What we can say is that, in the engagement scenarios we believe are likely, certain pilots are better than others. And by better I mean that the opponent dies.

What we can say is that, in the engagement scenarios we believe are likely, certain pilots are better than others. And by better I mean that the opponent dies. It is true that a computer will sometimes come up with stronger responses, even against strong human players, but it usually does this by choosing certain death for some of the pilots. While human pilots do accept certain death directions in combat (for example, to defend a high value asset like a trip transport), that still likely means that the planner failed. It's going to be interesting when we have robot planes; not only can they pull 10Gs for >10s, they can do it very quickly, and you can commit them to high risk engagements.