Habit Tracker

January + February 2025 · 59 days · temporal correlations · confidence-rated · Updated 4 March

59
Days tracked
6.3h
Avg sleep
64
Avg pickups
1:06
Avg screen
+0.9h
Bed@9pm → sleep
89%
Anchored avg

Month Overview

🗓 January

Habits/day4.7
Avg sleep6.3h
Avg pickups68
Avg screentime1h 21m
ContextTravel + hospital

🗓 February

Habits/day5.2 ▲ +0.5
Avg sleep6.4h ≈ same
Avg pickups60 ▼ 8 fewer
Avg screentime50m ▼ 31m less
ContextNormal life + GFO start

January Context

January was not a normal month. Only 12 of 31 days were "normal life." Raw rates aren't comparable to February.

☀️ SummerDays 1–10
Sleep: 6.7h
✈️ TravelDays 11–22 · Phuket + Singapore
Sleep: 6.5h · Days 11 & 22 = planes
🏥 HospitalDays 23–29 · Hudson RSV + ARG1-D
Sleep: 5.4h
🔄30–31
Days 11 & 22 were travel days on a plane — rare exceptions to everything. Days 23–29, Hudson was hospitalised with RSV complicated by his ARG1-D — a stressful week sleeping and living in a hospital room (avg 5.4h sleep, lowest 3.0h). Data from these periods is kept but contextualised.

Jan vs Feb — Common Habits

Habit Jan Feb Change
📈 Improving
Draw +27pts 19% → 46%
Calm vs Irritable +15pts 77% → 93%
Read +12pts 77% → 89%
Good Food +11pts 39% → 50%
Walk +25pts 7% → 32%
Workout +6pts 90% → 96%
Pickups ▼8 68 → 60
Screentime ▼31m 1h21m → 50m
📉 Declining
Bed@9pm −14pts 39% → 25%
This is the only habit that got worse — and it's the single biggest lever for sleep (+0.9h per hit). Needs attention in March.
➡️ Stable
Journal −1pt 87% → 86%
Social Media −2pts 97% → 95%
Atomic +3pts 26% → 29%

The Structure Story — 59 Days Strong

Your habits split cleanly by whether they have a time, place, and cue. This held across summer, travel, hospital, and normal life. It's not a one-month pattern — it's structural.

Anchored — Morning (5:30am)
Wake → workout → shower → groom → water/protein → tracker → journal → kids at 6:50. Workout 93% · Journal 86% across 59 days. Workout survived travel, hospital, everything — only missed 4 days total. These are identity-level habits.
Workout 93% Journal 86%
HIGH — 59 days, survived all contexts
Anchored — Bedtime + Environment
Read 83% across both months. Social Media 96% — essentially automatic, solved through app deletion. These don't cost willpower. Read dipped to 64% during Phuket travel but bounced right back.
Social Media 96% Read 83%
HIGH — 59 days, structural
Unanchored — No Time, No Cue
Draw 33% · Bed@9pm 32% · Walk 22% · Atomic 27%. All lack a when/where/trigger. They happen when you "find time." The gap between anchored and unanchored is 56 percentage points. Same person, same willpower — structure is the only variable.
Draw 33% Bed@9pm 32% Atomic 27% Walk 22%
HIGH — 59 days, consistent across all phases
Feb-Only Habits
Gratitude 71% — part of the bedtime routine (bed → gratitude → read). Dips when getting to bed late and skipping straight to reading. Anchored to bedtime, so it's moderately strong.

AOK (Act of Kindness) 30% — unanchored, no time or cue. Follows the same pattern as other unanchored habits. Started tracking Feb 8.

Chess 90% (Jan only) — tracked in January, dropped in February. Was a holiday/travel activity.
Gratitude 71% AOK 30% Chess 90% (Jan)
The Big Improvers ▲ Jan→Feb
Draw: 19% → 46% (+27pts, biggest jump). Calm vs Irritable: 77% → 93% (+15pts). Read: 77% → 89% (+12pts). Good Food: 39% → 50% (+11pts). Most of the improvement came from leaving the chaos of January behind — but the momentum is real.
MEDIUM — improvement is real, but Jan was unusually disrupted

The Sleep Chain — 59 Nights

With both months of sleep data, the Bed@9pm → Sleep → Everything cascade is confirmed across different life contexts.

