⚖️ Scoring rubric

Category	Weight	Observable questions
Quiz depth	25%	Types, banks, randomisation, media, logic, controls, reuse
Author and learner UX	20%	Authoring steps, clarity, accessibility, responsive behaviour, publishing
Logic, results and reporting	20%	Scoring, feedback, analytics, privacy, assessment workflow
Maintenance and trust	15%	Update recency, installs, rating, support resolution, current compatibility
Integrations and data ownership	10%	Hooks, services, LMS/commerce, portability, storage boundary
Value	10%	Free benchmark coverage, limits, licensing, price transparency, lock-in

Each criterion uses a 0–10 value. missing remains null; it never becomes a convenient zero. pro-gated is retained as a visible status and scores zero for this free-tier comparison. A final score exists only when all six categories have evidence.

$S=10{\sum }_i{w}_i{c}_i$

A provisional known-weight score normalises only the known weights and is never used as a final rank. Ties resolve by fewer unknown and paid gates, then quiz depth, logic/results, maintenance, benchmark completion, lower transfer weight and finally slug.

See the raw rubric and evidence tiers.