Joseki — Directional-bias supply & demand for TradingView

The problem

What is timing?

Most analysis answers what and why. It rarely answers when — and when is the only part that costs money. If a setup reads well now, it will probably still read well in an hour, a day, a week. The story is patient. Price is not.

The usual response is to add confirmation. More indicators, more confluence, more reasons to feel certain. But every tool you add draws both directions at once, and a chart that argues both sides isn't giving you context — it's giving you permission.

A chart that shows you both sides isn't context. It's permission to change your mind at the worst possible moment.

Joseki takes the opposite approach. It removes rather than adds. You establish the bias once — an anchored mean, an anchored VWAP, or both — and from that point the chart refuses to draw anything that contradicts it.

Not greyed out. Not deprioritised. Deleted. A supply zone that forms while price is above the line never appears. One that survives until the bias flips is removed along with every signal that fired inside it.

What remains is not a recommendation. It is the reduced set of places where your thesis and the structure of the market agree — and nothing else competing for your attention while you decide.

定石 · joseki

Candlesticks were not invented on a screen.

They came from the rice exchange at Sakata, three hundred years ago, where merchants learned to read a market from the shape of its trading. The vocabulary you already use is theirs — harami is in this indicator's source code.

In Go, a joseki is a sequence known to be correct.

Not calculated at the board. Learned beforehand, then recognised when it appears. The strength isn't in seeing further than your opponent — it's in already knowing which shapes matter.

A board offers ten thousand legal moves.

Perhaps thirty are worth considering. The player who has to evaluate all ten thousand has already lost on time. Skill is the discipline of not looking at the rest.

Preparation is mostly subtraction.

This was never about seeing more. It was about refusing to entertain what you had already ruled out — before the moment arrives, when refusing is easy. That is the whole design.

Three engines,
one gate.

Each runs independently, and none of them prints without permission from the bias line.

Zones

A base candle, a displacing impulse, and a third candle that never fills the gap. Two timeframes at once. You choose wick or body for every boundary, and whether a zone dies on first touch or survives to be retested.

gap impulse must fully displace
atr impulse ≥ ATR threshold
state fresh → tested → mitigated

Reversal counts

Perfected setups and completed countdowns, with recycling and cancellation intact. But a count on its own is noise — Joseki only prints one where it lands inside a zone that survived the gate.

S perfected setup
C completed countdown
req zone confluence + bias side

Fair value gaps

Displacement gaps that must originate from a zone touch within your lookback — a bull gap needs demand behind it, a bear gap needs supply. Optional inside-bar origin, size ceiling, and a live intrabar ping before the candle closes.

origin zone touch within lookback
shape inside bar / harami
live provisional pre-alert

The gate

Bias isn't
a suggestion.

Most indicators draw a moving average and hope you look at it. Joseki treats the line as a permission system — and it keeps enforcing it after the fact.

Permit
Above the line, demand only. Below it, supply only. Nothing on the wrong side is ever drawn in the first place.
Cull
Regime changes, the chart changes. When price closes through the line, existing wrong-side zones are deleted — along with any signal that fired inside them.
Anchor
Two lines, your choice. Anchored SMA on any higher timeframe, anchored VWAP by session, N-hour grid, or one specific bar you click. Run either, or require both.

bias gate · resolved every bar

// demand needs room to rise, supply room to fall
demandAllowed = close > aSMA and close > vwap
supplyAllowed = close < aSMA and close < vwap

// and the same test runs on zones already drawn
if stale or mitigated
    box.delete(z.bx)
    deleteSignalsFor(z.id)

Zone lifecycle

You always know
what's been touched.

Fresh

Untraded

Drawn on the close of the third candle. Price has not returned since. The orders are still assumed to be resting.

Tested

First touch

Price traded back in without breaking through. The zone re-colours and the tag flips, so a retest never looks like a first entry.

Mitigated

Broken

Price closed through the far edge. The zone and everything that fired inside it come off the chart together.

Culled

Wrong side

Never broken, just no longer relevant. Bias flipped, so the zone is removed before you can talk yourself into it.

Alerts

One alert.
Every signal.

Set a single TradingView alert on Any alert() function call and point it at your webhook. Choose JSON for your own relay, or Discord to post straight to a channel with no infrastructure at all.

Live pings arrive early. A forming fair value gap fires the moment it satisfies every filter, marked gold so you know it's provisional. The close-confirmed alert is the one that counts.

discord · confirmed & provisional

C_BUY

SymbolALLOUSDT

TF3

Price0.48810

FVG_ZONE_SELL_LIVE

SymbolNZDUSD

TF240

Price0.58661

Our philosophy

Anticipating price, not confirming it.

Joseki is not a black box and it does not place orders. It is a framework for arriving at a decision point already knowing which decisions are available to you.

It works the same way across any asset, any interval, any style — because the underlying idea isn't a pattern to memorise. It's a constraint you agree to in advance and then don't get to renegotiate mid-trade.

Objective
The bias rule is arithmetic, evaluated bar by bar. It doesn't consult your position or your mood.
Universal
Crypto, FX, futures, equities. Any timeframe. The structure logic doesn't care what it's drawn on.
Yours
Every gate is a toggle. Run all of them, one of them, or none — and see exactly what each one removed.

Access

Three ways in.

Monthly

$29/ month

Full indicator. Cancel any time, free.

All three engines, both timeframes
Anchored SMA and anchored VWAP gates
JSON and Discord alerts, live and confirmed
Private Discord community
Every update while you're subscribed

Start monthly

Lifetime

$399once

Joseki, forever. No renewal.

Everything in monthly
All future versions of Joseki
Private Discord community
Priority support

Get lifetime access

Founding member

$599once

Every indicator we ever release. First 100 only.

Everything in lifetime
Every future release, included
Beta access before public launch
Founders-only Discord channel
Feature requests reviewed first

Checking availability…

Claim founding place

How access works

At checkout you'll enter your email and your TradingView username — the one on your profile, not your email. Access is added to that account automatically, usually within a minute, and appears under Invite-only scripts. If it hasn't shown up in ten minutes, message us and we'll add it by hand.

Questions

Before you buy.

To use the indicator itself, no. To send alerts to a webhook — Discord, Telegram, or your own relay — yes: webhook notifications start on TradingView's Essential tier, and you'll want two-factor authentication switched on. Confirm current tier requirements on TradingView's pricing page before you buy.

Anything TradingView charts — crypto, FX, futures, equities. The zone logic is timeframe-agnostic. The anchored VWAP needs volume data, so on symbols without it, run the anchored SMA gate alone.

Neither. It's an analysis tool that draws structure and removes what conflicts with your bias. It places no orders and gives no advice. Every decision is yours.

Yes. Create a channel webhook in Discord, set the indicator's alert format to Discord, and paste that URL into TradingView. No server, no bot token, no relay. Choose the JSON format instead if you're wiring signals into your own system.

Monthly access is removed at the end of the billing period you've paid for. Lifetime and Founding access don't expire. Either way you keep any charts you've saved — you just stop seeing the indicator on them.

A joseki is an established sequence of correct play in the game of Go — a pattern learned in advance and recognised when it appears, rather than calculated from scratch each time. It's also a nod to where candlestick analysis actually comes from: the Sakata rice exchange in eighteenth-century Japan. The vocabulary is still Japanese; harami is a filter in this indicator's source.