Price action is the study of how prices develop through time, usually using candles, swings and areas of previous reaction. It does not require an indicator, but removing indicators does not remove uncertainty. A chart is a summarized record; it does not reveal every order, participant identity or reason for a move.
A beginner benefits from describing what is observable before making a forecast. Record the timeframe, completed prices, trend or range definition and the condition that would invalidate an interpretation. This guide uses invented prices to build a repeatable study process. It does not recommend a live setup or suggest that a recognizable candle creates a profitable strategy.
Read the candle without inventing its cause
A candle records open, high, low and close for an interval. Its body spans open to close; its wicks extend to the extremes. A fictional bar opening at 100, reaching high 106 and low 98, then closing at 104 has a four-unit rising body, two-unit upper wick and two-unit lower wick. Those distances describe the result, not the sequence of every transaction inside it.
The same OHLC values can arise through different intrabar paths. That matters when a backtest claims both a stop and target were touched in one candle: OHLC alone may not establish which came first. Also check whether the chart uses Bid or Last and which price side activates the relevant order. A screenshot is not an executable quote history.
Describe structure at the chosen scale
An upward sequence is often described using higher confirmed swing highs and lows; a downward one uses lower highs and lows. A range lacks a sustained progression under the chosen rule. The word confirmed matters: a swing based on two candles on each side becomes known only after the two later candles exist.
A five-minute decline can sit inside a daily uptrend without either description being wrong. Use a higher timeframe for context and a lower one only for the explicitly defined question, rather than searching through timeframes until one supports the intended direction. If the swings are ambiguous under your rule, label the structure uncertain instead of forcing a trend.
Levels identify areas to observe
Previous swings, session extremes and repeated reactions can provide support or resistance references. Decide whether bounds use full candle ranges or bodies, how nearby levels are combined and when they expire. Zones can fail; repeated reactions do not establish that the next one must work. A round number is a reference to test, not a guaranteed barrier.
Candlestick names also need context. A long lower wick inside a noisy range answers a different question from the same shape at a previously declared area. Specify the size and position requirements and the time at which the condition becomes known. Avoid calling an entire move a successful pattern only after its outcome is visible.
Turn one idea into a complete hypothetical rule
For an educational example, suppose a range of 100–110 was marked using completed observations. A research rule waits for a completed close above 110, then records a hypothetical entry at 112, invalidation at 107 and target at 122. The gross price-distance reward-to-risk is (122 − 112) ÷ (112 − 107) = 10 ÷ 5 = 2:1. These numbers are fictional.
That ratio does not measure win probability or realized expectancy. Add when an order is placed, expiry, allowed volume, maximum positions and what happens if the first executable price differs from 112. A later fill at 115 with unchanged stop and target gives 7 ÷ 8 = 0.875:1. Execution changes the economics even if the chart idea remains recognizable.
| Decision | Example question |
|---|---|
| Context | Which completed observations define the range? |
| Trigger | Is a wick enough, or is a completed close required? |
| Execution | What price and order type could first be available? |
| Invalidation | What closes the idea, and is the stop sent to the broker? |
| Evaluation | Are unfilled orders, costs and failed breaks retained? |
Review outcomes without relabelling them
Save a screenshot before the trigger and a record after the outcome. Include setups that never trigger, limits that never fill, losses and cancelled ideas. Otherwise, a folder of attractive finished patterns selects only what was easy to explain afterward. Record both maximum favorable and adverse movement if your data supports it.
Use a separate period to evaluate frozen rules and keep the history of parameter changes. A small selected sample cannot establish a dependable success rate. Testing alternative exits after every loss can also make the final backtest an account of optimization rather than evidence that the original plan worked.
A beginner's checklist and questions
Is price action better than indicators? Neither label establishes an edge; indicators transform prices while price-action rules still require definitions and evidence. Can a pattern guarantee an outcome? No. How many patterns should a beginner learn? A small set of well-defined observations is easier to review than many vague names, but its usefulness must still be tested.
- Identify the instrument, feed, price side, session convention and timeframe.
- Describe completed candles and confirmed swings before writing a directional interpretation.
- Set the zone, trigger, invalidation and expiry before seeing the next move.
- Translate stop distance into account-currency exposure and include trading costs.
- Keep failed patterns and compare results on observations not used to choose the rule.