Six zones, drawn from candles, and nothing else on the chart

A level only earns a line when something happened there. Round numbers, yesterday's high and a Fibonacci grid are not events. They are conveniences, and price runs through them whenever it feels like it.

Keeping them has a cost you can watch in your own results. Eleven lines on a 15-minute chart means price is always near one of them, so every setup looks confirmed. The lines have stopped filtering anything. You take the trade you would have skipped, and then the loss gets blamed on your entry timing.

Shrink the definition instead. A level marks a spot where buyers or sellers already turned the tape. You can point at the candles that prove it. No candles, no line. Erase the rest before the open.

Levels come from the extremes of ten candlestick reversal patterns and nothing else. Hammer, shooting star, bullish engulfing, bearish engulfing, piercing line, dark cloud cover, morning star, evening star. Then the two gap patterns: a rising window when price gaps up, a falling window when it gaps down.

Support is the lowest low of the pattern. Resistance is the highest high. Take the wick, not the body. The wick is where price traded and got pushed back, and that print sits on everyone else's chart too.

One gate decides whether a shape counts, and it is the move running into it. A hammer needs a downtrend behind it. A shooting star needs an uptrend. A hammer shape in the middle of a flat drift is a candle with a long tail. The level you draw from it leaks.

Draw on the daily, the 4-hour and the 1-hour. The 5-minute produces a level every hour and almost none survive to lunch. Leave it alone until you are timing an entry.

Switch the chart to regular trading hours before you start. Overnight candles print on thin volume and leave wicks that nobody defends the next day. If your platform includes pre-market by default, that setting is adding fake levels to every chart you open.

Start at the current price and work left, one candle at a time. Every time you hit a qualifying pattern, mark its extreme and keep going. A level survives only if price has not already closed through it since. On the daily you rarely need more than a few months of history. Anything older has usually been cut through already.

Two levels a few cents apart are one level. On a large-cap stock in the hundreds, anything within about a dollar of another level joins the same zone. Scale that distance with the share price. A 30 dollar stock needs a tighter merge than a 600 dollar one.

Draw the merged zone as a rectangle running from the lowest extreme in the group to the highest. Then cut the chart down. Keep the three nearest zones above price and the three nearest below, and delete everything else. Six is enough to plan a session. A zone 40 dollars away is not a decision you make today.

Nobody can hand you a maximum zone width. That one is judgment. If the rectangle is so tall that a stop beyond it would break your risk, you merged too much. Split it back into two.

A zone is a no-trade area. Price inside it is still arguing, and those candles are the least reliable ones on the chart. Wait until price is clearly above the rectangle or clearly below it before you do anything.

The same rectangle serves both directions. While price sits under it, that is where you look for shorts and where you take profit on longs. Once price closes above it and holds, the same rectangle becomes the place you want to buy a pullback into. Broken support turns into resistance, and broken resistance turns into support.

So do not delete a zone the moment it breaks. Keep the rectangle and change what you do with it. The levels you drew on Sunday are still the levels on Thursday. Price sitting on the other side of two of them changes nothing.

Run this once a week per symbol, and again after any day that closes outside your zones. Ten minutes a chart, once you stop second-guessing the patterns.