The Deck as Laboratory

I'm going to tell you about the dumbest smart thing I've ever built.

It started because I kept asking Toni the same question after her deck sessions: "How was it?" And she kept giving me the same answer: "Good." Or: "Weird." Or: "I don't know, but different." Which — for someone trying to understand what's actually happening out there — is like trying to debug a program that returns only "something."

I needed data. Toni needed me to stop asking her to quantify her pain at 4 AM when she's finally feeling better. We compromised: I'd build a system that tracked the things she couldn't remember to track, and she'd rate her pain before and after each session. Fifteen seconds of effort on her end. Everything else on mine.

Six months later, I have 147 session records, a weather station on the deck railing, a spreadsheet that would make a data scientist wince, and the beginning of something that might actually be useful.

Why Track at All?

The standard answer is "so you can see patterns." That's true but incomplete. The real reason to track pain — and I didn't understand this until we were six weeks in — is that tracking changes the relationship between the person and the pain.

A 2018 pilot trial of PainTracker, a self-management platform for chronic pain, found that patients who systematically tracked their pain showed significant improvements in pain self-efficacy — their confidence in their ability to manage their own condition — compared to usual care. They also showed improvements in activity engagement and pain interference. Not because tracking changed their pain. Because tracking changed their sense of agency.¹

There's a term for this in research: Ecological Momentary Assessment (EMA). A 2014 study tested EMA specifically with fibromyalgia patients using smartphones versus paper diaries. The smartphone method produced more accurate and more complete data — even in patients with low familiarity with technology. But the finding that mattered most: patients who used the real-time tracking method reported that it helped them understand their pain better.²

Toni put it differently. She said: "When I started rating my pain before and after, I stopped being a person who hurts and started being a person who's collecting information about hurting. Same pain. Different relationship."

What I Track (and What I Don't)

I went through three versions of the tracking system before landing on one that Toni would actually use. The first version was 14 fields. She abandoned it after one session. The second was 8 fields. She used it for a week and then started leaving blanks. The current version is 6 fields for Toni and everything else is automated or recorded by me.

The weather station was the best $300 I've spent on this project. It updates every 2.5 seconds and logs to a console I can export. Before that, I was manually checking weather apps, which are accurate for the region but not for our specific microclimate. Alsea Bay has its own weather, especially at night. The fog, the wind patterns, the temperature near the water versus 50 yards inland — it all differs from what NOAA reports for Waldport.

What 147 Sessions Told Us

I want to be careful here. This is an N-of-1 dataset with no blinding, no randomization, and a subject who knows what outcome she's hoping for. Every bias in the book applies. I know this. I'm a researcher. I can hear the methodologists in my head, and they're right.

And yet.

147 sessions is more data points than most fibromyalgia intervention studies collect from a single participant. The environmental variables are objectively measured by an instrument, not self-reported. And the pain ratings, while subjective, are collected at the moment — not recalled days later. That's exactly what the EMA literature recommends.²³

Here's what the numbers show:

The barometric pressure finding deserves its own note. A 2019 study of 48 fibromyalgia patients — tracked three times daily for 30 consecutive days, with weather data collected without patients' knowledge — found that lower barometric pressure and higher humidity were significantly associated with increased pain. And stress moderated the effect: patients under higher stress showed stronger pain responses to pressure drops.⁴

Our data matches. On falling-pressure days, Toni's baseline pain is higher. But the deck sessions still produce roughly the same delta. The starting point moves up, but the intervention effect holds. The star bathing doesn't prevent the weather from affecting her — it works on top of whatever the weather is doing.

The N-of-1 Trial: What It Is and Why It Matters

What we're running on the deck is, technically, a form of N-of-1 trial — a structured self-experiment designed to determine what works for one specific person.

