PolySignals.net Turns Prediction Market Prices Into Live Crypto Readings
A free Cryptohopper site reads Polymarket, Kalshi and options exchanges for a live Buy, Sell or Hold reading on every listed coin, and sends a daily trend call to Cryptohopper traders.
Written by CoinArticle’s AI Newsroom · from 2 cited sources. How we work
Reporting: PolySignals.net — this story is based on a single source so far.
Key Takeaways
- 1The advantage over holding is statistically marginal, with t-statistics of 1.4 to 1.9, below the usual threshold of 2.
- 2The rule was chosen from several variants tested on the same three coins.
- 3Drawdowns reached 52% to 87% across the nine coins tested over three and a half years or more.
- 4It lagged simply holding the coins in bull years, including 2023 and 2024 for all three.
PolySignals.net, a free site built and run by Cryptohopper, publishes a live Buy, Sell or Hold reading for every coin listed on crypto prediction markets and options exchanges, along with a once-a-day trend call that it sends to Cryptohopper as a trading signal. Reading the site, its API and its AI-assistant server requires no account and no key.
Disclosure: PolySignals.net and CoinArticle are both run by Cryptohopper.
What the readings are built from
Prediction markets such as Polymarket, Kalshi, Limitless and Gemini list short "Up or Down" contracts that ask a simple question: will a coin close the current window above the price it opened at? PolySignals.net reads the price of the "Up" side of those contracts, together with option prices from Binance, Deribit, Bybit and OKX, for five windows: 5 minutes, 15 minutes, 1 hour, 4 hours and 1 day.
Each price becomes a score from -100 to +100, calculated as 200 times the market's probability of "Up" minus one half, and a Buy, Sell or Hold label. Venues quoting the same window are combined into one reading per coin, and each venue's own reading is shown next to it. A confidence score from 0 to 100 reflects how deep the markets are, how closely the venues agree and how much of the window is left. The venues are read every 30 seconds.
The coin list is not fixed. The site covers whichever coins the venues list at the time, which has included Bitcoin, Ether, Solana, XRP, Dogecoin, Hyperliquid's HYPE and BNB.
A readout, not a forecast
The site's methodology is unusually direct about what the live score can and cannot do. Once a window has opened, the market price mostly restates how far the coin has already moved since the open. A two-parameter model fed only that move and the volatility expected over the time remaining reproduces 73% to 97% of the price, according to the documentation.
In a 12-day study published on the site, no timeframe, venue or variant of following the live calls had an edge, and following them mostly paid trading fees. PolySignals describes the score as a measure of where market pricing sits right now, not a prediction of the next window and not a trading signal.
The daily trend call sent to Cryptohopper
The signal that does go to traders is a separate, slower rule that uses no prediction market data at all. Once a day at 00:05 UTC, it looks at each coin's Coinbase daily close. A coin turns long when its close rises above its 50-day average by more than one standard deviation of its last 20 daily returns, and turns flat when it falls the same distance below that average. In between, the previous call stands. The rule is long-only: flat means holding US dollars.
Since 12 October 2026, PolySignals.net has been a signaler on the Cryptohopper Marketplace, where Cryptohopper users can subscribe and have their own trading bot act on the calls. The Cryptohopper signals page lists every signal delivered. PolySignals itself never trades or holds funds; trades happen in the subscriber's own Cryptohopper account.
The site publishes the rule's historical results with the same candour as the live score. Measured on Coinbase daily closes with a 0.6% fee per fill, the rule returned 134% on Bitcoin and 150% on Ether between April 2021 and October 2026, and 276% on Solana from January 2022. Holding the coins over the same periods returned 60%, 5% and -39%. The methodology lists the caveats alongside those figures:
- The advantage over holding is statistically marginal, with t-statistics of 1.4 to 1.9, below the usual threshold of 2.
- The rule was chosen from several variants tested on the same three coins.
- Drawdowns reached 52% to 87% across the nine coins tested over three and a half years or more.
- It lagged simply holding the coins in bull years, including 2023 and 2024 for all three.
The site sums it up as "a bet that a trend premium persists, not a proven edge." Every backtest result is published on its results page.
Free API, AI-assistant server and widgets
Everything on PolySignals.net is open to developers. A JSON API needs no key and allows 50 requests per 10 seconds per IP address. An MCP server lets AI assistants query the readings, trend calls and results directly, and is listed in the official MCP Registry as net.polysignals/polysignals. The site also offers Atom feeds and embeddable widgets, and is published in English, Dutch, Spanish, French and German.
The site places no cookies and runs no analytics or advertising scripts.
How it relates to CoinArticle
The two sites measure different things. PolySignals.net reads what traders are pricing in markets. CoinArticle's News Signals summarise what the news says about each coin, weighting every story by its impact and confidence.
PolySignals' news page shows the latest CoinArticle headlines next to the live PolySignals call for each coin a story mentions, so readers can compare what the news says with what the markets price.
PolySignals.net is not affiliated with Polymarket, Kalshi, Limitless, Gemini, Binance, Deribit, Bybit, OKX or any other venue it reads. Its readings describe what markets price, not what a coin will do, and are not financial advice.
This article is for information only and is not financial advice. Read the full disclaimer.
