Section Mode is its own surface as of this morning. Since the Mock Exam went to public beta a year ago, Section has existed inside the Mock surface as a thin wrapper around a single section of a Full Exam — same items, same fixed pool, two of three sections removed. That worked. It was not what Section needed to be.
Section Mode now lives at /section, alongside the diagnostic, Practice, and the Mock. It carries a configurable scope, adaptive item selection inside the section, and a conditioning frame the Full Mock deliberately does not have. This is what changed and why.
What Section is for, now that it has one job
The Mock Exam is a test-day rehearsal — full sections, fixed pool, faithful to the live test's curation. The diagnostic is a 25-minute measurement that returns a banded score. Practice is open-ended adaptive sessions against the candidate's full mastery vector. The candidate who has been on the platform for three months and has six weeks left before sitting the test has a fourth job none of those three are right for: condition a specific section under pressure.
Specifically: a candidate whose Verbal pacing breaks in the back third of the section, who wants to run the back-third pattern repeatedly against the engine, under the timer, without burning a full Mock attempt each time. Or a candidate whose Data Insights accuracy drops on Multi-Source Reasoning items specifically, who wants twenty minutes of MSR-only items at full section pacing. Or a candidate who has six days until the test and wants a forty-five-minute Quant section, adaptive, today, every day, with no full-mock setup.
What it is
Three choices, in this order. Length: 20, 30, or 45 minutes (the live section's full budget). Scope: the whole section, or a single sub-skill cluster surfaced from the candidate's mastery vector. Selection: adaptive — the v1.1 engine with mastery routing runs inside the section — or fixed pool, the same curated behaviour the Full Mock uses.
The default is adaptive whole-section at the live section's full length. That setting is the surface's original use case — the candidate who wants a section under pressure, against the engine, against their current state. Every other combination exists for the cases the default does not cover.
What this changes inside the surface
Three behaviours are new.
Adaptive selection inside the section. The Mock's selection is fixed-pool by design — the live test is curated, and we mirror it. Section Mode runs the v1.1 engine with mastery-vector routing inside the section, choosing the next item the way Practice does, but under the section's full pacing constraint. A candidate who is weak on Combinatorics gets routed to Combinatorics items at section pacing, which is the rehearsal experience the candidate is actually asking for.
Sub-skill scope. The "scope" picker lists clusters the candidate's mastery vector has identified as worth attention — usually three or four clusters per section, refreshed after each session. A Verbal Section Mode run with scope set to "Critical Reasoning · Assumption + Flaw" serves CR items from those two families exclusively, at section pacing, with the engine routing within them. The cluster list is not the full topic taxonomy; it is the candidate's actual gaps.
The Mock-style fallback. A candidate who wants Section as it existed before — a single section of a Full Mock — picks "Fixed pool" and the runner serves the curated pool the Full Mock would. Section Mode does not take a feature away from candidates who liked it the old way.
What we cut
Two features that lived in the build through the autumn and did not ship today.
Sub-section pacing splits. Some candidates asked for the section split into thirds — a twenty-five-minute "front and middle of Verbal" run, separately. We did not ship it because the live test does not let a candidate skip sections of a section, and a feature that makes the rehearsal easier than the room is not a feature this surface should carry. The pacing analytics inside the section already tell the candidate which third is breaking; running just that third does not condition the section, it conditions a fragment.
A streak / daily-conditioning counter. Same argument as Practice and Roadmap. Section Mode reads density from response history, not consecutive- day pressure. A counter rewards the wrong shape of conditioning.
How the surfaces fit together now
Four surfaces, four jobs.
Diagnostic. Measurement, banded, returns a starting score and a sub-skill gap list. Practice. Open-ended adaptive sessions against the full mastery vector, no scoring band. Section Mode. One section under pressure, adaptive or curated, scope configurable. Mock Exam. Full test-day rehearsal, fixed pool, three sections.
A candidate working through a Pro window typically moves between the four in roughly that order over the course of the window — diagnostic at the start, Practice as the routine, Section Mode for the problem section once one becomes visible, Mock Exams at the milestones. The Roadmap, when set, sequences across the four.
Who gets it
Section Mode is available today on every Pro tier and Ultra. Pro 1-month gets it too — Section Mode is the surface a Pro 1-month candidate has the time to use well, even when the Library does not yet fit in a thirty-day window. Free diagnostic candidates do not see Section Mode; the surface depends on a current mastery vector and a scope list, both of which require the engine context only Pro provides.
The next note will be a small one — a calculator update we have been writing for the runner, in time for the autumn-quarter test sittings.
— Brightroom Product