Successes, runs and longest runs (Q5903147): Difference between revisions

This paper is devoted to the explicit calculation of various probabilities associated with success runs in Bernoulli trials, particularly probabilities determined by (i) the number of success runs of length \(k\geq 1\), (ii) the length of the longest success run, and (iii) the number of times that the length of the longest success run is equal to \(k\geq 1\), all in n Bernoulli trials. For example, if \(p\in (0,1)\) is the probability of success on any trial, if \(q=1-p\), if \(N_ n^{(k)}\) denotes the number of success runs of length k in n trials, and if \(x=0,1,...,[n/k]\), then \[ P(N_ n^{(k)}=x)=\sum^{k-1}_{i=0}\sum \left( \begin{matrix} x_ 1+...+x_ k+x\\ x_ 1,...\quad,x_ k,x\end{matrix} \right)p^ n(\frac{q}{p})^{x_ 1+...+x_ k}, \] where the inner summation is taken over all k-tuples \((x_ 1,...,x_ k)\) of nonnegative integers such that \(x_ 1+2x_ 2+...+kx_ k=n-i-kx\). Some open problems connected with the random variable \(N_ n^{(k)}\) are also posed.