Bed@9pm → +0.9h Sleep (59 nights) KEY LEVER
19 nights in bed by 9pm → 6.9h avg sleep. 40 nights not → 6.0h. That's nearly an hour more every time. Combined correlation r=0.39 (r=0.57 for Feb alone — Jan's hospital/travel adds noise but the effect holds).
r=0.39 combined · r=0.57 Feb · r=0.32 Jan · n=59
HIGH — 59 nights, clear mechanism, consistent both months
Sleep → Next Day Effects (Feb)
After 7+ hours: 7.7 habits next day. After <7h: 6.8. After <6h: 6.3. Sleep predicts walking (r=+0.35), journaling (r=+0.26), and drawing (r=+0.24). Less sleep = reach for phone more (r=−0.30).
Sleep → Walk r=+0.35 · Sleep → Journal r=+0.26 · Sleep → Pickups r=−0.30
MEDIUM — 27 day-pairs, moderate signals
Hospital Week: Sleep Collapsed
Days 23–29 averaged 5.4h sleep (lowest: 3.0h on day 25, 3.5h on day 23). This isn't a habit failure — it's a parent sleeping in a hospital room with a very sick child. These 7 nights drag January's average down by ~0.4h. On normal-life days across both months, average sleep is 6.5h.
Hospital avg: 5.4h · Normal-life avg: 6.5h · n=40 normal days
HIGH — context, not behaviour
Bed@9pm is DECLINING 39% → 25%
39% Jan → 25% Feb — the only common habit that got worse. Yet it's the single biggest lever for sleep. The blocker isn't willpower — it's schedule (0% on Wed-Sat in Feb). Even a "Bed by 10pm" target on social nights might capture some benefit.
Jan 39% → Feb 25% · ΔSleep = +0.9h per hit
HIGH — declining trend + high-impact lever

Day of Week Patterns (59 Days)

Monday & Sunday are the strongest days
Both average 5.8 habits/day. Thursday is the weakest at 4.3 — the midweek dip is consistent across both months. Wednesday has the highest screentime (1h32m) and most pickups (71).
Best: Mon/Sun 5.8 · Worst: Thu 4.3 · Highest screen: Wed 1h32m
MEDIUM — 8-9 data points per day
Friday: Worst Day for Sleep (Feb)
5.9h average — worst of any day. Fridays are socially loaded (GFO with Dad, dinners out). Saturday (6.0h) close behind. Tuesday is best (7.0h). Combined Jan+Feb sleep by day will sharpen this picture. MEDIUM — Feb only, 4 data points per day

January Phase Breakdown

Phone Usage — Battle Won

Phone Usage ▼ Improving
Clear improvement from Jan to Feb. January was inflated by hospital days (136 pickups, 5h54m on day 28 — when your kid is sick, you're on your phone constantly). Even excluding those, Feb is genuinely better. Social media avoidance: 96% across 59 days — this is solved through environment design, not willpower.
Jan: 68 pickups, 1h21m · Feb: 60 pickups, 50m · Social media: 96%
HIGH — 59 days, environment-based solution working
Diminishing Returns Zone
60 pickups, 50m screen, 96% social media avoidance. Further phone optimisation has diminishing returns. The bigger lever is sleep — the gap between Bed@9pm nights (6.9h) and non-Bed@9pm nights (6.0h) is worth more than shaving 10 minutes off screentime. HIGH — clear from the data

Five Things for March

1. Protect Bed@9pm on controllable nights
Mon/Tue/Sun are realistic targets. Even going from 25% → 50% on these three days would add ~3 hours of sleep per week. Don't fight Wed-Sat — set a softer "Bed by 10:30" target instead. Better yet, track actual bedtime so you can see whether "missed 9pm" means 9:20 or midnight. HIGH — structural, high-impact lever
2. Anchor Drawing to a specific time
It jumped 19% → 46% which is great momentum. But without a cue, it'll drift back down. Try: "After kids bedtime routine, draw for 10 minutes before anything else." The data shows unanchored habits average 33% — anchoring could double it. MEDIUM — logical from the structural finding
3. Don't touch Workout / Journal / Social Media
These are working. 90%+ habits don't need intervention — they need protection. Don't add complexity to things that are already automatic. Workout survived travel, hospital, and everything else January threw at it. Leave it alone. HIGH — 59 days of evidence
4. Decide what to do with Atomic
27% across two months. It's not happening organically. Either pair it with an existing habit ("read 1 page of Atomic after journal") or accept it's not a daily habit and make it weekly. A habit at 27% for 59 days isn't going to suddenly click without a structural change. MEDIUM — consistent low performance
5. Keep tracking sleep — it's the connective tissue
With 59 nights of sleep data, we can see the Bed@9pm → Sleep → Everything cascade clearly. More data will sharpen the day-of-week patterns and let us separate normal-life from crisis periods. March is where the real experiment starts — the baseline is set. HIGH — measurement compounds over time

The Keystone Insight

Your system works. Your schedule fights it.
Anchored habits: 89% average. Unanchored: 33%. Same person, same willpower, same 59 days — including travel, hospital, and normal life. The only variable is structure. The four unanchored habits (Draw, Walk, Atomic, Bed@9pm) don't need more motivation — they need what Workout already has: a specific time, in a specific place, with a specific trigger.

This held through summer holidays, Phuket, a hospital crisis, and settling into a new work routine. It's not a one-month fluke. It's who you are. The structure is the variable, not the willpower.
📐 Methodology: 59 days · Pearson r for correlations · Temporal ordering: Sleep(night N) → habits(day N+1), bedtime habits → same night sleep · Confidence: HIGH (structural, 59 data points, cross-context), MEDIUM (moderate signal, 28+ points), LOW (suggestive only, weak correlation or small sample)