Genuinely one of the most disgusting abuse of statistics I have ever seen. The first two populations are set at 30:1 against the unit. Drawing values per million from those has a relative sample size of n = 30. The trans population is set at .05 : 1 or 1:20 against the unit.