The 25-Variable Scoring Framework: How We Quantify What Most Research Leaves Qualitative

Every research process has hidden assumptions. Ours are explicit. The 25-variable scoring framework forces every token through the same 100-point evaluation across five categories, with no narrative exceptions and no momentum adjustments. ETH scores 79. HYPE scores 76. ZRO scores 33, which is a sell regardless of how many people are talking about it.

The framework exists because qualitative research compounds its own biases. When you evaluate a token without a structured rubric, you weight the factors you remember most recently, the narratives you find most compelling, and the teams you have met in person. These are not good filters for investment quality. The 25-variable framework is an attempt to replace those biases with a consistent signal across all 25 dimensions.

The five categories each contain five variables, and each variable is scored on a 1-10 scale. The sum is a 100-point composite. Green is 70 and above. Yellow is 40-69. Red is 39 and below. The color is not a recommendation — it is a starting point for position sizing and further investigation.

Protocol Revenue contains five variables: Protocol Revenue (absolute dollar figure annualized), Revenue Trend (is it growing or contracting), P/S Multiple (current market cap divided by annualized protocol revenue, not total fees), and two supplementary revenue quality checks. A token can score well on absolute revenue but poorly on trend — which is exactly the pattern we see in protocols that dominated a prior cycle but are losing market share to newer entrants.

Supply and Tokenomics contains: Supply Inflation (what is the annualized emission rate eating into holder returns), Unlock Schedule (are large tranches of team or investor tokens coming unlocked in the next 12 months), Circulating-to-FDV Ratio (how much hidden supply exists in the form of vested tokens not yet circulating), Buyback or Burn mechanics (does the protocol structurally reduce supply), and Insider Selling (is on-chain data showing team wallets distributing).

Ecosystem and Usage contains: Staking Yield (what do holders earn for participating in protocol security), TVL Trend (is total value locked growing or shrinking), Active Users (monthly active wallet count and trend direction), Developer Activity (GitHub commits over 30 and 90 days), and Ecosystem Growth (new integrations, partnerships, and protocol composability).

Competitive Position contains: Market Share (percentage of category volume or TVL controlled by this protocol), Competitive Moat (number one or two in category, or defensible niche), Institutional Interest (whether named institutions are allocating or building on this protocol), Exchange Depth (order book liquidity as a proxy for market maker confidence), and Regulatory Safety (whether the token has clear regulatory status or faces meaningful enforcement risk).

Risk and Market contains: Catalyst Calendar (specific dated events that could reprice the asset in the next 180 days), BTC Alpha (historical performance versus BTC in bull and bear conditions), Team Execution (track record of delivering on stated roadmaps), Treasury and Runway (does the protocol have sufficient reserves to operate for two-plus years without token sales), and Social Mindshare (relative conversation volume as a leading indicator of retail interest).

We ran 40 parallel research agents to complete the scoring pass on the tokens that survived the kill filter. Each variable was evaluated independently by a research agent using a standardized prompt with defined scoring anchors. The 40-agent parallel execution meant we could complete the full scoring matrix in a single session rather than over multiple days of sequential work.

The results validated the framework's ability to surface non-obvious conclusions. ETH at 79 out of 100 scores highest on Competitive Position and Regulatory Safety, with a meaningful drag from Supply Inflation (post-merge issuance is low but not zero) and Social Mindshare (ETH narrative has been dormant relative to competitors for six months). HYPE at 76 scores highest on Protocol Revenue and Team Execution, with a drag from Regulatory Safety — the protocol operates without a legal entity in a way that creates non-trivial regulatory optionality, positive or negative depending on how enforcement evolves.

ZRO at 33 is the clearest sell signal the framework has produced. The token has low absolute revenue, a high P/S multiple, a circulating-to-FDV ratio that suggests significant future dilution, no buyback or burn mechanism, and a competitive position that has weakened as cross-chain bridge alternatives have proliferated. The 33 score is not a judgment on the team or the technology. It is a judgment on the token as an investment instrument at current prices.

The framework is not a black box. Every score is derived from a specific data point. Protocol Revenue comes from DeFiLlama. TVL Trend comes from on-chain data. Developer Activity comes from GitHub API. Exchange Depth comes from order book snapshots. Social Mindshare comes from relative search volume and mention frequency. The inputs are public. The scoring anchors are documented. The output is reproducible.

Scores below 40 trigger an automatic review for exit or avoidance. Scores above 70 are eligible for position entry, pending valuation check. The 40-70 band requires a specific catalyst or structural improvement to justify deployment. This is not a trading signal. It is a capital allocation discipline.

Author: Early Thunder Research Data sources: DeFiLlama, GitHub API, CoinGecko, on-chain analytics, order book data Last updated: 2026-05-21

This content is for informational purposes only and does not constitute financial advice.