The Decodex is a periodic-table-style reference for body language. Every cell is a single behavior; rows are body regions head→toe; columns are stress/deception likelihood low→high. Hover any cell for full metadata. Below: the cell key, the chart's architecture, and how to use it to read deception in real time.
The Cell Key
↑ Hover or tap this example to see the full metadata panel.
- Symbol2–4 letter code
- Nameshort label of the behavior
- DRSDeception Rating Scale, top-right, 1.0–4.0
- Colorgesture type (see below)
- Gold cornermarks Decodex extensions (vs original BTE 2022)
Hover any cell for the full record: confirming, amplifying, and conflicting gestures; the microphysiological signal; cultural prevalence; sex skew; and number of variants.
Gesture Type (color)
Open — comfort, openness
Closed — pacifier, withdrawal
Aggressive — anger, dominance
Unsure — ambiguous; depends on cluster
Color is your fastest read. A face full of green/amber suggests comfort. Red appearing in a baseline-comfortable region is a flag. Slate (Unsure) cells need cluster context to interpret — never read them alone.
Architecture
Rows are body regions, head to toe: head, forehead, eyes, gaze, nose, mouth, lips, cheek/jaw/chin, face, neck, shoulders/arms, hands (gestural), hands (pacifier), torso, hips, legs/feet. The bottom two rows are voice/verbal and object interaction.
Columns within each row run left to right by Deception Rating Scale (DRS), from baseline/comfort (1.0) to strong tells (4.0). The most stressful behaviors live on the right edge of each row.
1.0 baseline
2.0 mild
3.0 high
4.0 max-tell
Timeframe (B / D / A)
For each cue we note when in the question/answer cycle to expect it:
B — Before answering (preparation, leakage)
D — During the answer (live cue)
A — After the answer (release, regret, scan)
BDA — can appear in any phase
The most diagnostic deception cues fire after an answer (Looking-for-Acceptance, Duping Delight, Cathartic Exhale). Cues that fire before an answer often signal preparation to deceive.
Reading Deception
Single cells are weak signals. Body language is read in clusters across a question/answer scenario.
Rule of thumb: tally the Deception Rating Scale value of every behavior observed during a question and its answer. Cluster total > 11 = high deception likelihood.
- Watch for any cell at DRS 4.0 — these are anchors and mark the entire cluster as high-stakes.
- A confirming gesture (per the Confirming field) and a conflicting one cancel each other out.
- Always benchmark against the person's baseline — deviations matter more than absolute behaviors.
Baseline First
Before you read any single cue, spend the first 30–60 seconds observing the person at rest. How fast do they normally blink? Where do their hands sit when nothing's happening? How loud, how fast, how relaxed are they when the topic is neutral?
That is their baseline. Every cell on the Decodex tells you what a departure from that baseline likely means — not what an absolute posture means. A still hand on someone fidgety is a red flag; a still hand on someone naturally still is nothing.
The 5 C's of Behavior Reading
Five filters every Decodex read should pass through:
- Context — the environment, the situation, what was just said. A scratched nose during allergy season means nothing.
- Clusters — behaviors don't appear alone. One cue is noise; three pointing the same direction is signal.
- Congruence — do the body, voice, and words agree? Mismatch (Fin, Nx, Easm, Dd) is itself a tell.
- Consistency — how does what you're seeing match the person's baseline pattern?
- Culture — the same gesture means different things across cultures (Esf, Pf, Cpt, Ts).
Above all five: CHANGE
The most diagnostic moment in any read is the instant a cue shifts — pace changes, posture breaks, gaze cuts away, breath catches.
Reading Clusters (Decodex)
The Decodex names three cluster patterns the BTE leaves implicit:
- Confirming cluster — multiple cells from the same Confirming list fire together. Strongest read.
- Amplifying cluster — a primary cell + amplifiers raises its effective DRS.
- Conflicting cluster — incongruent signals (face vs body, words vs face). Itself a high-confidence tell. Flagged by
Fin (Facial Incongruence), Nx (Nod Contradiction), Easm (Emotional Asymmetry), Dd (Duping Delight).
The fastest live use: build a 30-second baseline of the person at rest, then watch for departures from it. The Decodex tells you what kind of departure each behavior represents.
Variable Factors (VF)
Each cell carries a VF score from 1–4 — the number of distinct presentations the behavior takes. It's not a confidence rating; it's a count of how many different forms of the same cue you'll need to learn to recognize in the wild.
- VF 1 — single canonical form. Spot it once, you've got it. e.g., Ear Lean, Earlobe Pull.
- VF 2 — two main variants, usually mirrored or context-shifted. e.g., Head Tilt, Hair Tuck.
- VF 3 — three or more, often with strong baseline dependence. e.g., Blink Rate, Breath Rate, Happiness/Smile.
- VF 4 — highly polysemous; you need pattern memory across all forms. e.g., Eye Accessing Cues, Facial Touch, Hugging.
High-VF cues earn extra cluster discipline: the more presentations a behavior has, the easier it is to misread one of them in isolation. Pair them with their Confirming and Amplifying cells before drawing a conclusion — never read a VF 3 or VF 4 cell alone.