Field Notes
I run safe, well-established protocols on myself, log everything, and publish the write-up whether or not it worked. Every entry is pre-registered before day one and judged against the bar I set while I still had no idea what the data would say. N=1 proves nothing about you, and everything about how to read a study.
Experiments
06
Nulls
02
Inconclusive
02
Effects
02
Logged data points
200
Two of six experiments found nothing. Those are the entries I am proudest of.
FN · The ledger
Every experiment, on the record
The question, the pre-registered primary outcome, the duration, and the verdict. Nulls get the gold edge, because a null that survives publication is the strongest credential this module has.
FN-01
Mar 02 – Mar 22, 2025
21 nights
The fixed sleep window
Does holding an 8.5-hour sleep opportunity at fixed clock times for 14 nights raise my morning readiness score by more than 3 points over baseline?
The least fashionable intervention in this series is the only one that clearly worked.
Pre-registered primary
Morning readiness score (0–100)
FN-02
Apr 21 – May 25, 2025
35 days
Creatine, 21 days
Does 5 g per day of creatine monohydrate for 21 days raise my countermovement jump by more than 1.2 cm over baseline?
+0.6 cm on the jump, +0.9 kg on the scale, and a power calculation that says I never had a chance.
Pre-registered primary
Countermovement jump, best of 5 (cm)
FN-03
Sep 01 – Sep 26, 2025
20 practice days
Ice baths, alternating days
On alternating practice days, is my next-morning soreness at least 0.5 points lower after 10 minutes at 12 °C than after no immersion?
Soreness dropped 0.9 points, and I trust that number about half as far as I can throw the trough.
Pre-registered primary
Next-morning soreness (0–10)
FN-04
Sep 29 – Nov 03, 2025
4 weeks + baseline
The ankle project
Does four weeks of daily, actually-completed dorsiflexion work move my injured ankle's knee-to-wall distance by more than 1 cm?
+3.4 cm on the injured side, +0.2 cm on the control side. Objective, boring, and the best data in this series.
Pre-registered primary
Knee-to-wall distance, left ankle (cm)
FN-05
Nov 03 – Dec 03, 2025
31 days
Foam rolling: the null
Does eight minutes of pre-practice foam rolling, daily for three weeks, lower my next-morning soreness by at least 0.5 points?
Baseline 4.10, rolling 4.12. My flexibility improved for about an hour. My headline did not.
Pre-registered primary
Next-morning soreness (0–10)
FN-06
Jan 05 – Jan 25, 2026
21 days
Tart cherry, in real life
During a heavy training block, does two weeks of daily tart cherry concentrate move my soreness or my sleep quality by at least 0.5 points?
Two co-primary outcomes, zero movement on either, and a dosing footnote that matters more than the result.
Pre-registered primary
Soreness (0–10) + sleep quality (1–5), co-primary
Running tally: 6 experiments · 2 nulls · 2 inconclusive · 2 effects
FN · The method
The same eight sections, sealed in advance
Every write-up follows one pre-registered format: question, sealed pre-registration, protocol, log, result, why it is weak evidence, what I would change, and a verdict split in two. The discipline is the content.
01
Why pre-register
The version of me who writes the question does not yet know the answer, which makes her the only version qualified to set the rules. Locking the protocol, the duration, the primary outcome, and the definition of failure before day one means the me who wants a result cannot quietly renegotiate with the me who has one. Every entry shows its sealed record, timestamped, above its data.
02
Why publish the nulls
The published literature skews sunny because journals prefer effects and file drawers hold the rest, and recovery marketing is that same bias with a budget. A self-experiment series that only reported wins would just be an advertisement for my own habits. So the nulls run with the same layout, the same charts, and slightly more pride. They are the most expensive entries to write and the cheapest ones to trust.
03
Why n = 1 cannot generalize
I am one adolescent athlete: one genome, one training history, one sleep schedule, and one set of biases holding the clipboard. Nothing here estimates an effect in people. At absolute best it estimates an effect in me, that season, against that month's noise. Claims that travel between bodies come from group trials, and those live in the Library and the Truth Engine.
04
What it is actually good for
Learning to read a study from the inside. Run one honest experiment on yourself and every methods section you read afterward becomes three-dimensional, because you have personally felt the pull to move a goalpost, promote a flattering secondary, or round a number in your own favor. And some questions (does this work for me, at my dose, in my season) can only be asked this way. Badly. But better than vibes.
The question you write down before you start is the only question you get to answer.
House rules
The rules I follow
- Only protocols with established safety records and mainstream acceptance. Nothing prescription, nothing hormonal, nothing off-label.
- No extreme temperatures, no extreme durations, no fasting, and no weight manipulation of any kind, ever.
- Rehab protocols come from my physical therapist. I measure them; I do not invent them.
- Anything that hurts (sharp, joint-line, or getting worse) ends the experiment that day, logged as ended.
- My parents and my coach know every protocol before it starts.
- Nothing in this module is a recommendation. See the note at the bottom of every entry.
Boring on purpose. The interesting risks are the ones this module exists to avoid, and a seventeen-year-old self-experimenter earns trust exactly one unexciting protocol at a time.
Education, not diagnosis. This is a student-authored science platform. Nothing here replaces a physician, a physical therapist, or an athletic trainer. Sudden severe pain, numbness, an inability to bear weight, or visible deformity means stop reading and get seen.
FN · Keep going
Where the real evidence lives
My data ends at my own skin. The claims that generalize come from group trials, and these are the modules that read them.