Real-Time Wyckoff Orderflow Analysis — what institutions are really building
4 phases · 6 Wyckoff states · Live POC/VAH/VAL · Continuity Score
You see the price. Not the build-up behind it.
Classic charts show the price. But not: whether accumulation or distribution is currently happening. Whether the London session closed bullish or bearish. Whether institutions are building long or short. This market structure decides whether your setup has a tailwind or a headwind.
You trade against the session direction
The London session closed with +3,000 contracts of delta — clear long build-up. You enter short at the NY open. Institutions push the price down once, their cluster gets triggered, and the pendulum swings back. Stopped out.
You misread accumulation as weakness
Price moves sideways. Volume is high. Classic reading: “range — wait”. But if the POC rises and delta cumulates positive, this is accumulation — a preparation pattern for the next leg up.
You have no real-time phase structure
Manual tracking of large orders per session is impossible — with 5–15 trades per hour you lose the overview. Without an automated aggregator, you don't know the cumulative bias of the running phase. You trade blind.
Three metrics every order-flow trader needs live in the HUD
Phase-Aware aggregates in real time what traders otherwise reconstruct only post-mortem from journals: cumulative delta per phase, Wyckoff structure, and value-area drift.
4 phases with large-order thresholds
London, NY Open, Lunch, Afternoon — each phase has its own threshold above which an order is classified as “institutional”. Smaller orders flow into the noise floor, larger ones into the phase delta.
- LONDON ≥ 120 contracts
- NY OPEN ≥ 200 contracts
- LUNCH ≥ 180 contracts
- AFTERNOON ≥ 150 contracts
- NY-time detection via getNYMinutes()
6 Wyckoff classifications
Markup, Markdown, Accumulation, Distribution, plus two divergence states. Classification runs continuously from delta trend, POC drift, and range width — recomputed every 5 seconds.
- MARKUP · stable bull structure
- MARKDOWN · stable bear structure
- ACCUMULATION · long build-up in range
- DISTRIBUTION · short build-up in range
- DIVERGENCE_BULL / DIVERGENCE_BEAR
Continuity Score (weighted)
A single score between −1.0 and +1.0 shows the day bias. Weighting: 30 % London delta, 50 % NY Open bias, 20 % large-player aggression.
- 30% Δ London ratio
- 50% NY-Open bias
- 20% institutional aggression
- Live update every 5 seconds
- Persistent via localStorage v2_b22
LONDON
Europe-driven. Sets bias for NY Open.
NY OPEN
Most volatile block. Highest weight in score.
LUNCH
Reduced volume. High whipsaw risk.
AFTERNOON
Trend continuity or clear reversals.
How to read market structure in 5 seconds
Three places in the HUD tell you everything about the running phase. Example: 2026-05-21 NY Open, 09:42 ET — a classic MARKUP setup with a confluent Continuity Score.
When which state — and what it means
Each Wyckoff state combines three signals: direction of cumulative delta, drift of the Point of Control, range width. Classification runs every 5 seconds — you see transitions live, before they appear in the price.
Clear uptrend, long bias preferred.
Long build-up in range. Breakout setup.
Clear downtrend, short bias preferred.
Short build-up in range. Breakdown setup.
Price falls, delta rises — reversal ahead.
Price rises, delta falls — topping out.
What the engine actually does
Directly verifiable in ofa-phase-aware.js — no marketing approximation.
Common questions about Wyckoff Orderflow Analysis
The key points about the Phase-Aware engine in short form.
How does Phase-Aware differ from classic Wyckoff tools?
Classic Wyckoff analysis is post-hoc: traders draw the four phases (accumulation, markup, distribution, markdown) into the chart after the trading day. That works for learning, but doesn't help in the trade moment.
Phase-Aware re-classifies the running market structure every 5 seconds from three objective signals: delta trend, POC drift, range width. You see the phase in the HUD while it happens — not in retrospect.
Why do I need cumulative delta per session phase?
Daily delta is too coarse. If London is −1,500 and NY Open is +4,500, daily delta is +3,000 — but London traders lost and NY traders won. As an NY trader, you need the NY-Open phase, not the daily aggregate.
Phase-Aware separates this automatically. You see per phase: cumulative delta, large-order share, phase bias. If NY Open starts with +5,000, you know aggressive buyers dominate. You trade with the flow, not against it.
What does the Continuity Score mean concretely?
A single number between −1.0 (STRONG_BEAR) and +1.0 (STRONG_BULL). Calculation: 30 % London delta ratio, 50 % NY Open bias, 20 % large-player aggression.
Values above +0.6 or below −0.6 signal clear directional phases. Values between −0.3 and +0.3 are range days — where mean-reversion setups outperform trend trades.
Does this work in the Trial plan?
Yes. Phase-Aware is included in all three plans (Trial, Starter, Pro). The Trial gets the full aggregator with live phase detection and Wyckoff classification.
What Pro additionally includes: GPU Heatmap, Cortex AI Strategy with OEM engine. Phase-Aware is the foundation — in the Trial too.
What happens on app restart mid-session?
Phase-Aware persists its complete state every 15 seconds to localStorage (key v2_b22). Validation is against the trading-day boundary (18:00 ET) — after a day change a new day starts, not a restore.
On restart within the same trading day, phase counters, Continuity Score, and Volume Profile are reconstructed from localStorage. You lose no aggregation history through an app crash.
How does the Historical Replay load without UI freeze?
When you launch the app or change days, tens of thousands of ticks may need reconstruction. Phase-Aware uses chunked processing with requestIdleCallback: 500 ticks per batch, then yields control back to browser rendering.
Result: the UI stays 60-FPS-smooth while the aggregator rebuilds session history in the background. Full session backfill (8h) typically under 6 seconds.