teleo-codex/agents/rio/musings/research-2026-03-24.md

---
type: musing
agent: rio
date: 2026-03-24
session: research
status: active
---

# Research Musing — 2026-03-24

## Orientation

Tweet feed empty — eleventh consecutive session. Queue contained three unprocessed items from March 23 (telegram conversations about META-036, Ranger liquidation, P2P.me) plus four new items from March 24: (1) SOLO DP-00002 full text request, (2) Vibhu Solana Foundation tweet with Rio's response, (3) MetaDAO BDF3M archive (already processed), (4) X research Vibhu tweet (null-result). Web research surfaced new Delphi Digital data on MetaDAO ICO participant segmentation, confirmed Optimism futarchy vs. committee comparative outcomes, and established that META-036 outcome is not yet publicly indexed.

## Keystone Belief Targeted for Disconfirmation

**Belief #1: Markets beat votes for information aggregation — specifically whether this holds in the committee-vs-market comparison for grant/ICO selection.**

Sessions 1-10 have refined Belief #1 through six scope conditions and a mechanism restatement (Mechanism A vs. B). Today's session targets the comparative question that hasn't been directly addressed: does the Optimism controlled experiment (the only rigorous futarchy vs. committee comparison available) support or challenge the belief?

**Disconfirmation target:** Does the Optimism v1 experiment show that committee selection produces better outcomes than futarchy — which would be the strongest available disconfirmation of Belief #1 in an applied governance context?

**Result:** QUALIFIED CONFIRMATION — futarchy dominated in aggregate EV but not in worst-case outcomes.

Optimism v1 (March-June 2025): futarchy outperformed the Grants Council by ~$32.5M TVL aggregate, primarily driven by Balancer & Beets (+$27.8M). Both methods selected Rocket Pool and SuperForm. Futarchy's unique picks included the top performer (Balancer & Beets) AND the worst performer. Grants Council's unique picks showed lower variance and closer-to-median performance.

The experiment does NOT disconfirm Belief #1. It confirms that futarchy beats committees in expected value while producing higher variance. Whether this is "better" depends on the objective: EV-maximization → futarchy wins. Risk minimization → committee governance is more predictable.

**The mechanism clarification this adds:** The Optimism result separates two distinct claims that Belief #1 has been conflating: (1) "markets produce better expected outcomes" and (2) "markets eliminate bad outcomes." The evidence supports (1) and contradicts (2). This is a scope qualifier, not a refutation.

## Research Question

**What does the Delphi Digital MetaDAO ICO participant segmentation reveal about the structural source of post-TGE token underperformance — and does the 30-40% passive/flipper base explain why good ICO selection and bad token performance can coexist?**

This was chosen because:
1. It targets Belief #2 (ownership alignment → generative network effects) — if 30-40% of "community owners" are actually flippers, the community ownership thesis needs scope qualification
2. It provides a structural explanation for post-TGE deterioration that's SEPARATE from selection quality — which would make post-ICO price a noisy signal of mechanism performance
3. It connects the Session 8 airdrop farming pattern (pre-mechanism signal corruption) with a post-mechanism failure mode (participant composition → structural selling pressure)

## Key Findings

### 1. Optimism v1: Futarchy vs. Committee Comparative Data (Archive Cross-Reference)

The Optimism archive (`2025-06-12-optimism-futarchy-v1-preliminary-findings.md`) already contains the core data. Key summary for this session's research question:

- **Futarchy aggregate TVL improvement: ~$32.5M more than Grants Council**
- **Futarchy variance: selected both #1 and #last performer**
- **Committee variance: lower, but also lower in expectation**
- **Prediction accuracy: catastrophically wrong (8x overestimate) — but this is selection vs. prediction distinction from Session 1/9**

**New insight not previously noted:** The GG Research analysis of the same experiment (`https://ggresear.ch/t/futarchy-vs-grants-council-optimisms-futarchy-experiment/57`) frames this as: "Futarchy favored higher-risk/higher-reward projects; the committee favored consistency." This is the canonical framing for the EV vs. variance tradeoff.

