Section

The Book

The encyclopedia. Neutral, citation-dense entries on how hackathons actually work — history, formats, judging, problem statements, sponsorship, integrity, the AI era, named failure modes.

The Book

Stub. Drafted during Phase 1 IA skeleton seeding. The piece below is a one-paragraph placeholder — replace before publishing.

The encyclopedia. Neutral, citation-dense entries on how hackathons actually work.

All entries (10)

  • A Field Guide to the Ten Working Hackathon FormatsGrowing

    A field guide to the ten working hackathon format archetypes — Single-Problem Competition, Themed Multi-Track, Sponsor-Bounty Federation, Government Civic, University Season, Internal Corporate, Platform Ecosystem Challenge, Code Sprint, Game Jam, and Blind Replication Sprint — with worked examples and named failure modes for each.

  • How to Write a Problem Statement Worth SolvingGrowing

    A problem statement is the artifact a hackathon team builds against. This entry walks Murby's three-part test in operational depth, presents the two valid architectures (organizer-issued and participant-authored), and names the failure modes that recur most often in practice.

  • AI lowered the floor, AI raised the ceilingDraft

    Cataloguing the tools that lowered the floor (Claude Code, Cursor, Lovable, v0, Bolt, Replit, Kiro) and the evidence the moat at the top got higher.

  • Hackathon cultureDraft

    Why the best events feel like a discipline, not a contest. The norms that travel across formats and the ones that do not.

  • Named failure modesDraft

    Demo-stage bias, hardware-stage bias, rubric mismatch, theme drift, sponsor capture, ringer mislabeling, peer-judging gaming. Each named to be diagnosable.

  • A short history of hackathonsDraft

    From OpenBSD's 1999 code sprints to TechCrunch Disrupt to MLH-supported student season to AI-era vibe-coding events.

  • Integrity mechanics: ringers, disclosure, embargoDraft

    The Class A/B/C labeling system, .gitcheck, Devpost project-validation, and the convergence-as-integrity model in no-ground-truth domains.

  • Judging fairly: tracks, rubrics, and the inter-rater problemDraft

    Conventional judging and its failure modes; pivot to integrity-through-convergence when no ground truth exists. The site's signature Book entry.

  • Rubric design without statistical liesDraft

    NASA's five-criterion model, calibration sessions, inter-rater reliability, why single-equal-weight rubrics hide bias.

  • Sponsorship economics and the bounty federationDraft

    How partner-prize architecture works, what good looks like (ETHGlobal), and where sponsor capture distorts agenda.