Consider the quality of engineer of the departed team that allowed Twitter to run without incidents for so long. Also, change is the most frequent cause of incidents, as no one was pushing new changes there were fewer chances of an incident.