THEORY-905: In simulations of the two imitation protocols, populations small enough to be finite often settle on a suboptimal option because pure imitation can remove a type entirely, and once settled the protocols give no way to re-explore when the environment changes; variants in which members compare quality estimates converge faster than blind weighted voting, at the cost of more communication and comparison

Source

Soma et al., LIT-1484, read in NOTE-1132: sections 4.2, 4.5 and 4.9.3 (pp. 12–14, 18), Figures 2b and 2d, appendix B.8.

What was actually shown

• Finite size. Ten options, 1000 seeds. N = 500 follows the replicator dynamic, and “a number of scout bees that is too small often yields convergence to a sub-optimal nest site option” (p. 13). The parameters are arbitrary and the result qualitative by the authors’ account.
• No way back. “Pure imitation can lead to situations in which a given type ceases to exist entirely in a finite population”, and the protocols “don’t provide a mechanism for escaping these strategies later on in non-stationary environments” (p. 18). Real colonies use spontaneous commitment, persistent stochastic allocation, or wandering workers.
• Speed against bandwidth. Three hybrids in which members compare qualities beat the weighted voter rule (Fig. 2d). The authors suggest blind imitation “uses minimal communication and cognition bandwidth” (p. 14), as an explanation, not a measurement.

What it does not show

A general result for finite populations; the no-escape point is argued from the protocol’s structure, not proved. Nor anything about real bees, whose nest-site process includes the mechanisms the model omits.