Recruitment: target players vs convenience testers, sample size per round

(5-8 finds most issues; more for balancing), screening criteria, consent and recording forms, remote vs in-lab setup and hardware capture plan.