Seeded Giveaway Winner Picker

Give every entrant one entry, or assign larger entry counts to reflect bonus actions, ticket bundles, or other published rules. Because the calculation never uses a clock or hidden random source, organizers can save the seed and input alongside the result, rerun the draw later, and demonstrate exactly how the named winner was selected.

Prepare an ordered entrant list with honest weights

Start with one row per entrant and keep the rows in a deliberate, stable order. Each row needs a non-empty name. The optional entries value is a positive whole number: omit it for one entry, use two for twice the selection weight, and so on. A participant with five entries occupies five positions in the conceptual ticket pool, while a participant with one entry occupies one position. The picker does not deduplicate names because repeated names may represent legitimate separate records, and silently merging them would change the published odds. If your source can contain duplicates, decide whether to consolidate them before the draw and preserve that decision in your campaign notes. Also freeze the final ordered list before announcing the seed. Sorting, adding a row, removing a row, renaming somebody, or changing an entry count changes the weighted ranges and may therefore change the winner. The calculation accepts up to 10,000 entrant rows and validates every weight rather than rounding fractions or accepting zero-weight placeholders. This keeps the input contract clear: every submitted row participates, and every declared entry count contributes directly to that entrant's chance.

Use a seed to make the draw reproducible

A seed is any non-empty text chosen for the draw, such as a campaign identifier combined with a closing-date label. The picker converts that exact text into a fixed unsigned 32-bit value using a documented hash, then maps the value to one position in the total weighted entry range. It walks the entrants in their supplied order and returns the row whose cumulative range contains that position. The same seed, entrant order, names, and entry counts always produce the same output on the API and in the browser. Capitalization, spaces, and punctuation in the seed matter, so save the exact original string instead of a retyped approximation. For a public process, consider committing to how the seed will be obtained before entries close—for example, a clearly identified external event value published after the deadline. This tool remains deterministic rather than cryptographically random: its purpose is repeatable selection and straightforward auditing, not generating an unpredictable secret. Do not reveal a freely chosen seed early if participants or organizers could still alter the entrant list in response to the projected result.

Record and verify the winning result

The result includes the winner's name, the zero-based index of the winning row, the selected zero-based weighted entry position, the total number of weighted entries, and the numeric hash derived from the seed. Keep the complete input with those fields in your giveaway record. A reviewer can submit the same ordered rows and seed, confirm that the total matches, and receive the identical winner without trusting a screenshot or an unexplained random button. The row index also distinguishes two records that happen to share the same displayed name, while the weighted position shows where the selection landed inside the conceptual ticket pool. If a rerun differs, compare the saved input character by character: even a trimmed name, reordered row, adjusted weight, or differently capitalized seed represents a different draw. The browser version is useful for a transparent manual selection, while the API costs $0.002 per request when you need to automate the same procedure in campaign software or an audit trail. In either channel, the algorithm has no network calls, current-time dependency, or mutable state, so execution environment and run date do not alter the answer.