This section constructs the probabilistic tools of the double-spend analysis in Sections 4 to 6: exponential clocks, which model block discovery; the attribution sequence (Proposition 3.4), which reduces the race in continuous time to a sequence of Bernoulli trials; the negative binomial law of the attacker’s progress (Proposition 3.5); and the normalisation of a first-to- race (Lemma 3.6). Probability spaces, conditioning, independence, random variables and expectation are those of the Math Guide, §“Probability, asymptotics, and computation”; countable additivity, continuity along monotone events, density-defined laws and infinite sequences of independent trials are those of its §“Beyond finite spaces: countable additivity, limits, and densities”.
A random variable is exponential with rate
An exponential satisfies
Let and be independent exponentials with rates and . Then
and the winning time is exponential with rate
Integrating over the time at which fires,
For the minimum,
and similarly for . Thus the winner is independent of the winning time. □
Model the honest network and attacker as independent exponential clocks with rates and , restarting when either fires (Section 4). Successive block attributions form an infinite sequence of independent, identically distributed Bernoulli trials in the Math Guide’s sense: attacker with probability , honest network with probability , independent of all earlier attributions and all elapsed times.
By Lemma 3.3 the first block is the attacker’s with
probability
Proposition 3.4 is the paper’s reduction (its §3, with the clock rates of its footnote 1) of the race in continuous time to a sequence of trials. Whether the attacker’s branch ever overtakes the public chain depends only on the order of block discoveries, not on their times; every result below is therefore a statement about a / trial sequence, and the time constant does not appear in any of them.
In an attribution sequence with attacker probability
— the negative binomial distribution.
The event constrains the first
In the same sequence, let be the event that the honest network reaches blocks before the attacker does, and the reverse. Then and partition the sample space, and
Among the first
The exponential model is a limit, not an axiom. A miner’s work is modelled as a sequence of candidate evaluations, each an independent Bernoulli trial with a small success probability; the number of trials to the first success is geometric, and a geometric law with small success probability, rescaled by its mean, converges to the exponential of Definition 3.1. The limit requires progress-freeness: a failed candidate carries no information usable by the next one. Whether Zcash’s Equihash satisfies this is examined with the deployed parameters in Section 8.3.