Sounds perfectly rational to me: the host does not tell the participant that they don't know which is the winning cup. The participant then assumes, quite rati

Sounds perfectly rational to me: the host does not tell the participant that they don't know which is the winning cup. The participant then assumes, quite rationally, that they are being tricked into losing. This is a really poor experiment, to be honest, because it is measuring the host's appearance of honesty, not the risk appetite of the subject.