Intention
What I planned to do.
03 / 2026Data Visualization / Behavioral Research
Second
World
A synchronized behavioral portrait comparing what I intended to do, what my digital avatar actually did, and how my body unconsciously responded.
- Duration
- 2,996 recorded seconds → 60-second animation
- Media
- Data audit · p5.js · interactive animation
- Structure
- Six synchronized temporal channels
LIVE BEHAVIORAL PORTRAIT
Play, pause, scrub, or inspect 15 / 38 / 60 seconds.
One experience was recorded through three different selves.
The visualization does not score the player or explain behavior psychologically. It compares three temporal accounts of one gaming session: the route I intended to follow, the actions performed by my digital avatar, and the changing posture of my physical body. Their agreements and mismatches accumulate into one behavioral trace.
Digital self
What the avatar actually did.
Physical self
How the body responded.
01 / SIX SYNCHRONIZED CHANNELS
The source was not one table, but six clocks describing the same session.
The digital and physical workbooks were recorded independently. Physical 00:04 and digital 00:18 are aligned as the same real-world moment, then every state and event is reconstructed inside its own temporal channel.
QUEST
4 intervalsChapter markers carried to timed digital rows.
INTENTION
14 intervalsPlanned, detour, and an explicit initial unknown state.
DIGITAL ACTION
117 intervalsSeven avatar behaviors sustained until the next action change.
PHYSICAL POSTURE
99 intervalsSix body states parsed independently from gesture timestamps.
HAND GESTURE
16 eventsInstantaneous physical events with no invented duration.
TRIGGER
3 eventsCollect and explore remain point events on the actual route.
02 / FROM LOG TO AUDITABLE DATA
Duration had to be derived without inventing relationships between rows.
The original sheets mix sustained states with instantaneous events, and every duration cell is blank. A gesture timestamp cannot be paired with a posture merely because they share a spreadsheet row. Each channel is therefore parsed, sorted, aligned, and validated independently before visualization.
- 01
Raw workbooks
Mixed temporal grains and inconsistent timestamp formats.
- 02
Data audit
Independent channels, endpoint assumptions, and interpretation limits.
- 03
Derived intervals
Aligned seconds, duration, source rows, and validation status.
- 04
Validated JSON
Six reconciled p5-friendly collections with unchanged sources.
- 05
Animation
A proportional 2,996-to-60-second transformation.
03 / VISUAL GRAMMAR
Every visual mark answers to one recorded variable.
The system behaves like an instrument rather than a literal game map. Time controls horizontal position; category, duration, posture, and event type control the visual properties around that shared route.
Planned intention
Warm-ivory dashed lineThe uninterrupted expected route.
Actual behavior
Solid cool-blue lineLeaves and reconnects during six recorded detours.
Digital action
Cool categorical dotsDuration sets dot count; timestamp fixes route position.
Physical posture
Lavender ripple fieldPosture changes scale, eccentricity, opacity, and pulse rate.
Hand gesture
Pink temporary ringsA short ripple leaves one permanent notch in the route.
Trigger + quest
Green nodes + bracketsPoint events and duration-proportional journey sections.
04 / ONE SESSION, ACCUMULATING
The route grows; its earlier marks remain as history.
Previously emitted dots keep deterministic positions while later events accumulate. At 59.02 seconds the digital route stops; the body continues through the final physical aftertrace.
Quest 1’s dominant detour establishes the difference between plan and behavior.
Multiple quest sections and action rhythms accumulate without resetting the frame.
The completed trace remains alive through an autonomous physical ripple.
The strongest recorded body state expands and brightens the physical-energy field without reclassifying the digital action.
05 / QUEST PORTRAITS
Chronology becomes comparable without becoming interchangeable.
The complete session answers when physical and digital states overlap. Three completed quest portraits compare duration, detour share, action rhythm, and bodily response. Hovering a portrait opens a fixed 120-second detail window with its own moving playhead.
30.0% detour
0% detour · original Q2 + Q3
12.1% detour · original Q4
06 / DATA INTEGRITY
Uncertainty stays visible inside the finished portrait.
The project preserves the limits of the recording instead of converting them into confident-looking graphics. Every derived record retains its source row, raw timestamp, and assumption flags.
An instrument for reading one recorded experience—not a dashboard and not a literal game map.
The final animation makes three temporal selves visible without collapsing them into one measure of intensity. Intention supplies a reference, digital action builds behavioral matter, and the physical body changes pressure and depth. Their mismatches are not errors to correct; they are the subject of the work.