Plain language

The words,
without the fog.

Definitions tied to what we actually measure.

Research vocabulary

How much prices moved around in a measured period, not which direction they will move next.

How closely two measured series moved together. It does not prove one caused the other.

Outstanding futures contracts. Rising OI does not by itself mean buyers are winning; each contract has two sides.

Amount traded. Trading volume is not the amount of new money entering a coin.

Recorded exchange executions, not a count of unique traders.

The median coin in the comparison set, rather than a capitalization-weighted market index.

A later period reserved for a specified check. Once inspected, it is no longer fresh for new choices.

Looking only at assets that survived while ignoring those that disappeared.

A group selected under a stated rule, using information available at the selection time.

A possible shift in activity between groups. Price/volume/OI patterns do not trace actual transfers.

The lead-up, matched chart window, and its later continuation.

Whether a signal ranked later outcomes correctly across coins. A small positive value is not automatically tradable.

A check on whether an observed difference is unusually large under a reference model, not a guarantee about the future.