Abstract

We observed the pattern of voting in the first round of 72 episodes of the UK version of the TV quiz-show „The Weakest Link‟ (WL). The first round culminated as each of the nine contestants carried out an eight-alternative-forced-choice task by voting for one of their peers as the WL. Rudimentary probability theory was used to generate the frequencies of votes that would be expected purely due to chance for all eight relative positions of voter-to-candidate spatial relationships. The observed frequencies from the episodes were then compared to the expected pattern. Consensus, the number of contestants voting for the eventual WL, was also recorded in each of the 72 first rounds. Two main findings emerged:- i.) contestants avoided voting their direct neighbour as the WL, although the propensity to vote for a peer was not a simple function of distance per se; ii.) the „neighbour-avoidance‟ effect increased as the group consensus as to the identity of the WL decreased.