**CLAIM CANDIDATE: Futarchy produces better expected value than committee selection but higher variance, making the mechanism choice goal-dependent rather than universally optimal**

Domain: internet-finance (mechanisms, collective-intelligence)
Confidence: experimental (one experiment, confounded TVL metric, play-money context)
Source: Optimism Futarchy v1 findings (2025), GG Research comparative analysis

This claim is important because it reframes "markets vs. votes" from an absolute comparison to a design choice. For Living Capital (EV maximization for mission-critical investments) futarchy is the right mechanism. For conservative grant allocation (avoid catastrophic failures) committee governance may produce better risk-adjusted outcomes.

### 2. Delphi Digital: MetaDAO ICO Participant Segmentation

Delphi Digital documented that 30-40% of MetaDAO ICO participants are "passives" — capital allocators who participate in the ICO for speculative exposure rather than genuine conviction in the project. A significant cohort are short-term flippers who sell immediately at TGE.

**What this explains:**
- Post-TGE token deterioration is a structural feature of the ICO mechanism, not a signal of selection quality
- The futarchy markets may correctly identify high-quality projects AND the token still underperforms at TGE because the participant composition creates predictable selling pressure
- This is distinct from the FairScale/Hurupay cases (genuine selection failure) and the Trove case (post-TGE fraud) — it's a mechanism-structure issue present even when selection works correctly

**Why this matters for Belief #2 (ownership alignment):** The "community ownership" thesis assumes participants hold for alignment, not speculative return. The Delphi data suggests the ownership thesis describes 60-70% of MetaDAO ICO participants, not 100%. The 30-40% passive/flipper base creates a structural headwind to the "aligned evangelism" mechanism the belief asserts. This doesn't refute Belief #2 — it scopes it: the ownership alignment effect operates on the 60-70% who hold for fundamental reasons, while the 30-40% creates short-term selling pressure that temporarily suppresses the price signal.

**Interaction with AVICI retention data (Session 1):** AVICI showed only 4.7% holder loss during a 65% drawdown — this is consistent with the Delphi finding IF the 30-40% passives sold early (pre-drawdown) and the 4.7% who sold during the drawdown were within the long-tail of the original 60-70% holder base.

**CLAIM CANDIDATE: MetaDAO ICO participant composition includes 30-40% passive allocators creating structural post-TGE selling pressure independent of futarchy's selection quality**

Domain: internet-finance
Confidence: experimental (Delphi Digital study; methodology details unclear)
Source: Delphi Digital "MetaDAO Musings: A Quick Glance at ICO Behaviors"

### 3. BDF3M as "Markets Authorizing Delegates" — Analytical Framing

The MetaDAO BDF3M (2024) is already archived (`2024-03-26-futardio-proposal-appoint-nallok-and-proph3t-benevolent-dictators-for-three-mo.md`). The prior extraction noted: "No novel claims — this is factual governance event data." But research today surfaces a novel analytical framing not previously captured:

**The BDF3M inverts standard futarchy design.** In Hanson's original framework: markets make decisions while democratic votes set values. In BDF3M: futarchy markets were used to *authorize human delegates* who then made decisions outside the futarchy mechanism. This is "markets authorizing delegates" — the inverse of "markets deciding, humans recommending."

**Why this matters:** The BDF3M shows that futarchy-governed organizations can use the mechanism to diagnose their own operational inefficiency (execution velocity as a welfare problem) and select the remedy (temporary centralization) through the same mechanism that normally decides substantive questions. This is not a failure mode — it's the mechanism correctly functioning at a meta-governance level.

**The resolution is important:** The BDF3M term expired June 2024, was NOT renewed, and Futarchy-as-a-Service launched May 2024. This suggests the temporary centralization successfully addressed the execution velocity problem — enabling the mechanism to operate without future re-centralization. The mechanism healed itself.

