Same four numbers. Completely different market.

On Bitcoin, $100,000 is a price. It can be $90,000 tomorrow or $120,000. On a Polymarket YES contract, $0.63 is a 63% chance the event happens. The dollar is a unit of account for a claim that settles at $1 or $0. There is no third number.

That single fact changes every subsequent reading. A “20% dump” on Bitcoin is a change in a free price. A 20-point drop on a prediction market is a 20-point change in a collective estimate. The candle looks the same. The thing it is measuring is not.

Prediction-market OHLC is the field guide for those four numbers when they sit on a chance. This page is the contrast with the charts most traders already know.

Crypto is unbounded above. It is bounded below at zero in theory and not in any way that helps you next week. A prediction market is bounded on both sides, always, and everyone in the book knows it.

That bound is why blow-off tops look different. A 95% market cannot do what a crypto blow-off does. The remaining five points are not a runway. They are the last argument, and they are usually the most expensive points in the whole history of the contract because the payout is so close.

It is also why oversold readings near 8% are not the gift they look like on RSI. The market can stay at 8% until it is 0. There is no “it can’t go lower” the way a beaten-down alt sometimes can’t. It can go to zero on purpose. That is one of the two legal endings.

Crypto does not resolve. It just keeps printing. A prediction market has a clock. As that clock runs down, the destination starts to dominate the path. Patterns that were informative three months out become noise the night before settlement, because the people still in the book are no longer expressing a long-horizon estimate. They are getting flat, squeezing a side, or holding to settlement.

Crypto TA that ignores the calendar is often still fine. Prediction-market TA that ignores the calendar is a method for a different instrument. The technical-analysis guide treats time-to-resolution as a first-class input, not a footnote.

On crypto, a long wick is a liquidity event: stops run, a market order hits a thin book, a perp cascade. On Polymarket a long wick is a change of mind that did not stick. Someone paid 70% and the book would not stay there. The mechanism can still be a thin book — plenty of Polymarket events are thin — but the meaning is an estimate being tested, not a spot price being raided.

That is why a probability line is such a bad default. The wick is the test. A line deletes the test and keeps the settlement of the hour. Crypto traders would not accept that deletion on ETH. There is no good reason to accept it on an election contract.

Hyperliquid, which Predm also reads, is the other half of this contrast. It is a perpetual-futures book. Its candles behave like the crypto ones. Putting a Polymarket event and an HYPE perp on the same terminal is useful precisely because they are not the same kind of chart, and you can see both without pretending they are.

Crypto perps trade all day. The candle at 04:00 UTC is as legal as the one at 14:00. Prediction markets have a human rhythm. Pew’s 2026 look at Polymarket flow found activity concentrating into European and U.S. business hours as the user base matured. A wick printed at 03:00 on a political market is often one person. A wick printed at 15:00 is more likely a crowd.

That does not make the 03:00 wick fake. It makes it a different object. Crypto traders already know to discount the Sunday overnight. Prediction-market traders have to learn the same discount, on a book that is thinner to begin with.

Volume is the tell. A quiet interval should not get a loud interpretation, no matter how dramatic the wick. Predm will not draw a candle on a market that cannot fill one. That rule is the practical version of this paragraph.

Steal: the habit of reading range before direction. The habit of checking more than one timeframe. The refusal to take a close-only chart seriously. The respect for prior highs and lows as places the book has already been.

Leave: unbounded trend systems, blow-off templates, “it can’t go lower,” and any indicator you only trust because it worked on a two-year BTC long. Also leave the idea that the order book on the other side is irrelevant. On crypto that can be a workable simplification. On a prediction market the other side is the estimate. Concentration per side is the extra pane crypto never asked you to open.

Two venues are live. Polymarket events get reconstructed OHLC, holder concentration, and the multi-timeframe board. Hyperliquid perps get the same terminal layout on a book that is already a price, so the reconstruction step is unnecessary and the candles are the venue’s own.

Aster, Binance, Kraken and Kalshi are marked coming soon on the homepage. They are not live, and nothing in this guide should be read as a chart of them.

You cannot trade in Predm. The terminal is read-only. How to chart Polymarket is the practical walkthrough if you want the rebuild itself, not the comparison.

Is a Polymarket candle the same as a Binance candle?

Same schema — open, high, low, close over an interval. Different meaning. One is a probability that will be forced to 0 or 1. The other is an unbounded price. Predm does not read Binance today; the comparison is structural, not a product claim.

Why put Hyperliquid and Polymarket in one terminal then?

Because a lot of the same people trade both, and because a news event that moves a perp often moves a related prediction market. The charts stay honest about what they are. They just live in one place.

Can I use the same stop-loss rules?

A 5% stop on Bitcoin and a 5-point stop on a 60% event are not the same risk. The event can gap to zero on a headline in a way a large-cap perp usually does not. Size the stop for a binary, not for a coin.

Do crypto-style chart patterns work on Polymarket?

Sometimes, in the middle of the axis, far from resolution, on a liquid event. Rarely near 0 or 100, and less and less as the date approaches. The TA guide is the longer, more cautious answer.