A Confluence Checklist for Trade Entries (with Scoring)

What is a confluence checklist for trade entries?

A confluence checklist is a fixed set of binary questions you ask before every entry, each answerable yes or no from data you can point at, scored so the decision to take or skip comes from the count rather than from how the chart feels. Confluence means independent factors agreeing — a level that exists for a mechanism you can name, flow that reacts at it, structure that permits the direction, derivatives positioning that does not contradict it, and a regime in which the pattern has room to work.

The word doing the work is independent. Three moving averages pointing the same way is one factor measured three times. A support level, absorption on the tape at that level, and open interest falling into the low are three different measurements of three different things, and only that kind of agreement reduces anything.

The checklist below is the same decomposition The Confluence Show uses on air: five blocks, two binary checks each, ten points total. It is a structure for organising a read, not a system that generates entries.

The gate: no invalidation, no score

Before any point is counted, one condition has to be met: you can state, in one sentence, what would make the idea wrong. Not a stop distance — the condition. "The level fails on a closed bar." "The absorption gets run through and price accepts beyond it." "Structure breaks in the opposite direction."

If you cannot write that sentence before entering, the setup scores zero no matter how many boxes it ticks. This is not a discipline slogan; it is a definitional point. A thesis without a falsifier is a direction, and a direction cannot be scored, reviewed or improved. Everything published on the Show carries its invalidation stated in advance for exactly this reason.

Which confluences should you look for in level, flow and structure (checks 1–6)?

Six checks across three blocks: two on the level itself, two on the flow reaction arriving at it, and two on the structure and geometry around it. Each is written so that "maybe" is not an available answer — if you find yourself arguing for a yes, it is a no.

Checks 7–10: derivatives, regime, and where checklists cheat

The last two blocks are the ones that require leaving the price chart, which is precisely why they carry information the price chart does not. Derivatives ask whether positioning agrees with the level or crowds against it; regime asks whether the tactic has room to work in the volatility state actually present. Both are read off screens the chart does not show, which is why they are the checks most often waved through.

Two structural notes. Check 4 is where most retail order flow analysis quietly cheats: marks placed on the bar that created them, using information that only existed several bars later. Separating where a mark is drawn from when it became knowable is the difference between a checklist you can audit and one that flatters itself. And check 7 is the one traders skip most often, because open interest lives on a different screen and rarely confirms the story already forming on the chart.

How do you score a trade setup, and what does each band mean?

Count the yeses across the ten checks and read the total against the bands below. Before reading the band, apply one composition rule: the score only counts if block 1 (Level) and blocks 3–4 (Flow) each contribute at least one point. Without that rule, a trader can reach seven on structure, derivatives and regime alone — a coherent macro story with no actual level and no actual reaction, which is a narrative rather than a read.

That last row is deliberate. Letting the score drive position size means one wrong high-conviction read costs several correct ones, and conviction is the exact input the checklist exists to distrust. Let the score control how often you trade, not how much.

The scoring also has an afterlife. Because every entry carries a number and a block breakdown, a month of trades becomes a dataset: which block was missing on the losers, which block you keep waving through, whether your 9s actually behave differently from your 7s. Most traders cannot answer those questions because their entries were never recorded in comparable units.

How the Confluence Engine renders these blocks live

On the Show, these five blocks are not read by eye. The Confluence Engine ingests raw Hyperliquid trades and book state for BTC, ETH, SOL and HYPE, stores the tape rather than candles, and maintains more than forty layers that map onto the blocks above — level construction, absorption and delta divergence, market structure, funding and open interest, and a volatility regime classifier.

Two implementation details matter more than the layer count. Every value carries two timestamps: the bar it is drawn on, and the moment it became knowable. Nothing is ever plotted earlier than it could have been observed, which is what makes check 4 enforceable rather than aspirational. And every published thesis ships with its invalidation condition attached, which is the gate rendered in software.

A typical live reading is a sentence of this shape: level block satisfied by an untested value area edge; flow block satisfied by aggression arriving without displacement, confirmed two bars later; structure permitting; derivatives contributing one of two because OI is cooperating but spot flow is flat; regime contributing zero because volatility has expanded past what the tactic needs. Seven of ten, composition rule met, invalidation stated. That is the whole output — a read and its falsifier, not an instruction.

What the score cannot tell you

It cannot tell you the trade will work. Ten independent factors agreeing describes the state of the auction at the moment you looked; it says nothing about a hit rate, and any number offered as one is invented. Absorption at a level can hold and turn price, or the absorber can be overwhelmed and the level can break violently — the checklist confirms the defence exists, not that it holds.

It also cannot see hidden liquidity, cannot tell you why anyone traded, and degrades on thin books where a handful of participants can produce every pattern in the table. And a checklist is only as honest as the moment you fill it in: scored after entering, it becomes a rationalisation engine. Fill it in before, or do not bother.

When a checklist is the wrong tool, and what else to use

If your edge is systematic — a rule set you can express in code and test over years of data — a discretionary checklist adds nothing a backtest cannot state more precisely. Score-then-decide is for discretionary reads at levels, where the decision is genuinely being made in the moment.

On tooling, be honest about fit. For raw book visualisation and watching liquidity move in real time, Bookmap (bookmap.com) remains the reference and is the better choice if that is your primary read. For deep footprint and delta configuration across crypto venues, Exocharts (exocharts.com) is excellent and highly configurable. For building an automated system end to end, Sierra Chart (sierrachart.com) is the more serious platform, and it is not close. Each of those is a better answer than we are for its own use case.

The Confluence Show is a different object: a live show where an AI analyst reads more than forty layers on real market data and narrates the thesis, the confluences behind it and the condition that would invalidate it, out loud, in advance. It is a way to see the checklist applied under time pressure, not a charting terminal and not a signal service.

See how the engine works, or watch the show free on a delay. — @TheConfluenceShow

Educational analysis, not financial advice.