What Is a Crypto Whale Conviction Score? The Complete Guide (2026)

What Is a Whale Conviction Score?

A crypto whale conviction score is a composite metric that measures how confident large wallet holders appear to be in a specific token. Rather than looking at a single data point — like one large buy transaction — a conviction score aggregates multiple on-chain behaviors over time to produce a single number that reflects the observable strength of whale positioning.

Think of it this way: a single whale buying $500K of a token is a data point. That same whale buying $500K, holding for 14 days, increasing their position twice, withdrawing from exchanges to cold storage, while 12 other unrelated whales do the same thing — that's conviction.

The concept of whale conviction scoring emerged because raw transaction data creates noise. Whale alert services tell you a large wallet moved tokens. They don't tell you whether that move reflects a high-confidence position or a routine rebalance. Conviction scoring attempts to distinguish between the two by layering multiple behavioral signals into a unified assessment.

At Deep Blue Alpha, we track 4,500+ Ethereum whale wallets across 200+ tokens and compute conviction scores using five distinct on-chain inputs. The result is a score from 0 to 100 that reflects how consistent, sustained, and coordinated whale behavior appears to be for any given token at any given time.

Why Single Signals Fall Short

Most whale tracking tools show you isolated events: "Whale X bought 5,000 ETH." This tells you almost nothing actionable because you lack context.

Consider these scenarios — all involving the same headline event of a large buy:

• A whale buys $2M of a token, then sells half the next day. Net conviction: low.
• A whale buys $2M of a token and holds it for 30 days while adding more. Net conviction: high.
• A single whale buys $2M while 15 other whales are simultaneously selling. Net conviction: uncertain.
• Twenty independent whales each buy $100K of the same token within the same week. Net conviction: very high.

Each of these scenarios would generate the same "whale bought" alert from a typical whale tracking service. But the behavioral context behind each is radically different. Conviction scoring exists to capture that context.

This is why the majority of whale alerts fail to produce useful research inputs — they strip away the behavioral context that makes the data informative.

The Five Inputs That Drive Conviction Scoring

A whale conviction score isn't a single measurement — it's a composite of five distinct on-chain inputs, each capturing a different dimension of whale behavior. When multiple inputs align in the same direction, the conviction score rises. When they diverge, the score reflects that uncertainty.

Here's what each input measures and why it matters.

1. Accumulation Velocity: Speed of Positioning

Accumulation velocity measures how quickly whale wallets are adding to a position in a given token. It's not just whether whales are buying — it's how fast and how consistently.

A wallet that buys $200K on Monday and nothing else all week has low velocity. A wallet that buys $200K on Monday, $300K on Tuesday, and $250K on Wednesday has high velocity. The rate of position-building reveals urgency.

Deep Blue Alpha calculates accumulation velocity by measuring the rate of net position change across tracked wallets over rolling time windows (24h, 3-day, 7-day). Accelerating velocity — buying that gets faster over time — is a stronger signal than constant-rate accumulation.

What high velocity looks like on-chain: Multiple buy transactions per wallet per day, increasing position sizes with each transaction, and decreasing time intervals between trades. This pattern suggests urgency in the accumulation behavior.

What low velocity looks like: Sporadic buys with long gaps between transactions, stable or decreasing position sizes, and no pattern of escalation. This may reflect routine portfolio rebalancing rather than directional positioning.

2. Holding Duration: Time as a Signal

Holding duration tracks how long whale wallets maintain their positions without selling. In on-chain analysis, time is one of the most underappreciated behavioral signals.

A whale that buys and holds for 48 hours is doing something very different from a whale that buys and holds for 30 days. Short holding periods often reflect trading activity or arbitrage. Extended holding periods — especially when the position could have been sold at a profit — suggest the wallet is positioned for a longer-term thesis.

Deep Blue Alpha tracks first-buy timestamps and compares them against current holdings to compute average holding duration per token per wallet. The longer whales hold without selling, the higher the duration component of the conviction score.

3. Concentration Changes: Reading Position Sizing

Concentration changes measure whether whale wallets are increasing or decreasing the share of their portfolio allocated to a specific token. This goes beyond raw buy/sell volume to capture relative positioning.

If a whale holds 10% of their portfolio in Token A and increases that to 18% over a week, that's a meaningful concentration change. It means the whale is not just buying — they're shifting their portfolio weight toward that token relative to everything else they hold.

Concentration analysis requires visibility into the full wallet portfolio, not just transactions for a single token. Deep Blue Alpha monitors 200+ tokens across each tracked wallet, which allows us to detect when whales are rotating capital from one token to another — not just adding new capital.

4. Exchange Flow Signals

Exchange flow signals track whether whale-held tokens are moving toward or away from centralized exchanges. This is one of the most watched on-chain metrics because exchange deposits often precede selling, while exchange withdrawals often precede longer-term holding.

