Gambling on losing teams is a legitimate strategy, but more than betting on losing teams I think it is evaluating the risk/reward scenario with a specific bet. Sure, there might be a team that is likely to win, but maybe it pays a penny to the dollar. Basically, winning that bet doesn’t make a difference. I get more entertainment out of betting longshots and making sure the reward is worth the risk.
If the popular team or fighter become an underdog, I think it's good to bet them because usually they have a good chance to pull an upset. But for a mediocre team or fighter, I don't think it's wise to bet them because they mostly lose, only few times they win.
I never want to bet on a huge underdog, it's just like a lottery where the chance to win is near to zero.
Betting on a huge underdog and simply hoping for an upset which you yourself don't even know how to happen or where to come from is like betting on luck. That's not going to be profitable. Betting on an underdog because of a distinct analysis could be reasonable. Betting on the huge underdog Miami heat, for example, versus the most favorite team to win the NBA championship, Boston Celtics, could have some reasonable bases. This could turn out profitable.