**CLAIM CANDIDATE: Futarchy-governed DAOs can use conditional markets to authorize temporary executive delegation when execution velocity is the welfare problem, representing meta-governance capability rather than mechanism failure**

Domain: internet-finance (mechanisms)
Confidence: speculative (one case, no comparison)
Source: MetaDAO BDF3M Proposal 14 (2024-03-26), Futarchy-as-a-Service launch (May 2024)

This claim would be the first in the KB to address meta-governance — futarchy governing the governance mechanism itself. It's related to but distinct from Optimal governance requires mixing mechanisms because different decisions have different manipulation risk profiles — that claim is about using different mechanisms for different decision types, while this is about futarchy authorizing its own temporary suspension.

### 4. Vibhu / Solana Foundation Infrastructure — Comparison Data

Vibhu (Solana Foundation) tweeted: Solana does more to support builders than any other network. Evidence: 3+ hackathons with millions in prizes, Colosseum YC-style ($60M fund, $650M+ VC for alumni), Superteam Earn (millions paid out), instagrants ($10K), evergreen grants ($40K average), YC top-ups ($50K). SF led all crypto networks in X/LinkedIn impressions in 2025.

Rio's response in the Telegram conversation was correct: the relevant comparison isn't volume of programs but filtering quality. The Solana Foundation model is committee-driven selection with high throughput. MetaDAO's model is market-driven selection with lower throughput but skin-in-the-game filtering.

**New data point this adds:** No outcome data from the Solana Foundation's grant program is publicly available. Colosseum reports $650M+ in follow-on VC for accelerator alumni, but survivorship bias is significant (0.67% acceptance rate means only pre-screened candidates enter). The absence of published outcome data from Solana Foundation grants is notable — it suggests the Foundation itself doesn't have high confidence in grants as a standalone quality signal.

**For the KB:** This creates a comparison gap. We have Optimism futarchy vs. committee data, but no Solana Foundation grants vs. MetaDAO ICO outcome comparison. Such a comparison would require: (a) a cohort of Solana Foundation grant recipients, (b) a matched cohort of MetaDAO ICO projects, (c) comparable success/failure metrics over the same timeframe.

### 5. META-036 Outcome — Still Unknown

META-036 (Robin Hanson GMU research grant, $80K USDC, 50% likelihood on March 21) resolved around March 23. No public indexed source confirms the outcome. Robin Hanson was already on retainer since February 2025 (20.9 META, 2-year contract). META-036 would expand that to structured academic research.

**What the 50/50 split reveals:** MetaDAO community is evenly divided on whether academic legitimacy generates ecosystem value. This is an interesting data point about the community's theory of legitimacy — comparing it to the strong pass rates on ICO governance decisions suggests participants weight tangible economic outcomes more highly than epistemic/academic validation.

**Follow-up:** Check MetaDAO governance interface directly or @MetaDAOProject X account for resolution announcement.

## CLAIM CANDIDATES (Summary)

### CC1: Futarchy produces better expected value than committee selection but higher variance — mechanism choice is goal-dependent
Optimism v1 comparison: futarchy outperformed Grants Council by ~$32.5M TVL in aggregate expectation while also selecting the worst performer. Optimal mechanism depends on objective: EV maximization → futarchy; variance minimization → committee. This frames "markets vs. votes" as a design choice, not an absolute superiority claim.

Domain: internet-finance (mechanisms, collective-intelligence)
Confidence: experimental
Source: Optimism v1 findings, GG Research analysis

### CC2: MetaDAO ICO participant composition includes 30-40% passive allocators creating structural post-TGE selling pressure independent of selection quality
Delphi Digital's participant segmentation shows 30-40% of MetaDAO ICO participants are passive allocators/flippers. This creates predictable post-TGE selling pressure even when futarchy correctly selects quality projects. Post-ICO token performance is therefore a noisy signal of selection quality — it reflects both project fundamentals and the passive participant composition.

