It looks like a lot at first. Read it in this order and thirty seconds is enough.
The nine cells at the top right are the summary of everything else. The lit cell is where the market is now.
Columns are the 15-minute chart (fast), rows are the 1-hour chart (slow). Both bullish → top-left "Both Long". Both bearish → bottom-right "Both Short". The middle cell, "Doors Decide", means neither is trending.
Cells with a dashed border are where this state historically moves next. "Most likely" is first, "2nd" is second. That is a tendency in the data, not a forecast.
The lit cell carries a glowing amber number, for example 15W-Door 64,263.
That is the first door price will meet if it keeps going the way it is going:
The same door appears in three places so you can cross-check it: inside the cell, highlighted amber in the Key Levels column, and as the bright amber horizontal line on the chart.
This door changes by itself as price and state change. It is not fixed.
The timeline at the bottom right records every state change.
It records what happened — "15m turned bullish", "1h wave entered B" — not what you should do. From it you can tell how long the current state has held and how many times it flipped today.
Flipping often means the market is chopping. A long stretch without a flip means a steadier trend. That fact alone is information.
Everything else is supporting detail: charts for shape, Six States for the breakdown, Key Levels for the other prices that matter. You do not need to watch all of it — go back to it when something puzzles you.
This is the core of the whole system. A door is a price level left behind by a past turn.
The market once reversed at that price, so that level is remembered. When price comes back to it there is often a reaction — because the people who traded there still remember it.
Doors are computed by the engine from closed bars, never drawn by hand. The board shows the nearest live door above and below on both the 15-minute and 1-hour charts.
The centre of gravity for the session — where most of the trading has actually happened.
Price above VWAP means the average participant is in profit; below means the opposite.
A medium-term direction reference. Watch its slope and whether price sits above or below it — it is not an entry signal.
This follows the Taiwanese convention, which is the reverse of the Western one. Everything on the board is consistent with it:
| Colour | Meaning |
|---|---|
| Red | Bullish (bar closed up) |
| Green | Bearish (bar closed down) |
| Amber | Divergent — the two timeframes disagree, or one is ranging |
| White | Doors Decide — both timeframes ranging, direction is up to the doors |
| Blue-violet | Doors |
The "Green up" button at the bottom right switches to the Western convention.
These mark the wave cycle, placed on the bar where the wave actually changed.
It reads the direction and strength of momentum — which side the larger context sits on, and whether the current leg is pushing or resting.
| Code | How it is decided | In plain words |
|---|---|---|
| 1 | Context bullish, momentum just turned up | Impulse up |
| 2 | Context still bullish, momentum weakening | Pullback (a rest inside an advance) |
| 3 | Rest over, momentum turns up again | Second push |
| A | Context bearish, momentum just turned down | Impulse down |
| B | Context still bearish, momentum strengthening | Bounce (a rebound inside a decline) |
| C | Bounce over, momentum turns down again | Second drop |
B is a bounce, not a downtrend push. This is the single most common misreading — B appears while the larger context is still bearish, yet the bar itself is rising.
The cycle runs 1 → 2 → 3 → A → B → C → 1 and never skips. Crossing from the bullish half to the bearish half requires the larger context itself to flip — that is also where the cycle restarts.
Each timeframe is reduced to one of three states:
| State | Meaning |
|---|---|
| Long | Context bullish and momentum building |
| Short | Context bearish and momentum building |
| Flat | Momentum fading — the move is losing force regardless of direction |
The two corners are the clean states. Everything in between is some form of disagreement — which is itself worth knowing, because disagreement is when most mistakes get made.
Six independent readings, each pointing bullish (red), bearish (green) or neutral (amber).
| Reading | What it measures |
|---|---|
| 1h wave | Which of the six wave positions the slow timeframe is in |
| Momentum | Whether the driving force is building or fading |
| Last bar | Whether the most recent bar closed up or down |
| Between doors | Where price sits between the door below and the door above, as a percentage |
| Volatility | Current swing width relative to its recent norm — expanding or contracting |
| Swing structure | Whether the recent highs and lows are stepping up, stepping down, or mixed |
Underneath, a count: "4 / 6 bullish" means four of the six point the same way.
