YC Fantasy

Type: Software Status: In Progress (launch 2026-06-16) Deadline: Post-Demo-Day launch — 2026-06-16 (Zeitgeist’s chosen arc; site already live at draft.dealdeck.vc) Created: 2026-05-28

What is this?

A lightweight, web-only fantasy/ranking game for the current YC batch (Spring 2026, 177 companies). Visitors pick the companies they think are most interesting, then see how their picks compare to (a) the aggregate of all players and (b) a proprietary algorithm. The game is hosted on its own subdomain/site — not inside the DealDeck app — and acts as a top-of-funnel exposure play with a clear bridge back to DealDeck.

Motivation

Distribution: Demo Day is a moment of peak attention on YC; a shareable, low-friction game rides that wave.
Positioning: Demonstrates that DealDeck has a credible opinion (the algorithm) about which startups matter — not just a content app.
Lead capture: Email-gated reveal (TBD) seeds the DealDeck v2 waitlist with a self-selected VC/operator audience.
Re-use of existing assets: Taylor’s tracker is already a working daily-scraping + scoring pipeline. Wrapping it in a public game is the cheapest way to make it visible.

What does success look like?

Site live on or before 2026-06-14 with a clean, playable mechanic and a satisfying reveal screen
Clear, prominent path from result screen → DealDeck app waitlist / TestFlight
Some volume of plays + shares from the elite-VC and YC-orbit Twitter / LinkedIn audience
Post-Demo Day scoring window keeps the page sticky for ~4 weeks as event points accrue

Relationship to DealDeck

Codebase: separate — lives at ~/projects/yc-fantasy/ (stack TBD)
Documentation: separate — this folder
Brand: distinct surface, but co-branded (“by DealDeck” or similar) and visually consistent
Data: shares the YC ingestion logic conceptually with DealDeck’s existing YC scraper (M-42), but the source of truth for this project is Taylor’s tracker (Google Sheet + Apps Script)
Bridge: result screen CTA → DealDeck waitlist / app

Technical

Stack: static HTML + Tailwind CDN + SortableJS + Supabase JS (auth + REST) + Cloudflare Pages. Single index.html; build script copies it to dist/ and wrangler pages deploy ships.
Repo: ~/projects/yc-fantasy/ → github.com/djt53/yc-fantasy (private)
Deploy: Cloudflare Pages project yc-fantasy → https://yc-fantasy.pages.dev, custom domain https://draft.dealdeck.vc (DNS auto-wired since dealdeck.vc is on CF nameservers)
Backend: Shared DealDeck Supabase project (euogsdojqtejmbzxdzyi); yc_-prefixed tables (yc_companies, yc_drafts) + yc_leaderboard() RPC. Schema applied via dashboard, registered with supabase migration repair --status applied. Migration source-of-truth at supabase/migrations/20260614104701_init_yc_fantasy.sql.
Daily data refresh: Taylor’s Apps Script pushes the YC Fantasy Leaderboard sheet to yc_companies daily via REST upsert (service-role bearer). Spec at docs/apps-script-spec.md.
Intended users: VCs, operators, founders, YC-orbit Twitter — broader and more casual than the DealDeck app audience

Data sources

Primary: Taylor’s YC Company Updates Tracker.xlsx (Google Sheet, daily Apps Script scrape of YC’s API)
- YC Fantasy Leaderboard sheet — 177 Spring 2026 companies with draft_base_score, pre_demo_total_score, total_score, event counts, last-event dates
- YC Scoring Config sheet — full algorithm definition (~60 rules)
- Companies to Watch sheet — 31-company editorial list with qualitative notes (potential third “opponent” — editorial picks)
- YC Fantasy Events sheet — 4,954 scored event rows feeding the post-Demo Day momentum loop
Snapshot exported to assets/yc-leaderboard.json (177 rows) and assets/companies-to-watch.json (31 rows) on 2026-05-28 for prototyping

Design Decisions (all resolved this session)

Mechanic — portfolio-builder ($20M budget, 10 picks, drag-to-fund chips). Matches Zeitgeist’s live mechanic. We diverged from Zeitgeist’s “≥3 sectors” rule (no sector constraint here).
Opponents — algo + crowd via yc_leaderboard() RPC. Editorial (Companies-to-Watch) deferred.
Timing — rolling submission, post-Demo-Day settlement window (Zeitgeist’s June 16 arc).
Auth — email-gated via Supabase OTP (shared with DealDeck v1/v3 auth).
Branding — co-branded Zeitgeist + DealDeck (welcome modal + reveal footer).
Domain — draft.dealdeck.vc subdomain. Portable to draft.zeitgeist.bot later via CF Pages custom-domain swap.

Milestones

#	Name	Status	Notes
M-1	Scoping + 3 mechanic prototypes	✅ Done	Closed 2026-05-28 + Taylor walkthrough 2026-05-29
M-2	Pick mechanic + commit to design	✅ Done	Portfolio-builder; opponents/timing/auth/branding/domain all resolved 2026-06-14
M-3	Production build	✅ Done	Single index.html, draft + fund mechanic, results scaffold; T-24 polish open
M-4	Backend (drafts, leaderboard, auth)	✅ Done	yc_drafts + yc_leaderboard RPC + OTP wired. Daily refresh spec’d; Taylor implements.
M-5	Branding + content + DealDeck bridge	🔶 In Progress	Co-brand + domain done; methodology/FAQ copy + OG share card pending
M-6	Soft launch	🔶 In Progress	Live at draft.dealdeck.vc; QA + launch comms pending; Zeitgeist redirect/embed TBD
M-7	Post-Demo Day scoring loop	🔶 In Progress	Returning-visitor flow shipped 2026-06-15 (T-61a–T-61d); full-page Your Draft view shipped same day (T-66, closes T-61g); still gated on Taylor’s Apps Script (T-32) for the daily score refresh and weekly digest (T-61f)
M-8	V1 Feedback follow-ups	🔶 In Progress	T-V1a–T-V1d shipped 2026-06-15 (welcome copy, points panel, email gate consent, funded counter); T-63/T-64/T-65 deferred post-launch

Risks

R-1: Demo Day timing (high / high) — 17 days from kickoff to launch. Scope discipline is everything.
R-2: Cold-pick problem (medium / medium) — Pre-Demo Day, casual visitors don’t know these companies. Reduction mechanic + interesting one-liners must compensate.
R-3: Audience mismatch (medium / medium) — Game framing may feel too gimmicky for elite VCs; needs polish to land.
R-4: Dependency on Taylor’s sheet (low / medium) — Single source of truth is a Google Sheet + Apps Script. Fine for v0, may need to harden if it goes viral.