NanoTicker

What is the Nakamoto coefficient?

The fewest representatives that could together disrupt the Nano network. What the stall and control figures mean, and why hosting concentration counts too.

The Nakamoto coefficient is the smallest number of participants who would have to act together to disrupt a network. One is catastrophic. Higher is better, and a rising number means the network is getting harder to attack.

Where the idea comes from

Every decentralized network claims to be decentralized. The Nakamoto coefficient turns that claim into a number: count the largest participants in order of size until they add up to enough to break something, then report how many you counted.

It works across networks because it ignores how consensus runs underneath. On a proof-of-work chain it counts mining pools. On Nano it counts voting weight, since weight decides whether the network agrees.

The two thresholds NanoTicker tracks

Nano confirms a block when more than 67% of online voting weight votes for it. That one rule leaves two ways to interfere, so there are two coefficients.

  • Stall (33%) is the fewest reps holding more than a third of online weight between them. If they stop voting, the rest can no longer reach 67% and confirmations stop. This is the number that matters most. It is the cheaper attack, and nobody has to be malicious. An outage is enough.
  • Control (67%) is the fewest reps that reach quorum on their own. Together they could confirm blocks without the rest of the network agreeing.

Control is always the larger number, and stall is the one to watch. A network can be halted long before anyone can take it over.

How the figure is calculated here

NanoTicker snapshots every online representative and its weight once an hour, and computes both coefficients from that. Two details if you compare with another source:

  • Every rep in the snapshot counts. There is no cut-off at the top twenty or fifty, which would understate concentration by ignoring the tail.
  • Each rep counts for the weight it held at the time, so later growth does not rewrite last week's figure.

Weights are a share of online stake, the same basis quorum uses. Counting offline reps too would give a flattering number that describes nothing the network does.

Counting nodes is not the whole story

A coefficient of eight sounds reassuring until all eight of those reps turn out to be virtual machines at the same hosting company. Independent operators, independent decisions, and they all go offline together when that provider has a bad day.

So the Nakamoto page also breaks weight down by hosting provider, resolved from each node's IP address. Two charts sit side by side: one shows voting weight per provider, the other the number of nodes. A provider can host many small nodes holding little weight, or two nodes holding a great deal of it. Only the second is a real concentration risk.

Geography and jurisdiction deserve the same scrutiny, though NanoTicker does not measure them yet.

What moves it

The coefficient rises when weight moves from large reps to smaller ones. It falls when weight concentrates, or when a rep near the top goes offline and its weight stops counting toward the online total.

Ordinary holders move it by changing their representative. No protocol change or vote required. The health score is there to help you find a rep worth moving to.

See the live Nakamoto coefficient