The same six-state count drawn as two bars, plus the range position — where price sits between the session low and the session high, as a percentage. 0% is glued to the low, 100% to the high.
The volume ratio compares recent average volume against the session average. Above 1 means it is busier than usual right now.
The prices worth remembering today, in three groups:
| Field | Meaning |
|---|---|
| High / Low | The extremes of the previous and current sessions |
| Open | The first traded price of the session |
| VWAP | The session's centre of gravity — the centre of gravity for the session |
| 15M / 15W doors | Doors from the 15-minute chart. Faster to react, good for the shorter rhythm |
| 60M / 60W doors | Doors from the 1-hour chart. Heavier, not easily moved by noise |
That row is the door the grid is currently watching. The small tag after it is the distance:
▲93 = 93 above the current price▼54 = 54 belowThe bright amber line on the chart is the same door. Three places, one fact.
An M-door is a level left by a past top. A W-door is a level left by a past bottom.
Important: the name has nothing to do with whether it is currently above or below price. A W-door that gets broken downward ends up sitting above price — it is still a W-door, it has simply changed sides.
That is why the board never says "upper door" or "lower door". It always writes the full name — 15W-Door, 60M-Door — because on a real chart you will meet two doors, one above and one below, that would otherwise carry the same label.
| Field | Meaning |
|---|---|
| 5m / 15m / 1h closes | Time until that bar closes. Anything from an unclosed bar is provisional. |
| Nearest door | Of all live doors, the one closest to the current price |
| Structural Position | Distance from price to each reference level, signed |
| Signal Log | Only real state changes are recorded — not every tick |
A new panel at the bottom prints one candle every second — red for up, green for down, the same colour convention as the main chart.
The reason it exists is simple: the livestream sends one frame per second. If everything on screen sits still, it looks broken. This panel keeps growing new bars, so one glance tells you the board is alive.
Nobody traded during those seconds. For seconds with no trade we insert a flat candle (open = close) so the timeline stays continuous.
The Taiwan futures night session looks like this a lot — in one measured stretch, 75% of seconds had no trade at all. The first minutes after the day-session open are the opposite: several trades a second, and the panel moves visibly.
So a flat row is not a bug. The market really is standing still.
There is a speaker button next to the "1-second candles" heading. Click it to enable sound (browsers require one click before any page may play audio). The livestream carries sound too — nothing to click, just turn your volume up.
It is designed so that you do not have to watch the screen. Leave it on while you do something else. It is nearly silent while the market ranges; when you hear a bright bell or a low thud, that is when to look up — a gate has just opened or closed.
| What you hear | What it means |
|---|---|
| A warm, slightly rising note | Red candle (close above open) |
| A low, contained note | Green candle (close below open) |
| Louder | That candle has a bigger body — measured against its own recent typical range, not a fixed number of points |
| Timbre slowly brightening / darkening | A run of same-coloured candles is building (the third red sounds brighter than the first) |
| A bright bell | Gate opened — the candle closed above the gate |
| A low thud | Gate closed — the candle closed below the gate |
| A barely audible drone | The bed. Its only job is to tell you the picture has not frozen |
| Nothing at all | A doji (close = open) makes no sound. Ranging markets are quiet, and that is deliberate |
The bottom-left corner cycles through messages, one at a time, changing every 5 seconds. It exists so you do not have to read the whole board. If the 3×3 grid means nothing to you yet, the ticker says what is happening in plain words.
