Fig. 1
From: Information cascade final size distributions derived from urn models
Left: In the urn model, a trial of extracting and returning a ball is repeatedly conducted until p reaches zero. p is increased by \(f-1\) if a white ball is extracted and a black ball is returned with probability \(\lambda\). In all other cases, p is decremented by one after each trial. In the figure, “\(p\leftarrow p+f-1\)” denotes that the new value of p is \(p+f-1\). Right: One of the cases of \(B_{\tau }=5\). Event \(p=0\) occurs at the end of one of the intervals \(I_0, I_1, I_2, \ldots\)