Need help with classification give each possible outcome has individual features

Follow the full discussion on Reddit.
I am working on a prediction model for mma fights and am wondering how I am going to train my data. The issue I am having is that I am going to pump in the statistics for both fighters ie their striking accuracy, their wins, ect. Other models I've seen handle this have had two columns for each fighter ie fighter 1 accuracy, fighter 2 accuracy. Thing is I'd like my model to be symmetric. So if I make Jose Aldo fighter 1 and Frakie Edgar fighter 2, I'd like it to pop out the same probability outcome if I make Jose Also fighter 2 and Frankie Edgar fighter 1. I guess what I'm saying is I'd like to force the weights of similar stats to be the same so that there isn't any issue with switching around fighters if that comes up in the data.