KNOWLEDGE · PROCESS · PRACTICE
Move from a market observation to a written scenario, then test what actually happened. Separate discovery, direction, timing and execution.
Start with one instrument, one venue and a defined horizon. More indicators help only when they answer a new question. Prepare a scenario, distinguish disagreement from missing data and review the outcome without rewriting the original decision.
Trend, range, compression, chaos and transition change how the same signal should be interpreted.
Swing levels, ranges, breakouts and retests tell you where the market is.
Flow, derivatives and cross-market context should answer specific questions.
Fees, spread, slippage and liquidity can remove the apparent edge.
Record the time, instrument, venue, horizon and condition that would invalidate the idea. In a hypothetical range from 98 to 102, trading above 102 is only an observation. Decide beforehand whether your test requires a closed candle, continued acceptance or a retest. If price returns inside the range, record that outcome instead of moving the boundary to preserve the story.
A short rebound can coexist with weak four-hour structure. Check what each output predicts: end-of-window direction, a threshold touch and a swing setup are different targets. Read when the forecast was issued and which interval was evaluated. Do not average unlike scores. A longer horizon can include a session change or event absent from the short-horizon evidence.
An educational plan assumes a 0.8% move, 0.20% total fees and 0.15% spread plus slippage. The remaining 0.45 percentage points describe a cost-adjusted scenario, not expected profit: the target may never occur and losses still matter. After the horizon closes, compare the frozen plan with the observed path, available prices and invalidation. Keep paper analysis separate from executed trades.
Begin with the workflow guide, then study the specific question you need to answer. These guides cover independent analysis, not a signal subscription or a promise of profitable results. Each guide states the limits of its method and distinguishes educational examples from live market information.
Choose the venue, quote currency and horizon. A reference chart and an executable quote are not interchangeable.
Record structure, confirmation, costs and invalidation before checking the eventual outcome.
Track false alarms and missed opportunities. A busy model is not necessarily a better model.
A useful research question specifies an asset, executable market, decision time and outcome window. “Will this rise?” leaves all four open. “Does a close above the previous hourly range precede a move large enough to exceed estimated round-trip costs within…
A regime is a summary of several observations, not a property permanently attached to a coin. Separate direction, realised variability and persistence. A market can trend steadily with narrow candles, or move sideways with large swings. Compare successive…
RSI, moving-average slope and a momentum oscillator often reuse the same closing prices. Agreement may therefore reflect one underlying movement three times. A different view might come from executed trade imbalance, the cost of crossing the book or…
A four-hour candle that is still forming contains an unfinished result. When comparing it with a closed five-minute signal, note that their certainty differs. For historical testing, use only the higher-timeframe candle that had actually closed at the…
A scanner's HOLD usually describes the current opportunity, not an instruction to keep every asset already owned. Existing positions involve a different entry price, exposure and invalidation condition. Waiting can also mean different things: insufficient…
A compact review can ask: which asset and venue; which horizon; how fresh the data is; what the broader structure says; what confirms the immediate setup; what invalidates it; whether flow supports the interpretation; what liquidity is available for the…
Continue learning