I think it's more than just that simple game:

Indeed, 4-year-olds monitor apparent differences in reliability even when no obvious errors are involved. Having watched one informant name objects accurately and another informant make non-committal remarks about them (e.g., “Let me look at that”) or express ignorance, children subsequently invested more trust in the accurate as opposed to the non-committal informant (Corriveau, Meints & Harris, 2009) or the ignorant11 informant (Koenig & Harris, 2005). Fourth, reliability monitoring can reverse the normal pattern of vertical trust. Although preschoolers trust an adult informant over a peer, that preference is reversed if the peer proves more reliable (Jaswal & Neely, 2006). Finally, selective trust is not transient. When a second test phase was administered, either 3-4 days or an entire week after the initial test phase, 3- and 4-year-olds still invested more trust in the previously correct informant (Corriveau & Harris, 2009a).

-- Harris and Corriveau 2011