I was thinking more about the NBA martingale system. Really every team in the NBA will eventually win games. The important part is getting a large enough sample size and controlling the size of the bet made.
What about looking at road trips, say 5 games in length or more. What are the odds of a team winning all five in a row? Could a reverse martingale system be done to bet on a team to lose, i.e. picking the whichever team the subject is playing against? There is probably huge levels of attrition on road trips, in the sense that teams get more and more fatigued with each effort required to win games on the road. What about a bad team on the road? The worst teams in the league have long road trips too. The odds of them sweeping a road trip must be unfathomable.
Now that is a research subject.
I think the main reason I haven't tried to research it more is not a lack of time but a lack of opportunity to bet the games and a small amount of disinterest.