Inputs capture effort, such as minutes practiced or pages read. Processes describe how you practiced, like repetitions, varied contexts, or spacing. Outcomes show capability under constraints: accuracy at target speed, quiz scores without aids, or shipping a feature. Triangulating prevents single-metric blindness.
When a measure becomes the only prize, people game it and learning suffers. Guard against proxy traps with paired quality checks, random audits, and narrative notes. Prefer rate-of-improvement over totals, and weight freshness, difficulty, and context to keep incentives humane and constructive.
Before the first cycle, capture a baseline under realistic constraints, then retest on a fixed cadence. Normalize scales where helpful, track variance, and flag ceiling or floor effects. Benchmarks guide ambition, while personal bests protect joy by highlighting meaningful, individual progress.