edgedepth EARLY ACCESS

Percentiles and the P-chip

0.94 means nothing until you know it beats 94% of that market's own history.

Is 0.94 a big number? Wrong question.

Raw values do not travel

A reading of 0.94 might be ordinary on one market and once-a-month rare on another. A book tilt that would be shocking on bitcoin is a quiet afternoon on a thin altcoin, because the altcoin’s book swings that far all the time. Markets differ in size, volatility and habits, so their raw numbers live on different scales. Comparing raw values across markets is comparing temperatures in different units.

Rank it against its own history

The fix is to ask a different question: how does this minute’s reading rank against the same reading on the same market over its recent history? If today’s value beats 94% of that history, its percentile is 94. Now the number means the same thing everywhere: “this is unusual for this market.”

The top and bottom tenth stand out

A useful rule of thumb: readings in the top or bottom tenth of their own history are the ones worth a second look. Everything in the middle is that market being itself, and deserves no excitement. This is the fairest yardstick we know for comparing across markets, because every market is only ever measured against itself: the altcoin’s wild book is ranked against its own wildness, and bitcoin’s calm book against its own calm.

The P-chip

On EdgeDepth market pages, the small chip marked P next to a reading is exactly this: the percentile of the current value against that market’s own history. P97 means the current value beats 97% of that history; P50 means it is dead ordinary. It is how one glance tells you whether a book tilt or an open-interest surge is genuinely unusual, or just Tuesday.

See it live

Open a market page and read its P-chips.

Related: Book imbalance · Open interest