A candlestick is a compact record of four prices during a chosen interval. It does not show every trade, the order in which the high and low occurred, or whether an order could have filled. Those limits matter when you move from reading a chart to describing a trade rule.
Start with the chart, not the candle pattern name
Before interpreting a shape, identify the exact market and data feed. Write down the venue, pair or contract, price type, bar interval, timezone and chart type. BTC spot, a BTC perpetual contract and an index series can draw different candles even when the labels look similar. A line chart may still use OHLC data internally, while non-standard charts such as Heikin Ashi and Renko can display synthetic prices. TradingView’s current chart information documentation describes these chart fields and the OHLCV values exposed to scripts.
The open is the first recorded price in the interval, the high the highest, the low the lowest, and the close the last price at the interval boundary. Many chart themes color a candle green when close is above open and red when it is below; check the platform settings because the colors are configurable. A small body says only that open and close were near each other relative to the displayed scale. Traders use names such as “doji” for some small-body candles, but naming conventions and thresholds vary.
Work through one fictional candle
Suppose an hourly candle has open 100, high 108, low 97 and close 104. These are invented price units for arithmetic practice, not a market observation.
| Measure | Calculation | Reading |
|---|---|---|
| Full range | 108 − 97 | 11 units |
| Body | 104 − 100 | 4 units; close finished above open |
| Upper wick | 108 − max(100, 104) | 4 units |
| Lower wick | min(100, 104) − 97 | 3 units |
| Close location in range | (104 − 97) ÷ 11 | 7/11, or about 63.6% |
The body shows the difference between the first and last prices. The wicks mark extremes beyond that body. Here the close is above the open and sits about 63.6% of the way from the low to the high. That describes this interval. It does not reveal whether price reached 108 before 97, how long it stayed at either level, who initiated the trades, or what price comes next. Avoid turning “green candle” into “buyers control the market” without a defined, testable measure.
Read a sequence in context
A candle has more context when compared with prior bars and a level defined before looking at the outcome. For example, if a researcher defines resistance as the highest high of the previous 20 completed bars, then tests whether a later completed close exceeds that level, the condition can be written down and repeated. Calling a single wick a “rejection” after seeing the next move is a description made with hindsight, not yet a trading rule.
A useful chart-reading note can be short: “On the specified venue’s standard one-hour BTC-USDT spot chart, the completed close was above the prior 20-bar high.” That sentence records the observation without claiming a forecast. To study it, add the sample dates, exact level formula, entry timing, exit or invalidation, benchmark and cost assumptions. Then check later periods that did not help select the rule. See the guide to testing a crypto strategy for a fuller research process.
Treat a live bar as unfinished
While the current interval is open, its high, low, close and volume may continue to change. The close displayed mid-interval is the latest available price, not the final close. TradingView’s execution-model documentation explains that realtime values update as new feed data arrives and become confirmed when the bar closes.
If your rule requires a completed close, wait until the interval ends before declaring that the condition occurred. A backtest that sees the eventual close and enters at an earlier point in that same bar uses information that was not yet available. A consistent convention is to evaluate the completed signal at its close and model an action at a later bar or price, with costs and gaps stated. That still does not guarantee a real fill; order instructions and execution are separate questions.
OHLC also leaves an important ambiguity inside a bar. If a hypothetical long trade has a stop at 98 and target at 106, both levels fall between this candle’s high of 108 and low of 97. One-hour OHLC cannot establish which threshold was touched first. A researcher needs finer-grained observations or a disclosed conservative convention; silently choosing the favorable ordering overstates what the data proves. A wick through a price is not evidence that a resting limit order filled there.
Keep volume and candle type in their lane
Volume is another field, not a verdict. Its unit can differ by instrument: a chart may report base-asset units for a crypto pair and contracts for a derivative. It can summarize traded quantity, but by itself it does not show order-book depth, your queue position, or “net buying.” Check how the provider defines and aggregates it before comparing symbols or venues. If you need to reason about potential execution, review spread and depth separately.
Use standard candles when the research question concerns the underlying recorded prices. Alternative chart types can be useful visual transformations, but their synthetic OHLC should not quietly become assumed executable prices. The same care applies when comparing exchanges: bar start times, daily resets and last, mark or index price series may differ. For daily crypto bars, record the timezone used to form each day; for perpetuals, distinguish chart price from the venue’s liquidation reference. The guides to crypto market sessions and mark, index and last price cover those details.
SuperQuant’s backtest examples use hourly OHLC bars, and its bundled histories are synthetic teaching data, not live quotations. Hourly candles cannot validate a scalping fill, reconstruct ticks within a bar or prove order-book liquidity. Use a candle to describe the interval, then test any interpretation with identified data, explicit timing and assumptions. This is educational material, not a trading signal or investment advice.