Methodology & Glossary

How our scores work

MarketScanner Pros is an educational market-analysis platform. Every score it shows is a composite summary of evidence — not a statistical probability, not a forecast, and never a personalised instruction to buy, sell, or hold. This page explains exactly what each score, badge, and state means so you can judge the evidence for yourself.

Scores are composite strength, not probability

A number like Composite Strength 78 / 100 (also shown as MSP Score or Confluence in places) is a weighted summary of how much of the current evidence points one way. It is a heuristic — the weights are rules chosen for transparency, not calibrated against historical outcomes. A 78 does not mean a 78% chance of anything. Where a tooltip says “reflects how many indicators agree, not a probability,” that is the literal truth of the number.

Independent factor groups

Many indicators measure the same thing. EMA position, MACD, ADX, and Aroon are all expressions of trend. If four of them agree, that is one piece of trend evidence — not four. To avoid overstating agreement, we collapse correlated indicators into a small set of independent factor groups and count agreement across the groups:

Trend — EMA structure, MACD, ADX/DI, Aroon

Momentum — RSI, Stochastic, CCI

Participation / Volume — OBV, MFI, VWAP, relative volume

Volatility — BBWP, ATR, DVE phase

Relative strength — vs a benchmark (SPY for equities, BTC for crypto)

Market structure — levels, breakout proximity, EMA stack

Positioning — funding, open interest, liquidations (crypto)

Regime — trend / range / risk-on / risk-off context

Macro / cross-asset — DXY, rates, VIX regime

Catalyst — earnings and scheduled events

Evidence quality

Evidence quality grades the inputs behind a conclusion — how many independent factors are available, how fresh the data is, and whether factors conflict. It caps the composite: poor or stale evidence cannot produce a high headline number.

HIGH

Broad coverage, current data, no major conflicts.

MEDIUM

Adequate coverage; some data delayed or a factor missing.

LOW

Sparse coverage, stale data, or conflicting factors.

INSUFFICIENT

Too few independent factors, or data missing/simulated — treat as exploratory only.

Setup stage

A description of observable behaviour — where a move appears to be in its lifecycle. It is context for research, not a trade instruction.

Dormant

Quiet; nothing developing.

Building

Participation rising while price is still contained — the early signal.

Confirming

Independent factors agree and price is beginning to validate the move.

Expanding

Participation and price expanding together with rising volatility.

Extended

A significant move has already occurred; the early window has passed.

Fading

Participation or momentum deteriorating after a move.

Extension state

How far a move has already run, from distance-from-average, range, and volatility percentile. Early / Normal / Elevated / Extreme. An Extreme reading means the obvious part of the move may already be behind it — context, not a signal.

Data freshness & provenance

Every data source carries a freshness badge: Live (current within the expected provider cadence), Delayed, Stale, or Missing. Panel-level badges reflect the weakest underlying row — a panel will not show “Live” over degraded or cached data. Delayed data is never presented as real-time.

Derived vs observed data

Some readings are estimated, not directly observed, and are labelled as such. For example, a “Long/Short (funding-implied)” reading is derived from aggregated funding rates — it is a proxy for positioning, not exchange long/short account data. We label derived, delayed, estimated, and simulated data so you can weight it accordingly.

Market regime

Regime is the broad environment — trending, range, risk-on, risk-off, expansion, compression. The same indicator can mean different things in different regimes, so regime is shown as persistent context. When evidence is thin, the platform is comfortable saying Mixed / Conflicting or Unknown rather than forcing a call.

What we deliberately do not do

We do not present composite scores as statistical probabilities.
We do not issue buy / sell / hold instructions or personalised financial advice.
We do not connect to brokers or route orders — portfolio and backtest features are simulation only.
We do not publish fake testimonials, invented usage statistics, or guaranteed returns.
We do not present delayed, estimated, or simulated data as live, observed data.

MarketScanner Pros provides market data, analytical tools and educational research. It does not provide personalised financial advice. Trading involves substantial risk. See our disclaimer and compliance hub.