How to teach candlestick charts

Teaching a candlestick pattern means showing it clearly, then showing what it looks like when it fails, then showing the near miss that is not the pattern at all. Live markets supply none of these on demand.

The usual workaround is to hunt back through historical charts for an example, screenshot it, and crop. That costs time before every lesson, and what comes out carries a ticker, a date, somebody else's indicators and a watermark. The class reads all of it, and most of it is noise around the one thing you wanted them to see.

A drawn chart has the opposite property. It contains exactly what you put in it. The textbook example can be textbook, because you drew it that way, and the failed version can differ in precisely the one respect you want to talk about.

Start with the clean case. Place the candles the pattern needs and set each open, high, low and close until the shape is unambiguous. Nothing has to be realistic in the sense of having happened, only consistent: a body and its wicks have to agree with the four numbers behind them, and the canvas keeps that true while you drag.

Then copy it and break it. Move one close through the previous open and the engulfing pattern stops being one. Lengthen a wick and the hammer becomes something else. The variations are the lesson, and they are quick once the clean case exists, because you are editing a chart rather than searching for a new one.

If a longer sequence is wanted as context, generate one from market statistics and shape the part that matters. That is faster than placing forty candles by hand when only four of them are making a point. The guide to drawing a chart covers all four ways in.

A worked example shown complete teaches recognition of a picture. The harder and more useful skill is deciding what happens next while it is still uncertain, which needs the answer withheld.

CandleDraw does this two ways. A reveal boundary is a line you place on the chart: everything to the right of it is hidden, so the class sees the setup and not the outcome. Alternatively you can mark individual candles, drawings or studies as the answer key, which suits a question like "draw the trendline you would use here".

Hidden content is removed on the server, not merely covered up. What reaches the student is a chart that does not contain the answer, so the diligent student who opens the developer tools finds the same thing the rest of the class does.

The same chart becomes an assignment. Students reach it through a link, land in the class without configuring anything, and work on the canvas you drew on rather than a picture of it. They can move the candles you left movable and not the ones you locked.

Work comes back to you as a submission you can open, draw on, and return with written feedback. Drawing on it matters more than it sounds: "your trendline is two candles late" is a sentence, but a line on their chart beside theirs is the correction itself.

You decide when the answer key becomes visible: as soon as they submit, only once you have approved the work, or never. The first suits self-paced practice. The second suits anything you intend to mark, because it stops the answer arriving before you have read what they did.

It does not stream live prices, and it does not detect patterns for you or tell anyone what to buy. There are no signals. A chart here is something you made to explain a point, and it is worth being plain with a class about that: the candles in a drawn example did not trade.

Where real prices matter, import them. An OHLC file gives you the session that actually happened, and the same annotation and assignment tools work on top of it.