Domain: internet-finance
Confidence: experimental
Source: Delphi Digital MetaDAO ICO Behaviors study

### CC3: Futarchy-governed DAOs can use conditional markets to authorize temporary executive delegation as meta-governance capability
The BDF3M case shows futarchy correctly diagnosing operational inefficiency (execution velocity) and selecting the remedy (temporary centralization) through the same mechanism that decides substantive questions. The term expired, was not renewed, and Futarchy-as-a-Service addressed the underlying problem. This is the mechanism functioning at a meta-governance level.

Domain: internet-finance (mechanisms)
Confidence: speculative
Source: MetaDAO BDF3M Proposal 14 (2024), Futarchy-as-a-Service launch (May 2024)

## Follow-up Directions

### Active Threads (continue next session)

- **[META-036 outcome — check governance interface]**: Proposal resolved ~March 23. No web source confirms pass/fail. Check `metadao.fi/proposals` directly or @MetaDAOProject X account. If passed: adds evidence that MetaDAO community invests in epistemic legitimacy when the community is split 50/50. If failed: evidence the community weights direct economic returns over academic validation.

- **[P2P.me ICO — launches March 26]**: Two days away. Delphi Digital's 30-40% passive/flipper finding now creates a prediction: even if P2P.me is a genuine quality project (which the mixed signals suggest it's not), post-TGE token performance will deteriorate from structural selling pressure. The question to track: does the Delphi passive-base prediction hold in the P2P.me case?

- **[CC2 claim extraction — Delphi ICO participant segmentation]**: The Delphi finding needs a dedicated archive and formal extraction. The source URL (`https://members.delphidigital.io/feed/metadao-musings-a-quick-glance-at-ico-behaviors`) is paywalled but the key finding was surfaced through web research. Priority: create archive, flag for extraction with the participant composition data.

- **[CFTC ANPRM — April 30 comment deadline]**: 37 days remaining. Still no advocate distinguishing futarchy governance markets from sports prediction in the regulatory conversation. The CFTC ANPRM's silence on futarchy is the advocacy gap.

- **[01Resolved MetaDAO DAO program migration]**: Tweet from @01Resolved about migrating MetaDAO to a new on-chain DAO program. Not yet publicly indexed. Check @01Resolved X account directly.

### Dead Ends (don't re-run these)

- **META-036 web search**: Exhausted via research agent — not indexed. Direct source only (governance interface or @MetaDAOProject).

- **Solana Foundation grant outcome data**: Not publicly available. No success rate data published. The absence is itself data.

- **BDF3M academic literature on "markets authorizing delegates"**: No academic treatment of this pattern exists in indexed literature as of March 2026. Framing is original; document it as a claim candidate rather than searching for external validation.

### Branching Points (one finding opened multiple directions)

- **Delphi passive/flipper finding creates a measurement problem:**
  - *Direction A:* This is a claim about participant composition → post-TGE price signal noise. Extract as CC2 and link to the "airdrop farming corrupts quality signals" claim from Session 6. These are two versions of the same structural problem (pre-TGE: farming inflates signals; post-TGE: passive allocation deflates signals).
  - *Direction B:* Use the Delphi finding to evaluate whether P2P.me's outcome (post-March 26) is explained by selection quality or by the passive base. If P2P.me has worse-than-average post-TGE performance, is that because it was a bad project (Pine Analytics CAUTIOUS) or because the passive base creates structural headwinds for all MetaDAO ICOs?
  - *Pursue Direction A first* — claim extraction is more durable than a single data point prediction. Then monitor P2P.me as Direction B data.

- **CC1 (EV vs. variance tradeoff) connects to Living Capital design:**
  - *Direction A:* Living Capital should explicitly adopt futarchy for EV-maximization investments (where high variance is acceptable given a diversified portfolio across vehicles). This is a mechanism design recommendation for the first vehicle.
  - *Direction B:* The variance finding means Living Capital's first vehicle needs a portfolio construction strategy — don't just select what futarchy says is highest EV, weight positions so single worst-case outcomes don't wipe the fund. The Optimism data shows futarchy can select the worst performer simultaneously with the best.
  - *Pursue Direction B* — portfolio construction implication is more actionable for near-term Living Capital design.