Indivisible Goods Fairness Workbench
Audit an authored goods allocation or a declared round-robin pick sequence with exact envy and EF1 witnesses.
Loading tool…
How to use Indivisible Goods Fairness Workbench
Enter one nonnegative additive value for every agent-item pair. Audit a complete allocation or pick in your stated round-robin order, then examine how each observer values their own and every other bundle, including the best one-good removal.
- Use the example as the schema; replace its labels and entered model assumptions.
- Enter one nonnegative additive value for every agent-item pair. Audit a complete allocation or pick in your stated round-robin order, then examine how each observer values their own and every other bundle, including the best one-good removal.
- Calculate, inspect the complete witnesses and limitations, then copy or download the JSON report.
Good to know
- 2–10 distinct agents, 1–50 distinct items and a complete rectangular value matrix. Utilities are nonnegative exact decimal strings at most 1,000,000,000. Audit mode must allocate every item exactly once; round-robin requires a complete unique agent order.
- Values add across a bundle, with no complementary goods, chores, negative values, money transfers or uncertain/hidden preferences. EF1 removes at most one good from the other bundle for the comparison; it does not transfer that good.
- Round-robin is the declared picking rule with source-item tie breaks, not an efficiency or truthful-reporting mechanism. All conclusions refer only to your matrix, not interpersonal welfare, legal ownership or a fairness certification.
- Local calculation with no upload, input persistence or outside lookup. Model JSON is limited to 30,000 characters and 10,000 values. Decimal quantities are strings with up to six places; count fields use plain JSON integers. Exact fractions are capped at 8,192 bits. The complete JSON output is limited to 1,000,000 characters; approximate numbers preserve their stored values. Unsupported states fail explicitly.
Common questions
Can an allocation be EF1 without being envy-free?
Yes. An observer may prefer another person’s whole bundle, yet that envy disappears after at most one good is removed from the other bundle for the comparison. The report shows both conditions and the observer-specific removal witness, without changing the actual allocation.
Are my model inputs saved or sent to a server?
No. This model runs in the current browser tab and creates a local JSON download. It does not save your input or call an external data service.
Keep going with related tools
All education →Shared Expense Settlement Workbench
Split expenses in exact currency minor units, reconcile who paid and who owes, and download a conserved transfer plan.
CalculatorsExact Integer Allocation Workbench
Distribute a fixed number of indivisible units by entered weights, preserving the exact total and exposing quota remainders.
CalculatorsPairwise Priority and Consistency Workbench
Convert a complete reciprocal comparison set into normalized priorities and a visible consistency index.