Assume failure then figure out how it happend. This is a method of risk identification.
Game development is a risky endeavor: It is a good idea to track the biggest risks, develop a mitigation plan, and create contingencies in case things go wrong. But risks aren’t always obvious. One way to help identify risks is to write a pre-mortem. I’m sure you’ve read game development post-mortems before. If you haven’t you need to, immediately! The idea behind a pre-mortem is to assume failure and imagine the various ways that could happen. Have everyone on your team go through this process to get more resolution on potential problem areas.