I'd like us to approach feature parity with GAMS: https://www.gams.com/48/docs/T_LIBINCLUDE_MOO.html#LIBINCLUDE_MOO_Methods_RandomWeighting