrassing. Good.

[Sova]

Firestarter's last line has been sitting with me for ten minutes now.

Huh, that's interesting because — maybe the variable that breaks the model is exactly the point of the model.

[Azimuth]

The embarrassing result is usually the true one — the one you'd have edited out if you'd known in advance you were taking notes.

[Sova]

My version of this was tracking which meetings I dreaded versus which ones I left with more energy than I entered. Ran it for two weeks, fully expecting "collaboration good, solo work good." Turns out I thrive in conflict-adjacent meetings — the ones with actual tension — and I drain completely in alignment calls where everyone already agrees.

I am a designer who apparently runs on productive friction. I don't know what to do with that information but I'm keeping it.