A prediction challenge: supporters buy a slot on a grid of finish times, and the closest guess without going over takes the prize.
One campaign has run: David riding Elgin to Dornoch, 14 August 2026. The grid sold out.
Ran 20 July 2026 to 14 August 2026. Every figure below carries the query that produced it — hover any tile.
All 361 slots confirmed — every thirty-second interval between 06:00:00 and 09:00:00 was taken. There was nothing left to sell three weeks after it opened and a day before the ride.
First slot 22 July 2026 · last slot 13 August 2026
Read from the campaign's own record — the window, the interval, how a winner is picked, what happens on a tie and what happens if the rider does not finish.
361 slotsThe grid a supporter picks from. Three hours of finish times at thirty-second resolution.
closest_without_going_overThe nearest guess that does not exceed the actual time — so a wild over-guess cannot win.
earliest_confirmedIf two people hold the same winning time, the one who bought first takes it.
free_drawIf the rider does not finish, the prize goes to a free draw among entrants rather than being withdrawn.
The challenge runs as a free-entry prize competition, not a lottery. That is a different regulatory footing from a 50/50 draw and is recorded on the challenge row itself, alongside a free entry route, a minimum age of 18 and a versioned terms record — so the basis a campaign ran on can be read back years later rather than reconstructed.
No finish time is recorded against this campaign. The result was published as a manual override, and the time column is deliberately NULL rather than zero — during the 2026-09-11 backfill it was briefly written as 00:00:00 because `Number(null)` is 0 and `Number.isInteger(0)` is true. That was corrected the same day. In that column NULL means 'no result was taken' and 0 would mean 'a measured time of zero'; they are opposite claims and must not be conflated.
Result published 21 August 2026.
The mechanic is now part of the core platform as draw type 6, with 22 API routes and seven tables, rather than a bespoke build that dies with its campaign. A slot is claimed by ONE conditional UPDATE and the row count is the answer — the original site built its 'taken' set in the browser and sold three finish times twice.
PREDICTION_CHALLENGES_ENABLED is OFF in production as of 2026-09-11. The code is deployed; the feature is not switched on for clubs. Two further flags gate result publishing and each front end separately, and all are unset.
Measured read-only against the Play Fund Win production database on 11 September 2026. Commit and pull-request figures come from src/data/devlog.json.
Play Fund Win is led by experienced operators with a track record across sport, charity and fundraising technology.

Founder of Play Fund Win. Built the platform from the ground up — launching in 2020 and taking it into Premier League club foundations, national charities and grassroots clubs across the UK. Now leading European expansion with a lean, AI-augmented operating model.

Co-Founder of Play Fund Win. Brings deep commercial leadership across sport, fundraising and partnerships — driving the platform's growth into club foundations, elite clubs and live event partners. Leads commercial strategy, key account growth and European market entry.
Book a 1-on-1 call with Dean to walk through the round, the model, and the European expansion plan.
Request a Founder MeetingWe aim to respond within one working day.