The romans had a clear succession rule: The senate gave a bundle of special rights to an aristocrat, which makes him a princeps. The people of Rome and the legions acclaim. So much for theory.
Egon Flaig ("Den Kaiser herausfordern") calls them the three aceptance groups. You need the support of all three groups best case, even if the support of the legions would be enough for a start. So the senate could not enforce its decision, if not enough senatorial legates of legions are supporting the senate. Remember, legates were politicians and vice versa.
Unfortunately, there was not ONE army group or instance, which was strong enough to decide unopposedly. I guess one way to stabilize the roman system is to install such an instance, e.g. an huge and powerful central field army. But now, there is almost no way to avoid, that the leader of this army is the successor or even usurps himself, if he does not agree with the emperor or does not want to wait anymore.
The relatively stable adoptive system in the 2nd century was something similar. Nerva was a poor senator, and adopted Trajan, the strongest armyleader immediately. He was old, so Trajan could wait easily. Fortunately no other army group disagreed. Trajan had no co-emperor afaik, but after his death his wife immediately claimed, that he had adopted Hadrian, the leader of the biggest army, lately. Again lucky Rome. Antoninus and Aurelius made everything right now, they always had a designated successor and co-emperor.
Commodus is seen as a bad emperor, but that does not matter. He had no co-emperor and so the roman luck ended.
However as the late empire show, having always a co-emperor does not prevent usurpations. But even if the list of usurpers in the 4th century is long, successful usurpations against the prime Augustus were pretty seldom.