| You see | What it tells you | How to use it |
|---|---|---|
| Now Both Long · both aligned | The structural state of the instrument currently on screen | If you just arrived, this one line is enough — it tells you which way things lean |
| Nearest gate 60W gate 44,586 · 120 below | The closest gate to the current price | The most useful one. Knowing where the gate is tells you which price to watch, and what to pay attention to when price gets there |
| Forming 15m M gate 45,382 · price reached, waiting for the wave | A gate is forming but has not become one yet | See where the next gate may appear instead of noticing it only after it exists |
| Watchlist ZECUSDT · W S(2) · D W gate 612 open · 0.4% away | An instrument the scanner picked out as matching the configured conditions | If something interests you, note the symbol and go research it yourself (details below) |
| How to read What is a gate? … | Rotating basics | Anyone arriving mid-stream can follow along without reading this guide first |
These come from a scanner that sweeps each market on a schedule. Read each entry in four parts:
ZECUSDT W S(2) D W gate 612 open 0.4% away
| ZECUSDT | The symbol. Taiwan stocks also show their Chinese name |
| W S(2) | Which timeframe, which engine, which wave — here: weekly, S engine, wave 2 |
| D W gate 612 | There is a daily W gate nearby, at 612 |
| open / closed / potential | Whether that gate is currently open, closed, or not yet formed |
| 0.4% away | How far price is from it. Closer means it resolves sooner; far away means park it for now |
Three practical uses:
Taiwan stocks and ETFs are scanned once per day after the close. So the list you see in the evening is exactly the set of instruments whose structure reached an interesting position as of today's close.
Note the symbols and open the charts yourself — what you are looking for is whether that gate gets opened, and whether price holds above it, not a list to buy from.
Crypto is scanned every 30 minutes, around the clock. The ticker keeps moving even when the Taiwan market is closed.
When a dozen symbols show up, look at the close ones first — those resolve within days. Anything 5% or more away can wait.
The stream sends one frame per second. Continuously scrolling text sampled at that rate stutters, and compresses badly. Showing one message at a time for 5 seconds reads far better — especially on a phone.
Below the key levels there is now a "Forming gates · not gates yet" list. It shows positions that may become gates but have not yet, in two states:
| Label | Meaning |
|---|---|
| ⋯ Awaiting price | The structure is forming, but price has not reached that level |
| ⏳ Reached · waiting | Price got there, but the wave is not finished. Once the wave completes, it has a chance of becoming a gate |
The point is to know where a gate might appear in advance, rather than seeing it for the first time after it exists. But remember: a forming gate is not a gate, and it may never become one.
The signal log used to record only "price approaching a gate". It now records:
Gate opened · close above 15m W gate 45,500
Gate closed · close below 60m M gate 46,450
Judgement always uses the close; a wick poking through does not count, and only completed candles are considered. Both M gates and W gates have open and close events.
Price is effectively real time.
But waves and doors are only judged on closed bars. So you will sometimes see price moving while the state does not change. That is correct behaviour, not a freeze.
Why? Because judging from an unclosed bar makes the momentum flip back and forth around zero, the state flickers, and the signal log fills up with noise. Half a bar late beats a false signal.
Not necessarily. We tested this. Over the sample we measured, doors were not meaningfully better than random price levels at stopping price.
So treat a door as a level where a reaction tends to occur, not as a wall. The board marks it to tell you what price will run into next — not to predict whether it will get through.
A door is destroyed once price has crossed it enough times. The reasoning: a level that price keeps cutting through in both directions is a level the market has stopped respecting.
In honesty: we tested fast versus slow destruction and found no measurable difference in outcome. The rule is kept because it is part of the original method, not because we can prove it helps.
It was not designed for that. There are no entries, no stops and no win rates on it — all of that was deliberately left out.
What it does is make the structure of the session legible. What you do with that is your call.
| Not here | Why |
|---|---|
| Entry price / stop / target | That is trading advice, not a description of structure |
| Win rate / confidence score | An unvalidated number that looks like a probability is more dangerous than no number |
| "Buy", "sell", "go with the trend" | Same reason. This board says what is, never what you should do |
This is not caution for its own sake — it is a deliberate position: the goal is to help you read the session, not to decide for you.