The format
Thirty minutes is the whole idea
The constraint does the work. It is short enough that you cannot hide behind scope, and long enough that something real can exist at the end. Everything else about the league — the rubric, the disclosures, the leaderboard — exists to make that constraint mean something.
Season shape
Three weeks, fifteen challenges, one championship
A season runs about three weeks each month. Administrators can change the challenge count, season length, point system, and qualification rules — the numbers below are Season 3's.
Challenges
15
one per day
Build window
30 min
60 at the championship
Qualifying spots
12
reach the final
Participation
+10
points per submission
Season flow
Twelve steps, start to champion
The same sequence every season. Trail markers, not surprises.
- 01
Registration opens
Builders register for the season. Registration stays open mid-season — you can join late, you just have fewer challenges to bank points on.
- 02
Partners and judges are announced
The panel and the partner roster are published before the first challenge, along with which challenges carry a required tool.
- 03
A challenge is released
The brief goes live at midday: theme, format, deliverables, rubric, judges, and prize. You get the afternoon to think about it.
- 04
Thirty minutes on the clock
The build window opens live. Everything you ship has to be built inside it. Scaffolding ahead of time is disqualifying.
- 05
Submissions close automatically
The form locks at the deadline. Late entries require an administrator exception, which is logged publicly in the audit trail.
- 06
Builders demo live
Demos run in an X Space. You show what works, and you say plainly what doesn't — the disclosure fields are part of the score.
- 07
Judges score
Each judge scores independently against the published rubric. Judges cannot see each other's scores until their own is submitted.
- 08
Community voting opens
Voting runs for 48 hours. Signed-in votes carry full weight, anonymous audience votes count for less, and manipulation forfeits the season.
- 09
Points are awarded
Judge scores and community votes blend into a challenge score, plus sponsor bonuses, participation points, and any penalties.
- 10
The leaderboard updates
Standings recalculate as soon as judging closes. Rank movement is shown against the previous challenge.
- 11
The championship field locks
The top twelve after the final qualifying challenge advance. Points double on that day, which routinely reshuffles the cut line.
- 12
A champion is crowned
Finalists get sixty minutes and a live stage. The championship rubric shifts weight toward execution and presentation.
Judging
The rubric is published before you build
Nothing is scored on criteria you haven't seen. Weights are tuned so that a working, honest, well-explained build beats an ambitious broken one.
| Criterion | What it measures | Weight |
|---|---|---|
| Originality | Is the idea fresh, or a well-worn demo with a new coat of paint? | 15% |
| Execution | Does it actually run? How much of the promise is real? | 20% |
| Usefulness | Would a real person use this? Is the problem worth solving? | 15% |
| Technical Quality | Sound architecture and sensible tradeoffs for a 30-minute build. | 12% |
| Design & UX | Clarity, polish, and accessibility of the interface. | 12% |
| Effective Use of AI | Is AI load-bearing here, or decorative? | 13% |
| Completion in Time | How much shipped inside the clock, per the disclosure fields. | 8% |
| Demo & Presentation | Did the live demo land? Was the process explained well? | 5% |
Weighted at 80% of the challenge score. Averaged across every judge who scored without declaring a conflict.
Weighted at 20%, and scored relative to the top-voted project of that challenge rather than by raw count.
Partners may pick their own favourites and fund prizes for them. Those picks are labelled partner-selected and carry no season points.
Fairness
What keeps this honest
A competition is only worth entering if the results mean something. These are the mechanics that protect that.
The clock is enforced by the platform
Submission forms lock at the deadline. Exceptions require an administrator and are recorded in a public audit trail.
Disclosure is mandatory
Pre-existing code, templates, collaborators, and any work done before the start must be declared. Undeclared use carries a points penalty.
Judges declare conflicts
A declared conflict removes that judge's score from the average automatically. Judges cannot see each other's scores before submitting their own.
Votes are weighted and checked
Signed-in votes count more than anonymous ones. Duplicate-account detection runs continuously, and invalidated votes are announced publicly.
Sponsors cannot touch the scoreboard
Partners have no access to scoring, rankings, or judge notes. Sponsor bonuses are fixed, published in the challenge brief, and applied by rule rather than by discretion.
Every scoring change is logged
Corrections, penalties, and appeals all leave an audit entry naming the actor, the reason, and the original value.
Ready to build?
Registration is open. You can enter mid-season.