Start here

CryptoEdgeExplorers takes trading claims that spread online, about crypto, stocks and ETFs, and tests them with real prices, real fees and a pass rule written down before the data comes in. So far: 16 claims, 12 failed, 1 held, the rest still running.

Read these three first

1. Does Binance copy trading make the copiers money? About 1 in 5 lead traders had copiers in profit over 30 days.
2. What happens after a Telegram channel says a token is "trending"? More than half of 300 tokens fell 90% within a day.
3. Does a 200-day trend filter on Bitcoin hold up? The one that held, with its live state.

Tools

• Prediction market fees and break-even: what Polymarket, Kalshi and Binance Prediction charge, and the win rate you need after fees.

How a test works

1. Pick a claim people actually act on, and state it so data can prove it wrong.
2. Write the pass rule, data and costs down first, with a timestamp.
3. Run it on real prices with real fees, or record it forward when no history exists.
4. Publish the result with its uncertainty, and stamp it FAIL, HOLDS or PENDING.

The details are on the methodology page. Mistakes are fixed in public on the corrections page.

Who writes this

Ξmars, a pen name. See about and the disclosure, which lists what the author holds.