Tracking & Measurement

The goal of tracking is not to accumulate data: it is to identify patterns that explain your experience and guide your experiments. Less data tracked consistently is more useful than comprehensive data tracked sporadically. The person who writes five data points every morning for thirty days knows vastly more about their sleep system than the person who tracked twenty-three variables for four days before abandoning the practice.

What to Track

The Minimum Useful Dataset

For the tracking practice to be effective, it needs to be simple enough to maintain consistently for weeks. The minimum useful dataset is five data points recorded each morning: bedtime (when you got into bed with the intention of sleeping), lights-out time (when you actually stopped engaging with phones, books, or other input), estimated sleep-onset time (how long it took to fall asleep), wake time (when you woke for the day), and morning energy rating (a 1-10 subjective score of how rested and functional you feel).

These five data points capture the core dimensions of sleep: timing, duration, quality of onset, and subjective restoration.

To these five, add one to three key variables that you are currently monitoring or testing. These should be specific to your current hypothesis about what is affecting your sleep. If you suspect late caffeine is a driver, track your last caffeine time each day. If you suspect alcohol, track whether you drank and approximately how much. If you are testing exercise timing, track when you exercised.

The key variable slot is where the experimental tracking happens: the five core data points establish your baseline, and the key variables are what you are trying to learn about. Keeping the key variable count small ensures that when you observe a pattern, you can attribute it to something specific rather than having too many simultaneous variables to draw conclusions.

What to Add as You Progress

As the basic tracking becomes habitual and the initial patterns become clear, you can expand the dataset selectively. Evening variables that may be relevant include: last meal time (for those investigating nutrition-sleep connections), screen use in the final hour (for those working on the wind-down protocol), room temperature at sleep onset, and any notable stressors or emotional content from the day.

Morning variables worth adding for some people include: whether you woke naturally or to an alarm, whether you snooze, whether you remember dreaming, and whether you have morning headaches (a potential OSA signal worth noting). The principle is to add only variables that you have a specific hypothesis about, not to track everything comprehensively.

Wearable sleep trackers (Oura Ring, Whoop, Apple Watch, Garmin, Fitbit) can be valuable additions to self-tracked data if used correctly. They provide continuous data on sleep duration and, with varying accuracy, sleep staging and HRV. Their value is primarily in identifying patterns over longer timescales than is practical with manual tracking, and in catching objective data that the subjective experience misses.

Their limitations are equally important to understand: consumer wearable sleep staging is substantially less accurate than clinical polysomnography (the gold standard), individual stage accuracy varies significantly by device and individual, and the numbers they produce should be treated as directional indicators rather than precise measurements. The direction of change over time matters more than any individual night’s reading.

Subjective vs. Objective Data

Why Both Matter

The most important tracking principle in the nuyu method is to maintain both subjective and objective data streams simultaneously and to treat them as separate sources of information rather than collapsing them into a single “sleep score.” Objective data (times, numbers, wearable readings) tells you what happened. Subjective data (how you felt, your morning energy rating, your sense of restoration) tells you what the system produced. These two streams are complementary and sometimes diverge in ways that are the most diagnostically useful signals in your data.

The divergence cases reveal mechanism. When objective sleep duration is adequate but subjective restoration is consistently low, the problem is architecture, not duration. When subjective restoration is high despite a short night, you either got high-quality sleep with good architecture, or you are underestimating the subjective cost of the debt you are accumulating. When a wearable shows good deep sleep but your morning energy is consistently low, either the wearable’s staging is inaccurate for you personally (common) or something other than sleep architecture is driving your morning fatigue. Each divergence is a signal that there is something to investigate. The patterns do not always have simple interpretations, but they always have information.

The Subjective Rating System

The morning energy rating (1-10) is the single most important data point in the entire tracking system. It is the output variable: everything else in the tracking is an input or process variable, and the morning energy rating is what tells you whether the system is producing the result you want.

Rate it consistently and early: before checking your phone, before caffeine, as close to the wake moment as possible. Rate how you feel in your body and mind right now, not how you expect to feel after coffee, and not how yesterday went. A rating taken before caffeine on a calibrated scale gives you information about the actual state of your nervous system upon waking. A rating taken after your second coffee tells you mainly about caffeine.

