Define the concept before choosing data
Support resistance reversal is the idea that price interacts with a previously identified support or resistance area and then shifts direction. To assess it, you first need a clear, testable definition of what counts as a “reversal” and what counts as “support” or “resistance.” Without that, any data you collect can be interpreted in multiple incompatible ways.
A practical definition separates the concept into components:
- Level (support/resistance): a region where prior trading activity repeatedly showed rejection, acceptance, or range behavior.
- Reversal event: a measurable change in direction after interaction with that region (for example, the next swing moving back inside the opposite side of the level), using a clearly stated method.
Mechanism and the minimum set of inputs
To evaluate support resistance reversal without making claims about future outcomes, gather inputs that let you reproduce both the level construction and the event detection.
1) Price data with clear provenance
Use historical OHLCV (open, high, low, close, volume) data from a stated data provider or platform export. Data provenance matters because different feeds can differ due to aggregation, corporate actions, or symbol mapping.
You should record:
- Source (provider/platform/export)
- Instrument identity (how the symbol was defined)
- Time zone and session handling
- Data type (spot vs. derived candles; bid/ask ambiguity if relevant)
2) Timeframe and alignment rules
Support and resistance are timeframe-dependent, so you need consistent candle duration and a method for aligning events to candles. Record the timeframe (e.g., 1H, 4H) and how you label:
- the candle where price first “interacts” with the level
- the candle range used to confirm a reversal
3) Level-definition method
Decide how the support/resistance area is measured. Examples of common, checkable approaches include using:
- prior swing highs/lows
- repeated touch/near-touch behavior
- a simple band (range) around a central level
The key is that the rule must be explicit enough that someone else can apply it to the same data and reach the same level boundaries.
4) Event-detection rule for “reversal”
A reversal definition should include objective thresholds, not intuition. For example, specify:
- the direction of the next meaningful swing
- how many candles you require before declaring a reversal
- whether a break of the level versus a rejection determines the event
Evidence or example: what to compute and compare
A simple assessment workflow relies on reproducibility:
- Identify a support/resistance area using your chosen level rule.
- Mark candidate interaction points where price enters the area.
- Apply your reversal event rule to classify outcomes.
You can then summarize results using descriptive statistics (not predictions), such as the proportion of marked interactions that match your reversal criteria over the observed sample. Historical relationships do not establish future results, but descriptive checks show whether your definition behaves consistently in the past.
Limitations and failure modes to include
At least one material limitation should be treated as part of the evaluation, not an afterthought.
Common limitations include:
- Regime changes: what worked in one market environment can fail later.
- Level instability: small changes in the level-definition method can produce different classifications.
- Data aggregation effects: candle construction can hide intrabar behavior that would alter the “interaction” and “reversal” timing.
- Market microstructure and costs: even if a conceptual reversal occurs, real trading results can differ due to execution quality and trading frictions.
Also, assumptions for any threshold (touch distance, confirmation length, band width) should be stated and kept consistent so you can test how sensitive your conclusions are to those choices.
Verification and next questions to ask
Information is worth using only if it can be independently verified. To do that, you should be able to answer these verification questions:
- Can another person reproduce the same levels from the same dataset using your stated rule?
- Can another person reproduce the same reversal event classifications from your stated reversal rule?
- Are your assumptions documented (time zone, candle settings, interaction criteria, confirmation criteria)?
If you cannot clearly list the rules and assumptions, you cannot reliably assess support resistance reversal—because the “data” is incomplete relative to the definition.
You can also sanity-check by testing multiple timeframes and comparing whether the reversal definitions remain coherent or become contradictory.