Reading a chart without fooling yourself.
A chart is a record of the past, drawn to look like a map of the future. Candles, volume, and liquidity are real, useful information. The lines people draw on top of them are mostly a story the brain tells to feel in control. This lesson teaches the mechanics straight, then shows you the evidence on what actually survives.
not advice · not a signal · educationalA chart tells you what already happened, with total precision. It tells you what happens next with none. Everything below is about keeping those two things separate. When you catch yourself reading the past and feeling the future, that feeling is the thing this lesson is trying to disarm.
The candle. Four numbers, nothing more.
One candle is four numbers over a slice of time: where price opened, the highest it reached, the lowest, and where it closed. The thick part is the body, from open to close. The thin lines are wicks (also called shadows), the high and the low. Colour just tells you whether close landed above the open (up) or below it (down).
the lessonWhat a single candle can and cannot tell you
The information is real. The interpretation is where people leave the rails.
- Real: the range (high minus low) is exactly how much price moved. A long wick means price got there and got pushed back before the close. That happened. It is data.
- Story: "a long lower wick means buyers are stepping in." Maybe. Or one large order filled and left. The candle cannot tell you who or why, only what.
- The named patterns — doji, hammer, engulfing, and the rest — are just labels for candle shapes. They describe the past shape cleanly. They do not carry a reliable edge about the next candle; that is the part the evidence section deals with.
- Timeframe is everything. A daily candle on a 24-hour market is 288 five-minute candles stacked into one. The same move is a calm green day or a violent morning depending on which you open.
The same prices, drawn two ways.
A line chart connects the closes and throws everything else away. Candles keep the range and the open. Neither is more "true" — they are the same underlying closes with different amounts of detail. More detail is not automatically more signal. It is also more surface to draw stories on.
Volume. The one input people skip, and the one that gets faked.
Volume is how many units changed hands in the period. It is the closest thing a chart has to a truth serum: a move on heavy volume had real participation behind it, a move on thin volume can be a single order pushing price through an empty book. "Volume confirms" is one of the few chart ideas with a mechanical reason behind it.
But volume is only honest if the number is honest. In crypto it very often is not. This is the part almost no charting course tells you.
In 2019, asset manager Bitwise told the SEC that roughly 95% of reported Bitcoin spot volume was fake or non-economic — about $6 billion in reported daily volume, of which only around $273 million was real, concentrated on ten exchanges. (Bitwise / SEC filing, 2019)
A peer-reviewed study, "Crypto Wash Trading" (Cong, Li, Tang & Yang), later measured it directly and found wash trading averaged over 70% of reported volume on unregulated exchanges, using the statistical fingerprints above. It was published in Management Science in 2023. (NBER w30783)
Liquidity: the number the chart doesn't show you.
The chart shows the last price. Liquidity decides what price you actually get. It is the difference between the number on the screen and the number in your fill, and on small tokens that gap eats more accounts than any bad entry ever did. There are two shapes it comes in.
- Thin book, big candle, no follow-through. A green spike on a token with a shallow book often means one buyer swept thin asks. The next seller gives it all back. The candle looked strong; the liquidity was not there.
- The exit is the real test. Getting in is easy — you are paying up into the book. Getting out of size, into thin bids, is where the loss actually lands. Always look at the bid side before you size a position, not just the last print.
- Depth is verifiable, price predictions are not. You can pull the live book or the pool reserves and see exactly how much a trade will move price. That is one of the few honest, checkable edges on this whole page.
Market structure describes. It does not prophesy.
"Higher highs and higher lows" is a clean way to describe what a chart has been doing. Each new peak above the last, each dip shallower than the last — that is an uptrend, by definition, in the window you are looking at. Using the word is fine. It is a summary of the past, and an honest one.
The slip happens in one word: until. "Higher highs and higher lows, so it keeps going up" quietly swaps a description of what happened for a claim about what happens next. The structure is real right up to the candle where it isn't, and nothing in the pattern tells you which candle that is.
