Freezing development for five first-time playtests is a smart way to protect the signal. The most useful extra question may be: at what exact moment did you first feel unsure what to do next? That timestamp can separate an onboarding problem from a progression problem.