Asymmetric results in a constructive simulation

[Andre Kuroswiski]

Hello, I am working with Mixer in constructive simulations of BVR Combat for my Masters and, after some verification stress tests, I found some asymmetrics results.
For example, in 800 executions of an scenario, the aircraft comming from East won 15% more combats than another comming from West and changing the hemisphere the results are the oposity!!!
I know that its could be correlated to my specific implementations , but I would like to know if someone had similar problems and could give some ideia?

I got stuck for almost a month trying to identify the reason, I am using JSBsim standart F-16 and checked Coriolis Effect, Tracks calculations and Earth Models, reaching some changes in the results, but its didn't solve the assimetry.