Measured findings

block.space's measurements document collects "numbers that would make a chart", under the heading "Measured, not assumed". Its figures are from the July 2026 measurement run, and it says that "the Meter's live pages supersede them where they differ". This page gives the run's five findings with that date, then the live readings.

The document's own advice on reading them: ratios (B/A, C/B, bytes:vbytes) "are the durable findings — they hold as measurement advances. Shares of chain totals will move."

The July 2026 run #

1. The chain became a data medium, then partly stopped #

Data payload as a share of all block bytes, at five heights:

The document reads the fall as runes replacing inscription bodies with 20-byte runestones: "a protocol getting more efficient reduces the chain's data share while increasing the number of tokens."

2. The witness subsidy, per protocol #

Bytes divided by vbytes: 3.374 for inscriptions, 1.008 for stamps, 1.0000 for runestones and 1.0142 for Omni. "A 3.37× gap between two protocols using the same chain in the same year." What this means is on the witness discount.

3. The free-ride index #

B/A, the share of a token's inclusive book that is its own measured structure. Open Assets: 0.72. CoinSpark: three regimes, 1.00, then 0.91, then 0.09, with a per-claim floor of 0.026, "the most extreme free-ride measurement recorded". See carriage and the free-ride index.

4. EPOBC, a token with zero carrier bytes #

EPOBC encoded its colour entirely in the input-0 nSequence field, which every transaction already has. Its free-carrier fraction was 1.0: the marginal bytes attributable to the colouring are none. Under Book B it claims essentially nothing; under Book A it claims whole transactions. "The two books give opposite answers, and both are correct."

5. OLGA, a protocol upgrade #

Stamps' switch from bare multisig to chunked P2WSH made encoding 1.84× cheaper per content byte and 2.96× cheaper per file byte. Mean file size rose from 366 B to 3,421 B, 9.3× larger, and adoption reached 98.3 % within 2,500 blocks. The document's reading: "users spent the savings on more content rather than banking them."

The live meter now #

The books held (as of block 972,299, 2026-10-20 09:30:48 UTC):

The census, nearer the tip, had counted 777.05 GB over 972,307 blocks (as of block 972,306, 2026-10-20 11:06:05 UTC).

The largest families by Book A, with their share of the chain's bytes (as of block 972,299, 2026-10-20 09:30:48 UTC):

These are Book A shares, the bytes of transactions bought for tokens. The run's first finding measured data payload, a narrower thing, so the two series are not the same measure and should not be read as one.

What has not been measured #

Blockspace Value Theory proposes a test of its own thesis: across many assets, with inscription number, collection, format and mint era controlled for, price per kilobyte should converge. In its words, "It has not been measured." BVT is a living document and an untested thesis.

Related #

• The witness discount
• Carriage and the free-ride index
• The three books
• Blockspace
• The frontier and "as of"

Sources #

1. block.space docs: measurements · block.space/docs/measurements
2. block.space docs: Blockspace Value Theory · block.space/docs/bvt
3. block.space meter and summary APIs · block.space/api/summary
4. block.space protocols API · block.space/api/protocols
5. block.space live pages (/, /rankings, /leaderboard, /largest-inscriptions, /heaviest-tokens, /collections, /sync) · block.space