The Andromeda galaxy is observed to have a system of two large dwarf ellipticals and ∼13 smaller satellite galaxies that are currently co-rotating in a thin plane, in addition to 2 counterrotating satellite galaxies. We explored the consistency of those observations with a scenario where the majority of the co-rotating satellite galaxies originated from a subhalo group, where NGC 205 was the host and the satellite galaxies occupied dark matter sub-subhalos. We ran Nbody simulations of a close encounter between NGC 205 and M31. In the simulations, NGC 205 was surrounded by massless particles to statistically sample the distribution of the subsubhalos expected in a subhalo that has a mass similar to NGC 205. We made Monte Carlo samplings and found that, using a set of reference parameters, the probability of producing a thinner distribution of sub-subhalos than the observed NGC 205 + 15 smaller satellites (thus including the 2 counter-rotators, but excluding M32) increased from < 10 −8 for the initial distribution to ∼ 10 −2 at pericentre. The probability of the simulated sub-subhalos occupying the locations of the observed co-rotating satellites in the line of sight velocity versus projected on-sky distance plane is at most 2 × 10 −3 for 11 out of 13 satellites. Increasing the mass of M31 and the extent of the initial distribution of sub-subhalos gives a maximum probability of 4 × 10 −3 for all 13 co-rotating satellites, but the probability of producing the thinness would drop to ∼ 10 −3 .