# Lab 2 grading rubric (written before grading, 2026-10-19)

Unit: one query x one provider, judged on that provider's top 3 results
(title, URL, returned snippet). 20 queries: the first 4 of each category.

- 2 = at least one top-3 result directly answers the query from a primary or
      authoritative source (vendor pricing/docs/changelog, spec, official blog),
      current as of the test date.
- 1 = the answer is reachable in the top 3, but only via a secondary source
      (blog, aggregator, forum, video) or a page that is outdated / partly wrong.
- 0 = no top-3 result answers the query (off-topic, stale, or only tangential).

News queries: "current" means the item covers an event from the last ~60 days.
Graded by Jonathan Hildebrandt from each result's title, URL and returned
snippet, with linked pages opened where the snippet was unclear.

Amendment before grading the "compare" block (2026-10-19): comparison queries
have no single primary source. For them, 2 = a current (2026) comparison that
addresses the specific question asked (team size, use case, volume), or the
official pricing/compare pages that settle it; 1 = generic, outdated, or
forum/video-only; 0 = off-topic. Vendor-written comparisons count but are
flagged as vendor-written in the notes.