How the AI scores confluence out of 100

What counts as a confluence

A confluence is an independent reason pointing the same way. Independent is the operative word: two names for the same observation count once, not twice.

The readings that can agree are the higher timeframe bias, the market structure, a completed liquidity sweep, the quality of the zone being entered on, what the indicators say, and the precision of the entry against the optimal zone. Each is rated separately, and the score is the sum of what genuinely agrees.

The grade, and the number of signals behind it

• A+ requires four or more confluences, a pool of orders already swept, and an entry inside the optimal zone.
• A requires three confluences and a clear higher timeframe bias.
• B requires two confluences and an acceptable risk to reward.
• C is one or two weak signals.
• No Setup is what conflicting signals or insufficient data produce, and the analysis has to say why.

The grade is not a mood. It is a count, with conditions attached.

What each grade demands in risk to reward

A grade is not awarded on agreement alone. The plan has to be worth taking.

• B grade requires at least two to one.
• A grade requires at least three to one.
• A+ grade requires four to one or better.

This is why a setup with five agreeing signals and a poor ratio does not become A+. The ratio is computed from the plan itself, from the entry, the stop and the targets, not asserted.

Why one signal is an opinion

Any single reading can be right by accident. A chart that is bullish on one indicator, in a zone that matters, above a level that held, heading for a target that has not been touched, is a different proposition from a chart that is only bullish on one indicator.

The score exists to make that difference a number instead of a feeling, and to make it comparable between two charts you would otherwise judge by how good the story sounded.

What it refuses to count twice

The most common way to inflate a confluence score is to count the same evidence under two names. An equilibrium at 50 per cent and a Fibonacci retracement at 0.5 are one line, not two. A zone and the level at its edge are usually one observation.

The factors are therefore rated one by one and named, so you can see what the score is made of rather than trusting the total.

Questions people ask

What is a good confluence score?

Above 65 the analysis expects three or more agreeing signals and a clear higher timeframe bias. Between 45 and 64 a setup exists but one or two signals contradict it. Below 25 the grade becomes No Setup.

Why is the grade sometimes low when the score is high?

Because a grade also requires a risk to reward. B needs at least two to one, A needs three to one, A+ needs four to one or better. Agreement alone does not earn a grade.

What counts as an independent signal?

A reading that does not restate another one. Two names for the same observation count once. The analysis lists the factors separately so the total can be checked rather than believed.

Is a high score a reason to trade?

It measures agreement between readings of one image, nothing more. This page describes what the analysis reports, not what you should do with it.

Counting the signals without counting them by hand

Trading AI reads a photo or a screenshot of any chart and returns the confluence score out of 100 with every factor rated separately, the grade it earns, and the risk to reward computed from the plan itself.

The idea itself, explained plainly: What is confluence in trading?

The ratio every grade depends on: What is the risk reward ratio?

And the structure the bias comes from: What is market structure?

This page is educational. It is not financial advice, and nothing here is a recommendation to buy or sell anything.