Support and resistance are the foundation under every chart pattern on this page. They are horizontal price levels where supply and demand have repeatedly met. Below resistance, supply usually wins. Above support, demand usually wins. Most of the structural patterns you’ve read about — double tops, triangles, flags, head & shoulders — are really just different shapes of price interacting with horizontal levels.
Understanding S/R properly is the difference between trading a chart and reading one. This article covers what makes a level real, how to draw it, what happens when it breaks, and the failure modes that catch most retail traders.
A valid level has at least two clear touches — three or more is much better. Each touch should be a meaningful reaction: a sharp rejection, a strong reversal candle, or a clear pause. Levels marked by one wick on one bar are noise. Levels marked by three rejections over weeks are real.
Three rules for drawing them:
1. Use wicks for the most honest line. Some traders use candle bodies; both have a logic. Pick one rule and stay consistent. Constantly redrawing the line to fit recent price is curve-fitting, not analysis.
2. Levels are zones, not exact prices. A level is rarely held to the penny. A few ticks of slop above or below is normal. Mark a band, not a hairline.
3. Higher timeframe levels are stronger. A level the daily chart respects is more meaningful than one only the 5-minute chart sees. Always check the higher timeframe before drawing on the lower one.
Three forces stack at any horizontal level:
• Memory. Traders who bought at the level earlier defend it; traders who sold near it remember the price. Resting orders cluster there.
• Round numbers. Psychological levels (4500 on ES, 100K on BTC) attract attention even without prior trading history.
• Algorithmic targeting. Modern markets are dominated by algos that explicitly reference recent highs and lows. They reinforce the levels human traders are watching.
None of these is mystical. S/R works because enough participants act on it that the level becomes self-reinforcing — until enough new flow shows up to break it.
A wick poking through a level is not a break. The signal is a close on the other side — ideally with a body that fills most of the candle’s range, and ideally with conviction (volume, momentum).
Stricter traders add a buffer: the close must be beyond the level by some amount (often 0.25-0.5% of price, or some multiple of ATR). This filters out marginal breaks that immediately reverse.
The most useful S/R behaviour is the flip: once support breaks, that level becomes resistance on the next test. Once resistance breaks, it becomes support. The level itself does not disappear — its role just inverts.
This is where the cleanest setups live. A retest of a recently-broken level with a rejection candle gives you a tight stop, clear confirmation, and a defined target on the next structure. The first retest of a broken level is the highest-quality entry in most price action playbooks.
S/R levels fail constantly. Three most common failure modes:
A bar’s wick punches through but the body closes back inside. No close-through means no break. The level holds.
Price breaks the level, sweeps the stops resting just past it, then reverses hard. This is a liquidity sweep in textbook form.
Trader tilts the level to fit recent price, creating a “break” of a line that wasn’t really there. Overfitting masquerading as signal.
Key insight: Most chart patterns are S/R patterns wearing different costumes. A double top is two failed attempts to break resistance. A flag is consolidation against support. A head & shoulders is three rejections at a level. Once you read S/R cleanly, every other pattern reads as a variation on the same theme — not as separate things to memorise.
S/R is subjective. Ten traders draw ten slightly different lines. Most of the time the market does not care about exact prices — it cares about zones where significant flow defends the price. A level held three times on heavy volume is a real level; one you tilted yesterday to fit the chart is not. Draw lines you would have drawn last week and last month. If you keep redrawing them, you are not analysing — you are explaining.