[modeller_usage] about random seeds

Hello: I am using the random seeds when I building model these days. and I found that in the following line:

environ(rand_seed=-8123, restyp_lib_file='$(LIB)/restyp.lib', copy=None)

I try three jobs with rand_sees=-500, -1000, -1500

and I found that the -1500 get the lowest DOPE score and -500 get the highest DOPE score for the best solution. Does it means that if I set rand_sees=-2000 would get a better one?

Thank you very much