We consider mixing of the antidecuplet with three J P = 1/2 + octets (the ground-state octet, the octet containing N (1440), Λ(1600), Σ(1660) and Ξ(1690) and the octet containing N (1710), Λ(1800), Σ(1880) and Ξ(1950)) in the framework of approximate flavor SU(3) symmetry. We give general expressions for the partial decay widths of all members of the antidecuplet as functions of the two mixing angles. Identifying N 10 with the N (1670) observed by the GRAAL experiment, we show that the considered mixing scenario can accommodate all present experimental and phenomenological information on the Θ + and N 10 decays: Θ + could be as narrow as 1 MeV; the N 10 → N η decay is sizable, while the N 10 → N π decay is suppressed and the N 10 → Λ K decay is possibly suppressed. Constraining the mixing angles by the N 10 decays, we make definite predictions for the Σ 10 decays. We point out that Σ 10 with mass near 1770 MeV could be searched for in the available data on K S p invariant mass spectrum, which already revealed the Θ + peak. It is important to experimentally verify the decay properties of Σ(1770) because its mass and J P = 1/2 + make it an attractive candidate for Σ 10 .