Verified Aug 18, 20264 min read

Understand results

Verify scores, constraint evidence, pair meetings, assignments, and exports.

Open the intended result

Results History belongs to the current saved scenario. Each result can be opened, renamed, expanded, exported, selected for comparison, or deleted.

Latest is the newest result.
Active is the result currently loaded in the workspace.
Best is the lowest-cost snapshot-backed result comparable with the latest result’s configuration.
Different Config means the saved result used a different setup. Open the badge to see the differences or select Restore this result's configuration as new scenario.

A saved result keeps its own scenario snapshot when one was recorded. Restoring that snapshot creates another scenario; it does not overwrite the current one.

Read Score Breakdown

Expand Score Breakdown to see six cards.

Card	Meaning
Cost Score	Total lower-bound-normalized cost from contact gap, repeat gap, Balance Attributes, and other weighted preferences. Lower is better for the same configuration.
Unique Contacts	Number of distinct participant pairs that met at least once, shown against a diagnostic upper bound.
Avg Contacts / Person	`2 × Unique Contacts ÷ people`, shown against the corresponding diagnostic bound.
Repetition Penalty	Weighted repeat cost above a conservative diagnostic lower bound.
Balance Attributes	Weighted penalty from Balance Attributes preferences.
Constraint Penalty	Aggregate weighted cost from the other preferences.

The contact and repeat bounds are diagnostics, not proofs that the displayed bound is attainable. A ? means the bound is still loading. If its calculation fails, the card tooltip explains the fallback; the assignments remain available.

Compare Cost Scores only when scenario snapshot, constraints, objective weights, penalty functions, and solver semantics match. If Different Config appears, compare the configuration before treating the lower number as the better schedule.

Verify Constraint Compliance

Expand Constraint Compliance.
Expand a constraint family.
Expand an individual card with Toggle details.

The family row shows Kept X / Y and the raw violation count. Details depend on the rule:

Fixed Placements: required and assigned group by session;
Keep Together and Keep Apart: split or colliding assignments;
Encounters: selected people, actual count, target, and sessions;
Balance Attributes: session, group, value, actual count, and target count;
Attribute Totals: group total, bound, and raw-unit deviation;
Group Visits: person, group or group set, visits, target, and raw visit deviation.

No violations on a preference means its target was met. A preference violation is a valid result with added cost. A Fixed Placements, Keep Together, or Keep Apart violation should be reported.

Check meetings and assignments

Use the pair-meeting histogram to see how many pairs met 0, 1, 2, or more times. Use the matrix to inspect one pair. These views count meetings; they do not identify the rule that caused a repeat.

When Avoid Repeats is active, the pair matrix also shows each pair's objective cost, configured threshold, and penalty function.

Under Your groups, switch among Grid, List, Text, Lines, CSV, SGP, and Visualize when available. Check that every attending person appears once per relevant session and that assigned counts fit capacities.

Create constraints enters selection mode. Select people, group names, or session headings, then choose Pin selected to add Fixed Placements to the current scenario. This changes the scenario; it does not edit the saved result.

Compare or edit results

Select at least two Results History cards to open Result Comparison. Review solver configuration rows first. The dialog labels the lowest raw Cost Score Best Result and the highest Worst Result even when the selected configurations differ; ignore those quality labels for unlike configurations.

The Manual Editor starts in Warn mode:

Move an assignment and review the change report.
Use Strict to block invalid moves, Warn to confirm them, or Free to apply them directly.
Check score, contact, and compliance changes; use Undo or Redo as needed.
Select Save as new result.

Pull new people and Pull constraints update the draft from the current scenario. The original saved result remains unchanged.

Open Share & Export.

Quick use: copy the schedule table or participant itineraries, print, or save the active visualization as PNG.
Structured files: download the result snapshot, Zoom breakout CSV, spreadsheet-ready schedule, or raw schedule rows.
Audience-ready downloads: download session rosters or participant itineraries.

For Create share link…, review every inclusion option: Participant names, Group and session names, Attributes, Objectives and constraints, Solver settings, Solution schedule and metrics, Detailed solver telemetry, and Selected result view. The link contains those selected fields. Anyone with the link can open it; if Require password to open is enabled, they also need the password.

Zoom export uses zoom_email, email, Email Address, or Email. Assignments without an address are skipped and the notification reports the count. Multiple sessions download as a ZIP with one Zoom CSV per session.

Open in GroupMixer

These links use registered product destinations. Result-specific links remain unavailable until a saved result is selected.

Results historyReview saved results available in this browser.Current resultOpen one saved result and inspect its scores, evidence, and assignments.Manual editorOpen a saved result for explicit manual assignment changes.RunsReview recent solve attempts, linked results, and available explanations.

Run and improve a solveGenerate a result, configure advanced solving, save progress, or continue from history.Requirements, preferences, and objectivesChoose the correct rule type, scope, target, weight, and penalty function.Group VisitsSet and verify how often people are assigned to selected groups.Troubleshoot scenarios and solvesMatch the reported state to the editor, runtime, result, or storage fix.