Four layers. One confidence score. One clear read. Here is exactly how XLumience reaches its daily conclusion — and how it names what it isn't sure about.
Every morning you get one clear read: the direction, a 24/48/72-hour outlook, the on-chain and technical picture, the key catalysts on the horizon, and an honest confidence level — with the Oracle's plain-language voice explaining why.
What you do with it is entirely yours. XLumience describes the market — it does not trade for you or touch your funds.
XLumience reaches its daily conclusion by weighing four independent layers: technical (RSI, MACD, EMA — what the chart says), fundamental (regulatory developments, institutional activity, partnerships — what the market structure says), sentiment (Fear & Greed index, market psychology — what traders are feeling), and on-chain (escrow releases, whale activity, ETF flows — what the money is actually doing). Each layer gets a score. The scores combine into a single confidence rating. If a high-impact event sits within the prediction window — like a Senate vote on crypto regulation — the confidence is automatically adjusted downward, because no read should be confident into a coin flip. Only when all of this clears a 60% threshold does the Oracle publish its read.
Every read is assessed against one fixed 24-hour window. We tried an adaptive window that stretched with market conditions — and found that it drifted with price levels and made reads harder to compare. So we fixed it: one day, one claim, one verifiable outcome. The daily briefing still describes how the picture could develop over 48 and 72 hours, but the read that gets measured is always the next 24 hours. Honest scoring beats flexible scoring.
Every read is timestamped, logged and measured against what the market actually did — including the misses. Beta members see the full prediction tracker inside the app. A public track record follows once the first 90 reads have resolved, together with the full scoring methodology. Most providers never publish their methodology — we will.