Does anyone really believe that a 51.8% win rate for longshot bets in 'Military and Defense' topics is some kind of stable hierarchy?
This is simply noise, not a controllable variable for future actions.
One cannot assume these rates remain constant; think of how a sudden shift in international posture could alter everything.
For example, if a major peace treaty were signed, the past win rates based on conflict would become irrelevant for new bets.
Focus on the actual mechanisms of each specific bet, not broad, unreliable classifications.