MDPE.16:3.7 - Choose trials that distinguish the alternatives

Select a trial or observation only when a materially uncertain branch disposition or development decision needs its result, the full contribution warrants cost and delay, and performers, capability, access, rights, resources and an opportunity window make it obtainable. Keep a live, archive-only or stepping-stone branch on sufficient available grounds without designing a trial merely to finish its record. For a selected trial, prefer a contrast that exposes different failure modes over repetition of the representation that selected the current winner.

Useful contrasts include actual performers versus notation or model scores alone; varied partners, bodies, instruments, tempi, phrases, venues, audiences, or media; held-out contexts when retrieval and transfer matter; source and receiving branches when transmission matters; and changed cue, timing, latency, noise, attention, fatigue, or operational conditions.

A selected trial belongs to its direct Method. Keep branch availability, trial plan, performed trial, observation and resulting disposition distinct. At the set level, choose to explore, stabilize, redirect, suspend or retire a branch from the current question on the basis that supports that decision. A trial plan is not a performed trial, and an availability result is not proof of the later cultural or performance outcome.