The patterns transfer. The market structure does not.

Most “TA for Polymarket” advice skips the only step that matters. If the chart is a probability line, every indicator you drop on it is an indicator of closes. RSI of closes. MACD of closes. A moving average of closes. You are studying a compressed series and calling the compression a signal.

Rebuild open, high, low and close from the trade feed first. Read the range. Then decide whether you even want an oscillator. How a Polymarket candle is read is the primer; this page assumes you have that picture and asks what you are allowed to do with it.

Support and resistance as levels the book has already argued about. A 50% handle that has been tagged three times and held is a real object. So is a 70% ceiling that has rejected twice. You do not need a special prediction-market theory for that. People remember round numbers, and they remember pain.

Wick rejection. A candle that runs to 38% and closes at 52% is a book that tried a story and could not keep it. The wick is the failed story. On a line that hour is a fourteen-point rally and you never see the attempt.

Multi-timeframe agreement. A market that is up on the week, up on the day and breaking on the hour is a different trade from one that is aligned across the board. Multi-timeframe signals are how Predm puts that disagreement on one row.

Volume as a condition, not a vibe. A level taken on a dead book is not the same level taken on a crowded one. Thin prediction markets print theatrical wicks. Treat them as theatrical until size shows up.

Unbounded trend-following. A crypto chart can trend for a year. A prediction market is pulled toward 0 or 1 by a date. The closer you get to resolution, the less a “trend” is a trend and the more it is the destination asserting itself. Riding a 92% market because the moving average is up is not the same trade as riding Bitcoin because the moving average is up. One of those still has room. The other has eight points and a calendar.

Classical pattern folklore at the extremes. Head-and-shoulders, cups and flags were catalogued on markets that can make new highs forever. Near 5% or 95% the geometry gets crushed against the bound. A “bull flag” at 96% is a drawing on a wall.

Indicators that assume returns are roughly symmetric. Many oscillators were tuned on assets whose next tick can be up or down by similar amounts. A 90% YES contract cannot rally the way it can sell off. The distribution of the next move is skewed by construction. An RSI that says “overbought” at 92% is often just describing the bound.

Ignoring the other side of the book. Crypto TA rarely asks who is holding the offer. On a prediction market the price is the crowd, so holder concentration is not a side quest. A breakout through 60% on a crowded YES book is not the same breakout as one where YES is a crowd and NO is three wallets. The candle cannot see that. The concentration numbers can.

If you want a beginner method that does not fight the market structure, keep it to four questions, in this order:

1. Is there a real candle? Enough trades in the interval, on a timeframe the book can fill. If not, stop. You are looking at noise with a face.

2. Where is the range relative to the destination? A 20-point day around 50% is a fight. A 20-point day around 90% is a book that is not as decided as the close wants you to think. Same range, different meaning.

3. Do the timeframes agree? Daily up, hourly breaking is a warning, not a system. It tells you not to treat the daily close as a settled fact.

4. Is there a crowd behind the close? If the side you would be fading is three accounts, you are not fading a consensus. You are fading a desk.

Everything else — Fibonacci, Ichimoku, a twelve-indicator dashboard — is optional decoration. It is not banned. It is just not the thing that makes prediction-market TA different, and it is the first thing people add to feel like they have a process.

Traders treat 50% as a fair-coin line. Sometimes it is. Often it is just a round number the same way $100,000 is a round number on Bitcoin: people park orders there because it is easy to remember, not because the universe agreed.

Better levels are ones the book has already used. Prior day’s high and low. Prior week’s close. The price where a known whale’s basis sits, if you have that data. The strike implied by a related market. Those are levels with a reason. 50% has a reason only when the event is actually close to a coin flip, or when the book has treated it as one for long enough that the memory is real.

Round numbers still matter — 25, 40, 60, 75 — because humans type them. Just do not build a religion on the midline of a bounded axis.

A stock candle does not know about earnings until you mark the date. A prediction market candle sits on top of a known expiry. As that date approaches, the same pattern means less. A bullish engulfing three months out is a change in the estimate. A bullish engulfing the night before resolution is often just the last people getting flat.

This is the part of prediction-market TA that has no crypto equivalent. You have to read the calendar with the chart. A method that does not change as resolution approaches is a method that is pretending the destination is not there.

Not a strategy, and not a backtest. This page is a reading guide. It does not claim a win rate. Any screenshot of a “perfect” Polymarket short is a screenshot, not evidence.

Not financial advice, and not a venue. Predm is a read-only analytics terminal. You cannot place an order here.

Not a claim about every prediction market. The structure above is about binary event contracts of the kind Polymarket lists. Hyperliquid perps, which Predm also reads, behave like the crypto charts you already know. The comparison page keeps those two books from being mashed into one story.

Can I use TradingView indicators on Polymarket?

You can point any OHLC indicator at a rebuilt series. The maths will run. The interpretation has to change because the axis is 0–100 and the contract resolves to 0 or 1. If your indicator assumes unbounded prices, treat its output as a sketch.

Is candlestick-pattern trading a thing on Polymarket?

Single-candle folklore (doji, hammer, engulfing) is the same folklore it is everywhere: occasionally descriptive, rarely a system. The useful part of a candle on a prediction market is the range, not the name of the shape.

What timeframe should a beginner start on?

Daily, on a market that actually trades. Intraday candles on a thin event are how people convince themselves they saw a pattern. Once the daily range makes sense, add the hour. How to chart Polymarket covers the rebuild and the timeframe choice together.

Does holder concentration count as technical analysis?

It is market structure, not a price pattern. It belongs next to the chart because it tells you how much the close is allowed to mean. A technical level held by four wallets is a different object from a level held by a crowd.