Skip to main content
Scoring Methodology

The STRICT Methodology

A comprehensive framework for evaluating cryptocurrency projects across six critical dimensions

What is STRICT?

STRICT evaluates crypto projects across six pillars: Sustainability, Transparency, Revenue, Innovation, Community, and Tokenomics. Each measures a critical aspect of project health.

Unlike ratings based on hype or market cap, STRICT combines hard data with expert analysis. Every project scores 0-100 on each pillar, producing a weighted overall score.

Developed from analyzing hundreds of projects across multiple market cycles, STRICT separates fundamentally sound projects from speculation-driven narratives.

The Six Pillars of STRICT

How We Score

We combine quantitative data with expert analysis, verified by AI agents and cross-referenced with authoritative sources.

1

Data Collection

Data from blockchain explorers, GitHub, CoinGecko, and on-chain analytics. All sources verified.

2

AI Analysis

Six specialized agents analyze different project aspects in parallel, each focused on its domain.

3

Validation

Seven validators cross-check data, verify claims, detect errors, and score quality.

4

Weighted Scoring

Scores weighted by asset type and sector. The final STRICT score balances health and potential.

Scoring reference

Every number in this section is read from the scoring engine's constants registry when the page renders. It is not a transcription, so the published methodology cannot drift from the code that produces the scores.

The formula

cᵢ  = published pillar score, 0–10 (null when we have no basis to score it)
wᵢ  = weight of pillar i in the coin's sector profile; Σ wᵢ = 1

score = round( Σ round(10 · cᵢ) · wᵢ / Σ wᵢ )   over observed pillars only

Each coin dataset publishes the six pillar scores on a 0–10 scale. The engine scales them to 0–100, takes the sector-weighted mean, and rounds to an integer. A pillar with no published value is excluded from the weighted sum and from the denominator, so the headline score is always the mean of the pillars that were actually scored — never a silent zero.

Pillar weights by sector profile

A Store-of-Value asset and a lending protocol are not judged on the same axes. Every sector resolves to one of nine weight profiles; the weights inside a profile sum to 1. A weight of 0% means the pillar is deliberately not part of that sector's score.

ProfileSustainabilityTransparencyRevenueInnovationCommunityTokenomics
balanced22%17%16%18%13%14%
monetary26%12%0%12%24%26%
stableAsset28%24%8%8%12%20%
derivative24%20%8%12%12%24%
financialProtocol20%16%24%16%10%14%
networkProtocol20%14%10%24%18%14%
technicalInfra18%14%14%26%14%14%
adoptionLed16%12%8%14%26%24%
paymentsRwa24%20%16%12%12%16%

Which sectors use which profile

balanced
Other
monetary
Store of Value
stableAsset
Stablecoin
derivative
Wrapped Token, Liquid Staking
financialProtocol
DeFi, Exchange, Launchpad
networkProtocol
Smart Contract Platform, Layer 1, Layer 2
technicalInfra
Infrastructure, AI, DePIN, Privacy, Identity, Social, Wallet, IoT, Telecom
adoptionLed
Gaming, Meme, Entertainment, NFT, Sports
paymentsRwa
RWA, Payments

Pillars we do not score

Two cases are structural rather than missing data, and the engine records them explicitly: Store of Value → revenue; Meme → revenue. A store-of-value asset has no protocol revenue to score, so scoring it as 0 would be a fabrication.

Everywhere else, a null pillar means we found no defensible basis for a number. In the published score that pillar is dropped, not imputed — which is a real limitation worth stating plainly: a coin scored on four pillars is not penalised for the two we could not assess. Read the pillar bars on a coin page, not only the headline.

The engine also implements a stricter policy that is not what the published score uses: a fixed denominator over all applicable pillars, a missing pillar imputed at the sector mean minus 0.75 (0–10 scale), a missing pillar whose sector weight is 20% or more imputed at max(mean − 2 × 0.75, 2) and flagged as a critical gap, and a low-coverage flag whenever observed weight falls below 70% of applicable weight.

Version and reproducibility

Engine version
v5.2.0
constantsHash
e55bd44d69e51b0ec565edfb5abf36391546a01c40c13683ba39f1ad5ac3fcdc
Weights registered
prior · owner kamyar · set 2026-07 · revisit by 2026-10-01

constantsHash is the SHA-256 of the entire constants registry, serialised with sorted keys. Change any weight or threshold on this page and the hash changes with it. Each coin dataset carries the stamp the engine produced for it — engine version, inputsHash, constantsHash and the computation timestamp.

inputsHash is the SHA-256 of the exact research inputs, canonicalised the same way. The engine is pure: no clock, no randomness, no network. Identical inputs and identical constants therefore produce a byte-identical result, which is asserted by a determinism test and, for the STRICT score specifically, by a bit-exact parity test that recomputes every published coin dataset. That is what the two hashes buy you: a score is challengeable. If the hashes match and the number does not, we are wrong.

What this score does not do

The STRICT score measures published fundamentals. It is not a price forecast, and it is not financial advice. It says nothing about entry price, timing, market conditions, liquidity, custody, counterparty or regulatory risk — a high score on a token you buy at the top is still a loss.

The weights are registered as prior — hand-tuned judgement, not fitted to outcome data. They encode our view of what matters per sector, they carry a scheduled revisit date, and we make no claim that they are optimal. A different, defensible weighting would produce different scores.

The score is also bounded by what a project discloses: an opaque project is scored on fewer pillars, and pillar scores are analyst judgement over sourced data, not measurements.

The engine's forward-looking outputs — price bands, probabilities, cycle scenarios — are a separate system from the STRICT score, and every result it emits is stamped calibrationStatus: provisional. Predictions are written to an append-only ledger and re-scored against realised prices (direction hit rates, bear-breach frequency against implied probability, Brier score, 80% interval coverage), but no calibration results are published yet. Until they are, treat those numbers as uncalibrated.

Score Tiers

Five quality tiers from 0-100.

90-100
Legendary

Elite. Exceptional fundamentals across all pillars.

80-89
Gold

Strong fundamentals with proven track records.

70-79
Silver

Good fundamentals, some areas to improve.

60-69
Bronze

Mixed fundamentals. Higher risk, potential upside.

0-59
Standard

Weak fundamentals or significant concerns. Highest risk.

Real Examples

See how STRICT Scores apply to actual cryptocurrencies across different quality tiers.

FAQ

How often are scores updated?

Weekly for major cryptos, with real-time updates for significant events like upgrades or governance changes.

Does a high score guarantee returns?

No. STRICT measures fundamentals, not price. Strong scores indicate quality, but all crypto carries risk. Market conditions affect returns regardless of score.

Why do some cryptos have null Revenue?

Store-of-value assets like Bitcoin generate no protocol fees. Revenue is marked null and excluded from their overall score.

How does STRICT differ from other ratings?

Unlike market cap or sentiment rankings, STRICT analyzes six fundamental pillars with both quantitative data and qualitative assessment.

Can I use STRICT for day trading?

STRICT suits medium to long-term decisions. For day trading, use technical analysis. STRICT helps identify quality, then apply your trading strategy.

See STRICT in Action

Explore our comprehensive crypto rankings powered by the STRICT methodology