Does reading charts actually make retail money?
This is the section the courses selling you patterns leave out. The mechanics above are real. The question is whether acting on them, as a retail trader, produces money. The academic record is large, consistent, and not kind. Here is what the studies actually found, with links so you can read them yourself.
| Study | What it looked at | What it found |
|---|---|---|
| Day Trading for a Living? Chague, De-Losso, Giovannetti · 2020 |
Everyone who started day trading Brazilian equity futures, 2013–2015 | Of those who persisted more than 300 days, 97% lost money. Only 1.1% earned more than the minimum wage; 0.5% beat a bank teller's starting pay. No evidence of learning over time. SSRN |
| The Cross-Section of Speculator Skill Barber, Lee, Liu, Odean · 2014 |
Every day trader in Taiwan, 1992–2006 | Less than 1% reliably earned positive returns net of fees. Skill exists, but it is rare and mostly the same few names. The bottom lost −28.9 bps a day after costs. PDF |
| Profitability of Technical Analysis: A Review Park & Irwin · 2007 |
95 modern academic studies of technical trading rules | 56 positive, 20 negative, 19 mixed — but most positive results were undercut by data snooping and ex-post rule selection, and profits faded after the early 1990s as markets adapted. Journal of Economic Surveys |
| ESMA retail CFD disclosure EU regulator · ongoing |
Retail CFD accounts across EU brokers (leveraged short-term speculation) | Between 74% and 89% of accounts lose money. It is now a mandatory risk warning brokers must print on their own ads. ESMA |
Test enough patterns against enough history and some will look profitable by pure chance. Backtest 1,000 indicators and a handful will have "predicted" the past beautifully — and predict the future no better than a coin. This is the core reason a strategy that looks flawless on the chart falls apart with real money. The pattern did not find an edge. The search found a coincidence. That is the machine most of retail TA runs on.
The same chart, two stories.
There is a reason chart-reading feels so convincing even when the evidence says be careful. Your brain is a pattern-finding machine that cannot turn off. It finds faces in clouds and trends in noise — the name for it is apophenia. Point it at a price chart and it will hand you a clean story every time, whether or not one is there.
Apophenia hands you the pattern. Then confirmation bias keeps it: once you have decided it is going up, every green candle is proof and every red one is noise. You remember the calls that worked and quietly forget the ones that didn't. The chart becomes a mirror that agrees with whatever you already wanted to do.
What's actually worth keeping.
Strip out the fortune-telling and a real, smaller toolkit is left. These survive because each one is checkable — it is about the present or the past, not a claim on the future. That is the whole filter: keep the parts you can verify, drop the parts you can only believe.
- Liquidity levels you can pull. The live order book or the pool reserves tell you exactly what your trade will cost in slippage. Real number, checkable now, before you click.
- Volume as confirmation — once you trust the source. On an honest feed, a move on real volume had real participation. On a random token page, verify the volume is not the 70–95% that studies find is fake before you lean on it.
- Timeframe honesty. Name the timeframe out loud and match it to your actual holding period. A 5-minute chart tells you nothing useful about a position you plan to hold for weeks, and vice versa. Most "conflicting signals" are just two timeframes argued as one.
- Structure in the past tense. "It has been making higher lows" is a fair description. Keep the description, drop the prophecy.
Every other lesson here comes back to one number: sample size. A pattern that won five times in a row tells you almost nothing — five is noise. The only way to know whether anything you do on a chart has an edge is to journal every trade — entry, exit, why, timeframe, and outcome — and let the count grow until the number is big enough to trust. Most traders never build the sample, so they run on the last three trades, forever. A losing method with a lucky week feels identical to a real edge until the N gets large. Small samples lie, and a chart is the easiest place to get lied to.
If you want to see this sample-size discipline applied to a real strategy with real costs loaded in, the Polymarket 5-min walkthrough → shows what happened to a great-looking edge once every fill was cost-loaded. And the options course → covers the seller-side mechanics this lesson stayed out of.