Anchoring the scale with consistent reference points makes the ratings more useful over time. A 10 is how you feel after an exceptional night of sleep on a calm, low-stress day when you wake naturally at the right time: alert, clear, physically good, motivated. A 1 is how you feel after a night of total sleep deprivation. A 5 is “functional but not good”: you can do what you need to do but you are clearly not at your best. Keeping these anchors consistent makes the numbers comparable across weeks, so that “average morning energy rose from 5.2 to 6.8 over three weeks” is a meaningful signal rather than a product of rating drift.

The Change in Score Is the Signal

Why Single Data Points Are Meaningless

Individual data points in sleep tracking are almost entirely uninformative. A morning energy score of 6 tells you very little by itself. Sleep that night might have been affected by anything: a single stressful conversation, a slightly warmer room, a different pillow, a passing illness, random biological variation.

Single nights of unusual sleep (excellent or terrible) are noise: they reflect the sensitivity of the biological system to minor perturbations and the inevitable variation in any complex system. The appropriate response to a single poor night is to note it and continue; the appropriate response to a consistent pattern over two or more weeks is to investigate.

Pattern, not performance, is the signal. A morning energy score that was 4.1 on average for two weeks and is now consistently 6.5 after making a specific change to your evening routine is strong evidence that the change made a difference. A score that is consistently low regardless of what you varied over six weeks suggests either that the primary driver has not yet been identified, or that the problem requires a different class of intervention than you have been experimenting with. The timescale to look for patterns is two to four weeks: less than two weeks produces too much noise to be reliable, and more than four weeks without any pattern emerging suggests the intervention being tested is not the right one.

Controlling for Confounds

A basic principle of self-experimentation that most people skip is changing one variable at a time and tracking it long enough to produce a reliable pattern before changing something else. When multiple things change simultaneously (new sleep schedule, new diet, new exercise routine, new wind-down protocol), it is impossible to attribute any observed change to any specific intervention.

This is the primary reason that the dramatic multi-protocol overhaul approach to sleep improvement tends to either produce unclear results (you feel better but don’t know why, so when things inevitably drift, you don’t know what to restore) or fails entirely when the volume of change is too much to sustain.

The structured approach is to establish a baseline tracking period (one to two weeks with no intentional changes, just recording), identify the primary pattern or question from the baseline, choose one intervention to test, run it for at least two weeks while tracking consistently, and evaluate the result before adding or changing anything else. This approach is slower in the short term but produces genuine understanding of your specific system rather than generic compliance with someone else’s protocol. The understanding is what makes the improvement durable: you know what works for you specifically, and you can return to it when disruptions occur.

What Not to Obsess Over

Orthosomnia and Data Anxiety

Orthosomnia is the term for sleep anxiety driven by excessive attention to sleep tracking data: a person becomes so focused on achieving good sleep scores on their wearable that the anxiety about the tracking actually worsens their sleep. It is a real phenomenon with documented cases, and it is particularly common in people who are already prone to anxiety and who adopt aggressive sleep improvement practices without appropriate framing.

The wearable score that was supposed to help becomes another source of pre-sleep arousal: lying in bed worrying about whether the device will register good deep sleep tonight is the opposite of the mental state conducive to sleep onset.

The antidote is the framing established from the start: tracking is a tool for pattern recognition over time, not a performance metric for individual nights. A single night of poor wearable scores is noise. A week of low scores while changing nothing is a signal. The appropriate response to a poor sleep score is not frustration or corrective intervention: it is to note it, continue the routine, and look for what it correlates with over time. If checking your wearable score in the morning makes you feel worse or more anxious about sleep, stop checking it daily and review weekly aggregates instead. The data serves you; you do not serve the data.

The Felt Experience Remains Primary

No tracking system, objective or subjective, replaces direct attention to how you actually feel. The morning energy rating exists precisely because how you feel is the primary outcome. A wearable that says you slept well while you feel terrible is a wearable giving you less useful information than your own experience. A perfect sleep schedule that leaves you consistently tired is a schedule that is not working for you specifically, regardless of whether it should work theoretically. The nuyu method uses tracking to surface patterns that are too subtle to see in real time, but the ultimate evaluative criterion is always lived experience: energy, clarity, emotional stability, functional capacity across the day. The numbers serve the experience, not the other way around.

This is also why the left-right journal separates objective data from subjective analysis rather than blending them into a single interpreted record. The felt experience, written on the right side without regard for what the numbers say, is information in its own right. The journal practice honors both streams as independent sources of signal and lets the relationship between them become visible over time. When they agree, confidence is high. When they diverge, something interesting is happening, and the investigation becomes more targeted.