Prototyping metagames and progression systems

We’ve got lots of tools for iterating rapidly on high frequency loops at the heart of most games. Rapid prototyping and playtesting works great. However, lower frequency loops such as progression systems or metagames are harder to prototype and balance. Often issues only show up when the game is live and while spot changes are easier than ever due to analytics, structural issues may only get 1 or 2 passes over the lifetime of a game. So we typically reduce risk by just copying what we think another successful game is doing.

In the past, progression systems arguably didn’t matter as much. But now, when 30-day retention can make or break your company, these systems are now in the spotlight. How do we not mess them up? How do we innovate? What are the best practices? Are there new tools (like simulation approaches like Machinations) that could help?