Audit: pmxt/v2

DownloadDocs

What this page is for: the problems with this corpus and the evidence behind them. The files are on the download page, and what the columns mean is in the format notes.

What is wrong with this data

v2's time span is solid and its market coverage is not complete.

• Coverage ratio 0.8707 against on-chain fills. The ratio is the share of the markets that actually traded in an hour, counted independently from on-chain fills, that have a book in this corpus for that same hour. 0.8707 means about 87 of every 100 do. Across its six pre-registered audit hours, 2,703 traded markets have no book, and the gaps are non-phantom in all six. The audit predicted v2 would come back clean and it did not; that falsification is on the record and is why this number is worth trusting. how the ratio is computed, and what it cannot tell you. audit instrument against on-chain fills, 2026-08-31
• On the seven hours it shares with AG6, v2 misses 3.09% of the market universe of that moment. AG6 misses 0.55% of v2's over the same hours. audit instrument, 2026-08-31
• Its per-hour market count grew about sixfold across its own span. 19,431 to 117,945. Any corpus-level average about v2 is therefore meaningless: use per-hour denominators. audit instrument, 2026-08-31
• The time span is solid. 2,835 hours with no zero-row objects. A warning that lists only faults is not a measurement, so this is stated with the rest. hour census against the published v2 SHA256SUMS.txt, with object sizes, 2026-08-31

Measured against the chain

Every figure here is measured against on-chain fills for the same hour, by an instrument that shares no input with this archive. Six hours were drawn, and which six was written down before anything was measured. Six hours is a floor, not a sample large enough to describe a corpus, so nothing here is a population rate.

Against our own capture, which scored 1.0000 on the same instrument: this corpus is at 0.8707, so roughly 13 in every 100 markets that demonstrably traded have no book here at all. That is what a low coverage number means: not that the data is imprecise, but that the market is missing, and nothing downstream can recover a book that was never recorded. See all four side by side.

Pre-registration 5ae4e8765d601145, draw 02a94091dea9c11f, measured 2026-08-31.

Markets held, against markets traded

A corpus holds a book for many markets that nobody traded in a given hour. That is not waste: an untraded market with a book is breadth, and it is the difference between an archive you can ask questions of and one that only answers about the busy markets. This counts both sides for each drawn hour.

This corpus does not have one number. Across the drawn hours it runs from 5.1x to 33.7x, and in date order the direction is plain, so a median of 18.7x would describe a state it was never in for long. It is shown by hour, oldest first.

Our own capture holds about 52 books for every market that traded; this one holds about 19, roughly 2.8 times fewer. What it holds can still be trusted: this is a statement about how much of the market it watched, not about whether what it recorded is right. See all four side by side.

Breadth is not completeness and a bigger number is not automatically better: this says nothing about whether the markets that did trade have books, which is what the coverage ratio above measures. The two share no arithmetic, and they rank the four corpora in the same order.