Reading order flow.

The chart answers a narrow question

A candle summarizes prices within an interval. Its open, high, low and close describe the range and endpoints; they do not preserve the order of every event inside it. Volume adds an aggregate measure of trading. Neither tells you the full depth of the market at each moment.

This is not a defect in the candle. It is a useful compression for a particular question. Trouble begins when a compressed view is treated as if it contains evidence it never recorded. Before comparing two screens, check the instrument, session, timezone and aggregation.

Resting interest and executed trading

A depth ladder displays interest available at price levels at the observed moment. In an aggregated book, that interest may combine many orders. It can be changed or cancelled. A large displayed bid is therefore not proof that a participant will buy at that price.

Time and sales records executions. A print’s price, size, timestamp and available side classification describe an event that happened. Classification often identifies whether the trade was associated with the bid or ask; it does not identify the participant or explain their full intent. Looking at the ladder and tape together keeps this distinction in view.

What a footprint adds

A bid × ask footprint groups executed volume by price inside each bar. It helps locate where classified buying or selling occurred. A delta is a difference between those classified sides under a chosen convention. An imbalance compares values using a defined rule, sometimes across neighboring prices.

Check that rule before interpreting the highlight. Tick aggregation, zero handling, minimum volume and diagonal versus same-price comparison can change what is marked. A bright cell is a calculation, not an automatic instruction to enter a trade.

• Executed volume. What was classified and recorded as traded at the selected price or interval.
• Resting liquidity. What was displayed in the book at the sampled time.
• Inference. What you hypothesize about behavior after examining the recorded evidence.

Profiles need a session

A volume profile distributes executed volume over price for a selected span. A TPO or market profile describes time-based participation under its own construction. Their points of control and value areas have different definitions. They should not be casually interchanged.

Choose the session deliberately. A level from one span can differ from a level calculated over a longer history. Keep the boundaries and the calculation settings with any observation you plan to review. Otherwise the apparent disagreement may simply be two different questions.

Missing data changes the story

An imported OHLC bar file cannot recreate missing prints by price and side. A tape without recorded book events cannot manufacture a historical heatmap. A sampled heatmap can miss changes between observations. Treat missing coverage as unknown, rather than zero.

For a fair analysis, record the source, available resolution, corrections and gaps. Synthetic data is useful for learning how a tool works, but does not establish actual market behavior. When using provider data, confirm the licence and entitlement as well as the technical connection.

A practical observation record

Write the observation before the conclusion. For example: the chosen level had displayed bid interest at the sampled time; classified selling traded into it; price remained within a stated range for a stated interval. Then write the hypothesis and what would contradict it. This keeps a description of evidence separate from a forecast.

During review, return to the same source and settings. Compare the actual sequence with the plan instead of choosing a later screenshot that makes the outcome look inevitable. Order flow can sharpen a question; it cannot remove uncertainty or guarantee a profitable decision.