For , its mean is , so the size-biased probability generating function is
Choose independent with distribution . Their shifted sum has generating function, initially for ,
with both sides continuous and zero at . Equality of probability generating functions proves
Equivalently, has possible positive summand pairs, each of probability , matching the size-biased geometric distribution probabilities.
For , differentiate the probability generating function term by term and multiply by :
This is the size-biased probability generating function identity. At , the finite-mean assumption gives the limiting value one on both sides; at zero, positive support gives value zero.
Condition first on . A Poisson distribution with parameter has conditional probability generating function , including the degenerate zero case when . The law of total expectation gives
For the inner argument is in , so this use of the generating function is justified directly. The result is a compound Poisson distribution with a Poisson distribution of unit parameter for each jump, also called a Neyman Type A distribution in this case. It is generally not an ordinary Poisson distribution.
Differentiate the probability generating function equation:
At , , giving
The derivative is finite because and the explicit square root has no singularity at . Alternatively, the expected population in generation is , so the expected number of splits in that generation is . Summing this nonnegative geometric series gives the same answer without initially assuming a finite expected value.
If the initial individual has no offspring, . If it splits, count that one event and the splitting events in its two independent descendant trees. Thus
where are independent copies of . Taking the probability generating function gives
For , the quadratic equation has two algebraic branches. A probability generating function must remain finite at , where it equals , so the admissible branch is
The rationalized formula is also defined directly at zero. At , , so , consistent with almost-sure finiteness. Extinction occurs after a finite number of generations almost surely, and each generation is finite, which justifies that finiteness claim for the number of splitting events in a binary branching process.
Set , so . Differentiate once and evaluate at , using , , :
A second differentiation gives , hence
The derivatives of a probability generating function give and . Consequently . The earlier geometric tail bound from a uniform escape probability ensures the needed moments are finite.
Write . The initial possibilities are with probability , after which a fresh wait begins; with probability , after which a fresh wait begins; or with probability , which ends the experiment. Accounting for the flips already made in each case gives the probability generating function equation
Therefore for . In particular, and there is no mass below two flips, as required.
By the definition of a probability generating function and the power series for the exponential function,
The series is absolutely convergent for every finite real ; in particular the expression is finite for all , as needed for the later tail estimate.
Insert the Poisson distribution probability generating function into the preceding bound:
For fixed , differentiate the exponent with respect to . Its derivative is , and its second derivative is , so its unique minimum occurs at . At that value,
Consequently
This Chernoff bound remains valid for noninteger positive thresholds ; the event is then the same as exceeding or equalling its integer ceiling.
Because , the function is increasing and nonnegative. Thus . Applying Markov's inequality to the nonnegative random variable , whose expected value is finite by assumption, yields
Under the usual probability generating function convention takes nonnegative integer values. The argument itself works for any real random variable if is understood as the finite expected value .
For a nonnegative integer-valued random variable, its probability generating function is
The series converges absolutely at least for , with in the term; it may converge on a larger domain.
The extra ancestor multiplies the probability generating function by . The preceding probability generating function of an independent random sum formula therefore yields
For the two roots are . A probability generating function lies in there, so the plus root is impossible. Also gives . Consequently the total progeny with critical geometric offspring has
At the value is , consistent with almost-sure finiteness. For complex , the same expression uses the analytic square root that equals at .
Differentiate the probability generating function inside its radius of convergence:
The first moment and second factorial moment are and . Consequently
Set . The preceding probability generating function becomes , identifying
Equivalently, for , . Since , also . This is the existing geometric count under Bernoulli thinning, with the zero-based convention specified.
The sum of independent Bernoulli random variables with possibly unequal success probabilities has a Poisson binomial distribution. Its probability generating function is : expand the product to read off the probability mass function. Linearity of expectation and additivity of variance for independent random variables give and . Equal recover the binomial distribution.
For three trials with success probabilities , expanding gives masses at . Their expected value is and variance is . Replacing all three success probabilities by their average preserves the expected value but gives binomial distribution variance , so it does not preserve the law.