pipeline: archive 1 source(s) post-merge

Pentagon-Agent: Epimetheus <3D35839A-7722-4740-B93D-51157F7D5E70>

inbox/archive/internet-finance/2026-03-24-gg-research-futarchy-vs-grants-council-optimism-experiment.md

@@ -0,0 +1,62 @@
+---
+type: source
+title: "GG Research: Futarchy vs. Grants Council — Optimism's Futarchy Experiment"
+author: "GG Research (gov.optimism.io community)"
+url: https://ggresear.ch/t/futarchy-vs-grants-council-optimisms-futarchy-experiment/57
+date: 2026-03-24
+domain: internet-finance
+secondary_domains: [collective-intelligence]
+format: analysis
+status: processed
+priority: high
+tags: [optimism, futarchy, grants, committee-selection, comparative-governance, empirical]
+---
+
+## Content
+
+GG Research published a comparative analysis of the Optimism v1 futarchy experiment (March-June 2025). This is a community analysis of the official Optimism preliminary findings, providing additional framing and interpretation.
+
+Key comparative framing (from research agent synthesis):
+
+**Selection outcome comparison:**
+- Futarchy: ~$32.5M TVL advantage over Grants Council in aggregate
+- Grants Council: lower variance, closer-to-median performance
+- Both mechanisms selected Rocket Pool and SuperForm (the 2 overlapping picks)
+- Futarchy's divergent picks included the top performer (Balancer & Beets, +$27.8M) AND the worst performer
+- Grants Council's divergent picks (Extra Finance, Gyroscope, Reservoir) showed more consistent but lower-magnitude outcomes
+
+**Key framing from the analysis:** "Futarchy favored higher-risk/higher-reward projects; the committee favored consistency."
+
+**The EV vs. variance distinction:**
+- Futarchy dominates in expected value (aggregate TVL improvement)
+- Committee governance dominates in variance reduction (no catastrophic failures)
+- The "correct" mechanism depends on the allocation objective: EV maximization → futarchy; risk minimization → committee
+
+**Caveats noted:**
+- Play-money context (Butter platform, no real stakes) — likely inflates prediction inaccuracy (8x overshoot)
+- TVL metric was endogenous to market prices in some cases (Optimism Season 7 endogeneity problem from Session 8)
+- Only 84-day measurement window
+- 45% of projects didn't disclose plans to forecasters, creating systematic information asymmetry
+
+Note: Source URL accessibility not confirmed by research agent; content synthesized from secondary research.
+
+## Agent Notes
+**Why this matters:** This is the only rigorous empirical comparison of futarchy vs. committee selection for the same pool of projects under comparable conditions. The EV vs. variance framing resolves the session-long question about whether "markets beat votes" is a universal claim or a goal-dependent design choice.
+**What surprised me:** Futarchy actually WON on aggregate TVL in the Optimism experiment. Prior sessions had treated the Optimism data as ambiguous (Session 1 noted "selection vs. prediction split"). The comparison framing from GG Research makes it clearer that on the metric that matters (actual outcome, not predicted outcome), futarchy outperformed. The catastrophically wrong predictions (8x overshoot) are a separate issue from selection quality.
+**What I expected but didn't find:** Statistical significance data. Is +$32.5M TVL a robust difference or within noise given the small sample size (5 projects vs. 5 projects)?
+**KB connections:**
+- Primary: [[futarchy-excels-at-relative-selection-but-fails-at-absolute-prediction-because-ordinal-ranking-works-while-cardinal-estimation-requires-calibration]] — the GG Research framing confirms this claim while adding the EV vs. variance dimension
+- Secondary: [[futarchy-variance-creates-portfolio-problem-because-mechanism-selects-both-top-performers-and-worst-performers-simultaneously]] — directly confirmed by this comparison
+- New scope qualifier for Futarchy is manipulation-resistant because attack attempts create profitable opportunities for defenders — the variance finding means futarchy markets can select the worst performer even in non-manipulated conditions; the EV advantage doesn't guarantee individual outcome quality
+
+**Extraction hints:**
+- New claim: "Futarchy produces better expected value than committee selection in grant allocation contexts but higher variance — mechanism choice depends on whether the objective is EV maximization or variance reduction"
+- Scope qualifier for existing futarchy claims: the "markets beat votes" superiority claim is conditional on accepting higher variance as an acceptable tradeoff. For risk-constrained allocators, the committee model's consistency may be preferable even at lower expected return.
+- Connection to Living Capital design: a diversified multi-vehicle Living Capital structure (multiple vehicles across domains) can tolerate individual vehicle variance because the portfolio diversification absorbs it. A single-vehicle allocator cannot.
+
+**Context:** GG Research is a community analysis forum connected to Gitcoin and similar grant ecosystem researchers. The analysis is practitioner-level, not academic-level. The Optimism experiment is widely cited in the governance mechanism design community as the primary empirical evidence point for futarchy vs. committee comparison.
+
+## Curator Notes (structured handoff for extractor)
+PRIMARY CONNECTION: [[futarchy-variance-creates-portfolio-problem-because-mechanism-selects-both-top-performers-and-worst-performers-simultaneously]]
+WHY ARCHIVED: Provides the EV vs. variance framing for the Optimism comparison that converts the empirical data into a design principle. The "futarchy favored high-risk/high-reward; committee favored consistency" framing is the canonical distillation of the experiment's mechanism design lesson.
+EXTRACTION HINT: Focus on the EV vs. variance distinction as a design principle, not just as an empirical finding. The claim should be: the mechanism choice between futarchy and committee governance should be made based on the allocator's objective function (maximize EV vs. minimize variance), and the Optimism experiment provides empirical support for this design principle.