That decision — are you really making it as your usual self?
Exactly when the decision scares you most. 1–3 minutes, PC only — your drift from baseline, made visible.
AXIOM never tells you whether to trade or play. It only makes your drift from "usual you" visible.
Click when the signal appears. Your reaction is compared with the adult population average and decomposed into speed, scatter, and attention lapses.
Then comes the real question — "compared with your usual" — which needs history, and is exactly what AXIOM does. Feel the difference before reading about it.
Several psychology and cognition tools exist for traders. AXIOM's design philosophy differs.
※ The table below summarizes publicly available product information as of March 2026. Check each vendor's site for current specs.
AXIOM doesn't instruct. When drift shows — check the numbers, size down, rest, or keep going. The choice is always yours. It raises the signal; the discipline stays yours.
What corrupts judgment is the small state change you can't feel. Four familiar moments, and AXIOM's answer to each.
"The bigger the stakes, the more you try to push through on willpower."
Willpower can't tell you what state you're in. AXIOM never gives instructions — it focuses on making "your current state" visible from reaction-time data.
"'I feel sharp today' — that feeling fails you most when you're tired."
Reaction time is answered by the body. Disturbances too small to notice get recorded as millisecond-level variance — data where self-awareness can't reach.
"What good is 'faster than average'? If your usual is 250ms, today's 300ms is the change worth seeing."
The judgment centers on drift from your usual self. A personal baseline built over 5 sessions is the sound reference for cognition, where everyone's pattern differs.
"You usually notice the crack after the loss."
1–3 minutes, your existing PC, done. Before the trade, before the session, before the big meeting — designed to fit right before the decision. No extra hardware.
AXIOM descends from the PVT (Psychomotor Vigilance Test, Dinges & Powell 1985–), rebuilt beyond sleep research into something usable right before decisions.
You don't need to score well. Recording today's state as it is — that's AXIOM's job.
Click when the light appears — that's all. 20–50 trials collect your reaction-time data. The adaptive session (SBPM) detects when precision is reached and cuts unnecessary trials. Device quality is verified before every session.
Ex-Gaussian fitting and Bayesian MAP estimation compare the session against your personal baseline. Online change-point detection (BOCD) tracks natural baseline drift, and a cognitive-state filter learns your personal patterns. No mood self-reports required.
Check today's condition on the dashboard. Deviations from your baseline are detected in real time, and a re-test is suggested automatically when needed. What you do with the information stays your decision.
Curious what step 1 feels like? — Try it in your browser now →
This is how your measurement comes out — from the free Core tier's "today's readout" to Standard+'s "comparison with your usual". All real app screens (values illustrative). Shown here in Japanese — the app itself ships with a full English UI (auto-detected from your system language, switchable in Settings). Labels are translated in the captions.
※ Screens are under development. All values and dates are illustrative, not real data.
Reaction-time data converted into indicators — reference material for checking your condition.
Connection types (USB / Bluetooth / remote desktop etc.) are auto-detected and classified into device tiers (Tier 1–3 + Quarantined). Jitter variance < 1.0µs² is verified before any measurement starts.
Decomposed into three parameters: μ (mean speed), σ (stability), τ (attention-lapse component). Not just an average — the internal structure of your reaction-time distribution, made visible.
Five sessions build a reference range that is yours alone. From then on, daily deviations are detected "versus yesterday's you". The standard of judgment is your own measurement history.
Monitors changes in reaction-time fluctuation patterns (rising attention-lapse component τ / growing instability σ). When the Z-score crosses your threshold, an alert is shown. Sensitivity is user-adjustable — you control the balance between false alarms and misses.
※ AXIOM gives no instructions — resting, continuing, or quitting is always your call.
Sequential Bayesian Precision Monitoring detects when statistical precision is reached. Four modes — Speed (20–30 trials), Balanced (20–40), Precision (25–50), Fixed (50) — cut unnecessary trials automatically.
Tracks change across time, not just single sessions — seasonality and drift in your performance, made visible. Baselines recalculate per 90-day season.
AXIOM is local-first. Measurements, baselines, and P&L records are stored in an on-device SQLite database by default, and the app performs no automatic outbound transmission.
Instead of flashy effect claims, AXIOM publishes the sources of its methods and the limits still under validation in the same place. The measurement method (PVT-based) and statistical models (Ex-Gaussian / BOCD) rest on peer-reviewed cognitive psychology.
※ The papers below are the academic basis for the measurement and statistical methods. AXIOM's own tilt-sign detection accuracy and personal-baseline stability are being validated progressively through beta feedback — we have not yet published peer-reviewed validation of AXIOM itself.
AXIOM's measurement foundation adopts the Psychomotor Vigilance Test (PVT) methodology, with Ex-Gaussian three-parameter decomposition for analysis.
The tilt AXIOM monitors is a reference concept: fluctuations in attentional control associated with emotional stress and losses.
AXIOM's measurement reliability rests not only on algorithms but on verification of the execution environment.
A resident app that measures input devices owes you explicit boundaries. Details in the privacy policy (Japanese).
What's measured is reaction time only — the delay between stimulus and click. Typed characters, URLs, passwords, trading strategies, or any other text content are never handled by the app. Only reaction timing is read, via standard Windows input-event APIs.
Behavioral data ≠ input contentReaction-time data, baselines, and P&L records are never sent to any cloud automatically, under any setting. They leave your PC only when you explicitly export them. Separately, only if you opt in via Settings, feedback / survey responses and crash reports are sent encrypted (none of which contain your measurement data). The app never phones home on launch — update checks are manual, version-query only.
Default = stays localThe operator is "PRO ORDER", a sole proprietor in Japan. Being small means feedback reaches the product directly. Should the service ever wind down, advance notice will be given per the terms, with time to preserve your data via export. Since storage is local, your data remains on your PC regardless.
Sole proprietor · your data stays with youAXIOM is currently in a closed beta for Windows. The app UI supports English (auto-detected from your system); beta documentation and support are Japanese-first for now. Leave an email to hear about wider availability — that's the only thing we'll send.