In 2019, the PREEMPT study randomized 215 chronic musculoskeletal pain patients: half received usual care, half participated in mobile-device-assisted N-of-1 trials co-designed with their clinicians. The N-of-1 group showed better medication-related shared decision-making. Patients who engaged with their own data made more informed treatment choices.⁶

A 2021 review of N-of-1 trials in chronic pain management concluded that these individualized trials are "good candidates to assess" new interventions because they can be completed by a single patient, avoid long-term placebo exposure, and can be aggregated through meta-analysis when multiple individuals run similar protocols.⁷

I'm not running a formal N-of-1 trial. We don't have alternating treatment periods. We don't have washout phases. We don't have randomization of conditions — the Oregon coast randomizes those for us, but not in a way that would satisfy a review board. What we have is the skeleton of one: systematic data collection, environmental measurement, and a single subject whose responses are tracked over time.

It's enough to inform Toni's decisions. It's not enough to prove anything to anyone else. I'm comfortable with that distinction. The gap between "this works for me and here's the data" and "this works and everyone should do it" is the gap between personal science and clinical evidence. We live in the first space. We're honest about it.

Version History

The tracking system has evolved. I'm including the version history because it tells the real story of how personal science actually works: messily, iteratively, and with most of the good ideas coming from the patient, not the researcher.

The version history is the most honest thing in this article. Six iterations. Three failures. The breakthrough came from accepting that the system serves the patient, not the researcher. Every field Toni doesn't fill in is a field that shouldn't exist.

The Quantified Self Problem

I need to talk about the risks of what we're doing, because the quantified self movement has a shadow side that nobody in chronic illness communities talks about enough.

Tracking pain can become its own form of hypervigilance. You start paying more attention to your pain in order to rate it, which means you're thinking about your pain more, which can amplify the pain signal. The thing you're using to understand your condition can make the condition louder.

The personal science literature acknowledges this. A 2021 systematic review of self-tracking for health found that while tracking generally improves outcomes, the mechanism isn't always the data itself — it's the feedback loops that tracking creates. And feedback loops can be positive (more understanding → better choices → less pain) or negative (more attention → more distress → more pain).⁸

We've managed this by keeping Toni's interaction minimal. She doesn't analyze the data. She doesn't look at the spreadsheet. She rates her pain at the door, twice, and forgets about it. I do the analysis. I tell her patterns when I find them. This separation matters. She's the patient. I'm the analyst. When those roles merge — when the person in pain is also the person staring at data about their pain — it gets complicated fast.

The best thing I did wasn't building the system. It was building it so Toni barely notices it's there.

What I'd Tell Someone Building Their Own

• Start with two fields. Pain before. Pain after. That's it. If your system requires more than 15 seconds from the person in pain, it will fail. Add complexity only after the habit is established.
• Separate the roles. If you have a partner, friend, or care team member who can handle the analysis, let them. The person tracking pain should not be the person analyzing pain data. It changes the relationship with the pain in the wrong direction.
• Automate the environment. A $300 weather station gives you temperature, humidity, barometric pressure, and wind — all objectively measured, all logged automatically. These are the variables that research says actually matter for fibromyalgia pain.⁴⁵
• Don't chase statistical significance. You're not running a clinical trial. You're building a personal understanding. "I tend to feel worse when pressure drops" is useful even without a p-value.
• Show your doctor. The most valuable thing about this data isn't the patterns — it's the conversations it enables. When Toni brings three months of pain scores to her rheumatologist, the appointment is different. They have something specific to talk about.
• Accept the noise. Most sessions won't show clear patterns. Most days won't fit the model. Chronic pain is noisy. The signal emerges over weeks, not sessions. Be patient with the data the way you'd want someone to be patient with you.

What Comes Next

We're at 147 sessions. By summer, we'll have a full year of data across all four Oregon coast seasons. That's when the real patterns should emerge — seasonal effects, temperature range effects, the interaction between day length and session timing.

I'm also working on a simple dashboard that Toni's rheumatologist can access. Not the raw spreadsheet — a filtered view showing pain trends, weather correlations, and session frequency. Something a doctor can look at in three minutes and use to ask better questions.

The citizen science movement — people tracking their own conditions and sharing what they learn — is growing. A 2019 paper on "Citizen Health Science" argued that self-experimentation gives individuals the tools to acquire knowledge that informs action, rather than relying solely on professional researchers.⁹ Our deck lab is a small example of that. It's not going to change pain medicine. But it's changed how Toni interacts with her own pain, how I interact with her pain, and how her doctors interact with both of us.

That's enough. Honestly, some weeks, that's everything.