Lines drawn after the move always look right

Draw them before the market opens, either the evening before or early that morning. Use charts whose candles have already closed: the daily, the 4-hour and the 1-hour. For stocks and ETFs, use regular trading hours only.

Start each session with a clean chart. Delete yesterday's lines and redraw from where price sits now. Old levels that still matter will show up again when you walk back through the candles.

Once the bell rings, the map is fixed. You can add a level if price runs into territory you didn't chart. You shouldn't nudge an existing line toward where price just turned.

The reason is simple. A level is a forecast, and the open is when the forecast gets tested. It's only useful if you wrote it down before you knew the answer.

Look at any chart after the close and you'll find lines that worked. Price turned at a prior high, bounced off an old low, stalled at a gap. It all looks obvious.

That's hindsight doing the drawing. Your eye goes straight to the places where price reacted and builds a line through them. The prices where nothing happened don't get a line, so they never count against you.

There's a second trap. A busy chart has a level every few cents if you look hard enough. Price will touch some of them by chance. Count the touches afterwards and almost any set of lines scores well.

This is why level-drawing feels easy and trading from levels feels hard. One happens with the answer on screen. The other happens before it.

To find out whether your levels work, freeze them before the open and score them after the close. Write the prices down with the time. A screenshot or a dated note is enough.

After the close, go level by level. Did price reach it? If it did, did it turn there, or close straight through? Keep a running tally over at least 20 sessions.

Then add a control. Each morning, also write down two random prices at a similar distance from the open. Score them the same way. Your real levels should beat the random ones. If they don't, your process is finding busy areas instead of turning points. That comparison is uncomfortable. It's also the only one that tells you anything about your levels.

Keep the tally somewhere dated, next to the day's results. A level that held on a day you lost money still held. A level that failed on a green day still failed. Score the map and the trading separately.

Start at the current candle and walk left, one candle at a time. Ask which high price must clear next, then the one after that. Do the same for lows below.

Mark the extremes of clear reversal patterns and gaps. When two levels sit too close to trade between, merge them into a zone. Inside a zone you wait. You act once price is clearly above or below it.

Stop at three levels above price and three below, counted across all three charts. A line, a zone and a gap each count as one. More than that and every move looks like it's near something.

How close is too close for a zone depends on the instrument's price and how much it moves. That part is judgment. Pick a width, write it down, and keep it the same while you score.

This fits in the half hour before the cash open. Times are Eastern.