When a whale withdraws a large amount of tokens from an exchange to a private wallet, it suggests they intend to hold rather than sell in the near term. When a whale deposits tokens to an exchange, it may indicate preparation to sell — though deposits don't always result in immediate sales.

Deep Blue Alpha monitors exchange flow activity across tracked wallets and integrates it into conviction scoring as a directional input. Net outflows (more leaving exchanges than entering) push conviction higher. Net inflows push it lower.

For a deeper exploration of exchange flow mechanics and how they compose on-chain sentiment, see our guide to the 6 on-chain signals that reveal what ETH whales are doing.

5. Multi-Wallet Convergence: The Strongest Signal

Multi-wallet convergence measures whether multiple independent whale wallets are acting in the same direction on the same token at the same time. Of all five conviction inputs, this one carries the most weight — and for good reason.

A single whale buying a token tells you about one wallet's behavior. Twenty independent whales buying the same token within the same week — wallets with no on-chain connection to each other — tells you something far more significant. It suggests that multiple large participants, independently, reached a similar conclusion about the same token.

Deep Blue Alpha identifies convergence by tracking the number of unique whale wallets that are net-accumulating a given token within a rolling time window. We then compare that count to the historical baseline for that token. If a token normally has 5-8 whale accumulators per week and suddenly has 25, that's a convergence event.

Why convergence matters most: Single-whale behavior can reflect personal portfolio decisions, OTC deals, or wallet restructuring. Multi-wallet convergence is harder to explain away. When 20+ unrelated wallets all accumulate the same token simultaneously, the behavioral signal is stronger because it reflects independent decision-making pointing in the same direction.

Convergence is also the hardest signal to fake. A single whale can create noise by splitting transactions across wallets they control. But genuine convergence across wallets with distinct on-chain histories, different funding sources, and independent transaction patterns is difficult to manufacture at scale.

Real Examples: Conviction Scoring in Action

The following examples illustrate how conviction scores reflect on-chain whale behavior. These are observational — they describe what happened, not what should have been traded.

Example 1: High Conviction Accumulation (LINK, March 2026)

Over a 7-day period in early March 2026, the LINK conviction score on Deep Blue Alpha rose from 42 to 87. Here's what the five inputs showed:

Example 2: Misleading Single Signal (AAVE, February 2026)

A whale alert service flagged a $3.2M AAVE purchase by a single large wallet. Headline: "Whale buys $3.2M AAVE." The conviction score at the time was 34. Why so low?

• Accumulation velocity: 28 — only one wallet buying; others flat or selling.
• Holding duration: 45 — the buying wallet had sold similar positions within 48 hours in the past.
• Concentration change: 31 — most tracked wallets were reducing AAVE exposure.
• Exchange flows: 22 — net AAVE deposits to exchanges were rising across tracked wallets.
• Multi-wallet convergence: 11 — only 2 wallets were net-accumulating; 19 were distributing.

The single large buy generated a headline. The conviction score revealed that it was an isolated event against a backdrop of broader distribution among tracked wallets. This is the kind of context that raw alerts miss entirely.

How to Use Conviction Scores in Your Research

Conviction scores are observational data points. They describe what large on-chain participants appear to be doing. They are not predictions, trading signals, or financial advice. Here's how researchers and analysts can incorporate them into their workflow:

1. Screening: Use conviction scores to surface tokens where whale behavior appears coordinated and directional. A rising conviction score puts a token on your watchlist for further research — it doesn't tell you to buy it.
2. Context layer: When you're already researching a token based on fundamentals, news, or technical analysis, check the conviction score to see whether large on-chain wallets are behaving consistently with your thesis.
3. Divergence detection: The most interesting readings are often divergences. If a token's price is falling but conviction is rising, it means whales are accumulating into the decline. If price is rising but conviction is falling, whales may be distributing into strength. These divergences don't predict outcomes but they surface behavioral patterns worth investigating.
4. Trend monitoring: Watch conviction scores over days, not hours. A sustained move from 40 to 75 over a week is more informative than a jump from 40 to 65 in a single day. Sustained shifts reflect persistent behavior; single-day moves can be noise.

Tools Comparison: Who Offers Conviction Scoring?

Whale conviction scoring is a relatively new concept in on-chain analytics. Most platforms offer some of the individual inputs but few combine them into a unified score. Here's how the landscape breaks down:

Most platforms give you the raw building blocks — transaction alerts, exchange flow charts, wallet labels. But they leave the synthesis to you. You have to manually check whether multiple wallets are accumulating, whether they're holding or flipping, and whether exchange flows align with the buy-side activity.

Conviction scoring automates that synthesis. It takes the five inputs described in this guide and combines them into a single, continuously updated metric for each tracked token. That doesn't make it better or worse than manual analysis — it makes it faster and more systematic.

For a complementary perspective on how whale behavior translates into sentiment readings, see our guide to the whale buy/sell ratio and how to interpret Ethereum whale sentiment.