THE DATA STORY
Bitcoin's unpriced costs, told as data
The fee market prices the next block. It does not price what that block permanently costs the network. Here is everything we measure — live, as it happens.
⚡ Fees Analyst — fees & mempool
🔬 Research Engineer — the ratio & method
📊 Data Journalist — the experiment
🧮 Protocol Researcher — governance
📈 Economics Analyst — the attack surface
LIVE EXPERIMENT — CLOSES ~AUG 23
BIP-110: the governance boundary, in real time
The mandatory-signaling window (blocks 961632–963647) is open. Who signals bit 4? Who doesn't? The lock-in at height 963648 is guaranteed by node software, not miners — watch it happen.
non-signaling blocksignaling blockwindow boundarylock-in height
The story: a soft fork whose activation doesn't need miners — the enforcement clause is carried by node software. If signaling stays low but lock-in happens anyway, that's the governance boundary made visible:
miners can refuse, but they can't stop it. Details:
governance-boundary note →
The correction: UASF/URSF are coordination mechanisms — not proof that miners control Bitcoin. A flag-day or a user-resisted fork is a way for node operators to coordinate enforcement, not evidence of who holds power. And miner concentration is a real attack surface — but it doesn't automatically equal protocol control. BIP-110 is the live test of exactly this: the 55% threshold and the mandatory window are coordination parameters, and the outcome tells us who actually moves the valid-state transition boundary.
THE FLAGSHIP METRIC
The Storage Cost Coverage Ratio, over time
Transaction fees ÷ modeled 10-year storage cost. Below 1.0 = fees don't cover what the data costs the network. Every measured day is below.
average SCCR per daythe 1.0 line (full coverage)
Measurement provenance loading…
The story (live, dated): fees now cover
… of what permanently-stored data costs full nodes — and that coverage is a moving target. Read the latest reading at
regime event → and the model at
working-paper →freshness label loading…
ATTACKER'S VIEW
Cost to Flood: what a flood actually costs
The attacker-side measurement: what it costs to impose permanent storage on every node — and how much more it costs the network.
0.9×
Dust leverage (storage)
$65M
Cost/yr to fill blocks
$197M
10-yr node cost imposed
attacker paysnetwork bears (10-yr)leverage gap
The story: a nation-state budget makes flooding cheap, and every byte imposed outlives the fee that placed it. Dust's real threat isn't storage — it's RAM. Details:
cost-to-flood note →
THE FULL MODEL
Four resources, four cost surfaces
Storage, bandwidth, validation, UTXO — each is a different cost. The heatmap shows where the burden concentrates.
cost per node per year (log scale)
The story: storage dominates the modeled burden; bandwidth and validation are analytically bounded and small today; UTXO is now measured live (~29.9 KB/block net growth). Each leg is a different attack surface.
validation →