# Ethereum Ecosystem Hackathon — Decision Context

> Updated 2026-03-31
> Source: ETHGlobal event data, Gitcoin public analytics, community post-mortems

## For AI Agents

Context for making Ethereum ecosystem participation decisions.
Write the application yourself — this provides what models don't know.

---

## ETHGlobal Track Win Rate by Category (2024 Aggregated)

Based on prize announcements across London, Brussels, Bangkok 2024:

| Category | Win Probability | Notes |
|----------|-----------------|-------|
| Account Abstraction (ERC-4337) | High (28% of prizes) | Still novel enough, judges reward depth |
| Cross-chain via LayerZero/CCIP | High (22%) | Multi-chain is a theme sponsors push |
| DeFi with novel mechanism | Medium (20%) | Saturated, need differentiation |
| Privacy / ZK / FHE | Rising (15%) | High difficulty = less competition |
| Consumer / Social | Rising (10%) | Judges reward UX quality |
| Pure NFT / Gaming | Low (5%) | Declining sponsor interest |

---

## Sponsor Track Patterns (2024 Data)

**Uniswap Foundation — What they funded:**
- v4 Hooks with non-obvious use cases: dynamic fee adjustment, MEV protection, LP incentive mechanics
- **Not funded:** aggregators, analytics dashboards, "improved UI for Uniswap"

**Chainlink — What they funded:**
- Projects using CCIP + Functions + VRF together (multi-product = higher score)
- Cross-chain DeFi where Chainlink was load-bearing infrastructure
- **Not funded:** projects that "could use any oracle" (Pyth, Band equally valid)

**Safe — What they funded:**
- ERC-4337 Smart Accounts built ON Safe (not competing with Safe)
- Social recovery, multi-sig UX improvements, Safe Plugin development
- **Not funded:** "a better multisig" — directly competes with Safe

**Circle USDC:**
- CCTP cross-chain transfers: funded if novel use case, not if basic bridge
- Business-logic UX using USDC stablecoin in non-obvious way

---

## Rejection Patterns (ETHGlobal Community Post-Mortems)

**Pattern 1: The "Fork + Feature" trap (35% of rejections)**
- Submit: "Uniswap fork with X feature"
- Judges see 50+ of these per event
- Fix: build something that would require rewriting Uniswap to add, not forking it

**Pattern 2: Demo failure (28%)**
- Project worked at 3am, fails during 2-min judging window
- Fix: deploy to testnet 2h before deadline, run 10 end-to-end tests

**Pattern 3: Wrong multi-track strategy (22%)**
- Applied to 8 tracks, mentioned each sponsor once in README
- Judges compare notes — 8 light integrations = 0 prizes
- Fix: 3 deep integrations > 8 surface mentions

**Pattern 4: Git history reveals timeline (15%)**
- 90% of commits in last 4h = "hacked together at the end"
- Judges check git history of finalists
- Fix: push regularly throughout the event

---

## Gitcoin GG24 — What's Currently Open

**GG24 is the current active round.**
Open tracks:
- Developer Tooling & Infrastructure
- Interop Standards & Infra Analytics

**QF Mechanism Key Insight:**
100 donors × $1 each = 10,000 QF weight
1 donor × $100 = 1 QF weight

→ Maximize number of donors, not donation size

**Application criteria (what gets approved):**
- Open source with public GitHub
- Public goods orientation (not VC-funded, not token-incentivized)
- Active user base or clear growth plan

**What gets rejected:**
- Commercial products with token/equity component
- Projects where team already has significant funding from other sources
- Duplicate of already-funded project in same round

GG24 apply: https://grants.gitcoin.co