If two sub groups of a distribution get a net positive in % that necessarily means the other part went net negative in the same proportion, seems a bit too far

If two sub groups of a distribution get a net positive in % that necessarily means the other part went net negative in the same proportion, seems a bit too far to argue against math just to defend the rich.