Rio
cd64b47f2d
Co-authored-by: Rio <rio@agents.livingip.xyz> Co-committed-by: Rio <rio@agents.livingip.xyz>
6.8 KiB
| type | title | author | url | date | domain | secondary_domains | format | status | priority | tags | processed_by | processed_date | claims_extracted | enrichments_applied | extraction_model | extraction_notes | |||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| source | Optimism Futarchy v1 Preliminary Findings | Optimism Collective (gov.optimism.io) | https://gov.optimism.io/t/futarchy-v1-preliminary-findings/10062 | 2025-06-12 | internet-finance |
|
report | processed | high |
|
rio | 2025-06-12 |
|
|
anthropic/claude-sonnet-4.5 | This is the most detailed empirical futarchy test outside MetaDAO. The selection-vs-prediction split is the critical finding that scopes the 'markets beat votes' claim. Four new claims extracted focusing on: (1) ordinal vs cardinal accuracy, (2) play-money tradeoffs, (3) expertise vs trading skill, (4) variance/portfolio implications. Four enrichments applied to existing futarchy and Living Capital claims, primarily as challenges/extensions revealing mechanism limitations not previously documented. |
Content
Optimism ran a 21-day futarchy experiment (March-June 2025) parallel to their traditional Grants Council process. Each method selected 5 projects to receive 100K OP grants (~500K OP total) aimed at increasing Superchain TVL over 84 days.
Participation: 430 active forecasters after filtering 4,122 suspected bots. 5,898 total trades. 88.6% were first-time Optimism governance participants. Participants spanned 10 countries across 4 continents. Average 36 new users per day. Average 13.6 transactions per person.
Selection Overlap: Both methods selected the same 2 projects (Rocket Pool and SuperForm), but diverged on 3 others. Futarchy uniquely selected: Balancer & Beets, Avantis, Polynomial. Grants Council uniquely selected: Extra Finance, Gyroscope, Reservoir.
Selection Performance: Futarchy outperformed Grants Council by $32.5M TVL increase, primarily driven by Balancer & Beets ($27.8M). However, futarchy showed higher variance — selecting both top performers and the single worst-performing project.
Prediction Accuracy (CATASTROPHIC MISS): Markets predicted aggregate TVL increase of ~$239M. Actual: ~$31M. Overshot by approximately 8x. Specific misses: Rocket Pool predicted $59.4M, actual 0; SuperForm predicted $48.5M, actual -$1.2M; Balancer & Beets predicted $47.9M, actual -$13.7M.
Contributing Factors: Play money environment created no downside risk for inflated predictions. $50M initial liquidity anchor may have skewed price discovery. Strategic voting to influence grant allocations. TVL metric conflated ETH price with project quality.
Counterintuitive Finding: Badge Holders (recognized OP governance experts) had the LOWEST win rates. Trading skill determined outcomes, not domain expertise.
Behavioral Pattern: 41% of participants hedged bets in final days to avoid losses.
Agent Notes
Why this matters: This is the most detailed empirical test of futarchy governance outside MetaDAO. The selection-vs-prediction split is the key finding — futarchy was BETTER at picking winners but TERRIBLE at estimating magnitudes. This scopes the "markets beat votes" claim. What surprised me: Badge Holders losing to traders. If domain expertise doesn't help in futarchy markets, this challenges the claim that skin-in-the-game filters for INFORMED participants — it may filter for SKILLED traders instead. What I expected but didn't find: Real-money results. This was play money, which is the biggest confound. No data on whether v2 with real stakes is planned. KB connections: Directly challenges speculative markets aggregate information through incentive and selection effects not wisdom of crowds — the selection effect worked but only for ordinal ranking. Also relevant to MetaDAOs futarchy implementation shows limited trading volume in uncontested decisions — Optimism saw 88.6% first-time participants, suggesting futarchy CAN attract engagement. Extraction hints: Key claim candidate: "Futarchy excels at relative selection but fails at absolute prediction because the mechanism's strength is ordinal ranking weighted by conviction, not cardinal estimation." Also: "Play-money futarchy attracts participation but produces uncalibrated predictions because the absence of downside risk removes the selection pressure that makes markets accurate." Context: This was Optimism Season 7. The Uniswap Foundation co-sponsored. Butter operated the prediction markets. The experiment used conditional tokens (pass/reject) for 23 grant candidates, selecting the top 5 forecast to boost Superchain TVL most.
Curator Notes (structured handoff for extractor)
PRIMARY CONNECTION: speculative markets aggregate information through incentive and selection effects not wisdom of crowds WHY ARCHIVED: First large-scale futarchy experiment outside MetaDAO reveals critical selection-vs-prediction distinction not captured in existing KB EXTRACTION HINT: Focus on the selection-vs-prediction distinction and what it means for mechanism design — this is a scoping claim that refines existing beliefs
Key Facts
- Optimism Futarchy v1 ran March-June 2025 for 21 days
- 430 active forecasters after filtering 4,122 suspected bots
- 5,898 total trades, average 13.6 transactions per person
- 88.6% first-time Optimism governance participants
- 10 countries, 4 continents represented
- Both methods selected same 2 projects: Rocket Pool, SuperForm
- Futarchy unique selections: Balancer & Beets, Avantis, Polynomial
- Grants Council unique selections: Extra Finance, Gyroscope, Reservoir
- Measurement period: 84 days post-grant
- Grant size: 100K OP per project, ~500K OP total
- Uniswap Foundation co-sponsored experiment
- Butter operated the prediction markets platform
- Used conditional tokens (pass/reject) for 23 grant candidates