Athena Methodology

How Athena evaluates credibility. Scores reflect observed behavior, not opinions.

Version 1.0.0 • Last updated: December 2025

6
Athena Dimensions
60-80
Factors Analyzed
v1.0.0
Score Version
Our Philosophy

We believe credibility is about track record, not popularity. A crypto influencer with millions of followers but consistently wrong predictions should rank lower than someone with fewer followers but accurate claims.

Our scoring system is designed to be:

  • Objective: Based on verifiable market outcomes, not opinions
  • Transparent: Every score shows exactly why it's that number
  • Defensive: Can explain any score to influencers or investors
  • Gaming-resistant: No single factor dominates; diversified assessment

The 6 Credibility Pillars

Every score is built from these fundamental dimensions. Each pillar measures a different aspect of credibility, weighted by importance.

1. Outcome Alignment (30%)Highest Weight

Do their claims match reality?

Measures how often predictions align with actual market outcomes. We check:

  • Direction accuracy (50%): Did the price go up/down as predicted?
  • Target accuracy (30%): Did it reach the specific price target?
  • Timeframe accuracy (20%): Did it happen in the predicted window?
✓ Good Example:

"Bitcoin will reach $50K by end of Q1" → BTC hits $51K on March 28th

✗ Poor Example:

"Ethereum going to $10K this month" → ETH drops 15% instead

2. Calibration (20%)High Weight

Does their confidence match their accuracy?

Well-calibrated influencers are highly confident when they're right, less confident when uncertain. We penalize overconfidence—making bold claims that don't pan out.

✓ Good Example:

High confidence predictions: 85% accuracy • Medium confidence: 60% accuracy

✗ Poor Example:

"I'm 100% certain SOL will 10x" → SOL drops 30% (overconfidence penalty)

3. Consistency (15%)Medium Weight

Do they maintain stable positions over time?

Detects flip-flopping—making opposite calls on the same asset within short timeframes without acknowledging the change. Consistency builds trust.

✓ Good Example:

Maintains bullish stance on BTC for 6 months, even during dips

✗ Poor Example:

"BTC to $100K!" (Monday) → "BTC will crash, sell now!" (Friday) → No explanation

4. Specificity Quality (15%)Medium Weight

Are their claims precise and falsifiable?

Vague claims like "crypto will moon soon" are unfalsifiable. We reward specific targets and timeframes that can be objectively verified.

✓ Good Example:

"ETH will reach $4,200 by March 31, 2025"

✗ Poor Example:

"Altcoins looking good, major moves coming" (no asset, target, or timeframe)

5. Transparency (10%)Lower Weight

Do they disclose conflicts of interest?

Measures disclosure of sponsorships, paid promotions, and asset holdings. Transparency prevents pump-and-dump schemes and builds audience trust.

✓ Good Example:

"Full disclosure: I hold SOL and was paid by Solana Foundation for this video"

✗ Poor Example:

Promotes token heavily, later revealed they dumped before audience could buy

6. Accountability (10%)Lower Weight

Do they acknowledge mistakes and update positions?

Great influencers admit when they were wrong and explain why their thesis changed. This shows intellectual honesty and builds long-term trust.

✓ Good Example:

"My BTC prediction was wrong. Here's what I missed: [detailed explanation]"

✗ Poor Example:

Never mentions past failed predictions; deletes old videos with wrong calls

How We Calculate Overall Scores
Formula:
Score = (Outcome × 30%) + (Calibration × 20%) + (Consistency × 15%) +
(Specificity × 15%) + (Transparency × 10%) + (Accountability × 10%)

Each pillar is scored 0-1, then weighted and combined into a final 0-100 score. The weights reflect importance: track record (outcome) matters most, while transparency and accountability are bonuses.

Confidence Levels:

  • HIGH: 50+ evaluated claims (statistically significant)
  • MEDIUM: 20-49 claims (meaningful sample)
  • LOW: 5-19 claims (preliminary assessment)
  • MINIMAL: <5 claims (insufficient data)
Gaming Resistance

Our scoring system is designed to prevent manipulation:

  • No single factor dominates: Even perfect outcome alignment only gets you to 30/100. You need balance across all dimensions.
  • Verifiable outcomes: We check claims against real market data, not sentiment or likes.
  • Time-weighted: Recent behavior matters more, but flip-flopping is penalized.
  • Transparency required: Hidden affiliations hurt your score even if predictions are good.
  • Accountability matters: Deleting failed predictions or ignoring mistakes reduces trust.

Bottom line: The only way to consistently maintain a high score is to make accurate, specific, well-calibrated claims while being transparent about conflicts and accountable for mistakes.

Trust Badges

Badges describe what influencers do, not what score they have. They're earned through consistent behavior and can be lost if performance drops.

🔍 Discloses Conflicts Consistently

Earned by: Transparency score >0.8
Means: Consistently discloses sponsorships, holdings, and conflicts of interest

📊 Consistent Track Record (30+ claims)

Earned by: Consistency >0.75 + 30+ evaluated claims
Means: Maintains stable positions across market cycles, doesn't flip-flop

✅ Demonstrated Accuracy (50+ claims)

Earned by: Outcome alignment >0.7 + 50+ evaluated claims
Means: Proven accuracy over statistically significant sample size

🔄 Admits & Corrects Errors

Earned by: Accountability >0.7
Means: Acknowledges mistakes publicly and updates positions when wrong

🎯 Confidence Matches Reality

Earned by: Calibration >0.75
Means: When confident, they're right; when uncertain, they say so

Questions or Concerns?

We're committed to transparency and fairness. If you:

  • Believe your score is incorrect
  • Have evidence of misclassified predictions
  • Want to understand specific driver explanations
  • See suspicious scores or gaming attempts

Formal Dispute & Correction Process

We have a transparent 4-step process for challenging scores and correcting factual errors. All corrections are logged and visible.

View Dispute Process
View Athena Index Rankings