Pattern Analysis
Identifying behavioral patterns in blockchain transactions
Overview
Pattern analysis examines transaction structures and behaviors to infer intent, identify entities, and detect suspicious activity. Unlike mixer detection or exchange identification which look for specific services, pattern analysis reveals how addresses are being used.
Key Pattern Types
1. Consolidation Patterns
Multiple inputs combined into fewer outputs—typical of wallets collecting funds.
Characteristics
- Many inputs (3-20+) from different addresses
- Few outputs (1-3)
- Output value ≈ sum of inputs minus fee
- Often precedes large payment or exchange deposit
Example
Transaction: Consolidation
Inputs:
1ABC... 0.5 BTC
1DEF... 0.3 BTC
1GHI... 0.8 BTC
1JKL... 0.4 BTC
1MNO... 0.6 BTC
Outputs:
1XYZ... 2.59 BTC (consolidated)
1ABC... 0.005 BTC (dust change)
Fee: 0.005 BTC
Analysis:
- 5 inputs → 2 outputs
- All inputs likely same entity (common input heuristic)
- Preparing for large payment or exchange deposit
- 1ABC... appears in both inputs and outputs → likely change
Investigation Implications
- Entity Clustering: All input addresses likely controlled by same party
- Intent Signal: Consolidating suggests imminent large action
- Tracking: Follow consolidated output—it's where funds are going
2. Split Patterns
One input divided into multiple outputs—distributing funds or layering.
Characteristics
- Single input or few inputs
- Many outputs (3-9 is trackable, 10+ is dispersal)
- Outputs may be similar sizes (equal distribution)
- May indicate money laundering layering
Example
Transaction: Split
Input:
1ABC... 5.0 BTC
Outputs:
1XYZ... 1.5 BTC
1DEF... 1.2 BTC
1GHI... 1.0 BTC
1JKL... 0.8 BTC
1MNO... 0.49 BTC
Fee: 0.01 BTC
Analysis:
- 1 input → 5 outputs
- Funds splitting into multiple branches
- No change address (all outputs different from input)
- Follow largest output (1.5 BTC) to continue investigation
Investigation Implications
- Layering: Classic money laundering technique
- Multiple Trails: Investigate multiple paths in parallel
- Priority: Follow largest amounts first
- Complexity: May require more resources to track
3. Peeling Chains
Sequential small payments from a large balance—characteristic of payment processing or gradual cashing out.
Characteristics
- Address starts with large balance
- Series of transactions sending small amounts
- Change always returns to same address
- Pattern continues until balance depleted
Example
Address: 1ABC... starts with 10 BTC
TX 1: 1ABC... → 0.5 BTC to 1XYZ..., 9.49 BTC change to 1ABC...
TX 2: 1ABC... → 0.3 BTC to 1DEF..., 9.18 BTC change to 1ABC...
TX 3: 1ABC... → 0.8 BTC to 1GHI..., 8.37 BTC change to 1ABC...
TX 4: 1ABC... → 0.4 BTC to 1JKL..., 7.96 BTC change to 1ABC...
...continues...
Analysis:
- Classic peeling chain
- Methodically sending to different addresses
- Suggests payment processor OR gradual cash-out
- Track each payment destination for pattern
Investigation Implications
- Payment Processing: May be legitimate business activity
- Gradual Liquidation: Or suspect cashing out slowly
- Destination Analysis: Check if recipients are exchanges
- Timing: Regular intervals suggest automation
4. Round-Trip Patterns
Funds that eventually return to original address—could indicate circular trading or shell game.
Example
1ABC → 1DEF → 1GHI → 1JKL → 1ABC
Analysis:
- Funds returned to starting point
- Possible explanations:
1. Circular trading to create fake volume
2. Layering with return to wallet
3. Failed payment returned
4. Testing or demonstration
Investigation Implications
- Obfuscation: May be deliberate complexity
- Circular Trading: Could violate exchange ToS
- Stop Following: Mark as circular reference in investigation
Temporal Patterns
Burst Activity
Sudden spike in transactions after long dormancy.
Address Activity:
2022-01: 0 transactions
2022-02: 0 transactions
...
2023-06: 0 transactions
2023-07: 47 transactions in 6 hours
2023-08: 0 transactions
Analysis:
- Address activated suddenly
- High-velocity activity then stops
- Suggests:
1. Stolen funds being moved quickly
2. Automated service activated
3. Owner finally accessing old wallet
Regular Intervals
Transactions occurring at predictable times.
Transaction Times:
Monday 09:00
Monday 09:05
Monday 09:10
Tuesday 09:00
Tuesday 09:05
...
Analysis:
- Precise regular intervals
- Clearly automated
- Suggests:
1. Payment processor
2. Salary payments
3. Automated trading bot
Value Patterns
Round Numbers
Amounts like 1.0, 0.5, 10.0 suggest human-initiated payments.
Transactions:
1.00000000 BTC
0.50000000 BTC
2.00000000 BTC
Analysis: Round amounts = human user, not automated
Precise Amounts
Odd-looking amounts suggest automated calculation.
Transactions:
0.12347891 BTC
0.08429371 BTC
0.19283746 BTC
Analysis: Precise amounts = likely automated/calculated
Equal Outputs
Many outputs with identical values.
Outputs:
0.05 BTC (appears 87 times)
0.0234 BTC (change)
Analysis: Equal outputs = mixer OR batch payment
Address Reuse Patterns
Single-Use Addresses (Good Privacy)
1ABC... receives once, sends once, never used again
Analysis: Modern wallet, good privacy practices
Heavily Reused Addresses (Poor Privacy)
1ABC... used in 500+ transactions
Analysis:
- Exchange deposit address OR
- Donation address OR
- Old wallet with bad practices
- Easier to track (more data points)
Network Analysis
Hub Addresses
Addresses connected to many others—centralization points.
Address 1ABC... connects to 200+ other addresses
Analysis:
- Exchange hot wallet OR
- Payment processor OR
- Mixer service
Implication: Critical node in the network
Isolated Clusters
Groups of addresses that only transact with each other.
Cluster:
1ABC ↔ 1DEF
1DEF ↔ 1GHI
1GHI ↔ 1ABC
No external connections
Analysis: Likely same entity testing or moving funds
Suspicious Pattern Combinations
High-Risk Pattern Stack
Multiple suspicious behaviors together.
- Receives funds → immediate split into 6 paths
- All 6 paths go through mixers within 2 hours
- Outputs consolidate at new address
- New address immediately sends to exchange
Analysis: Sophisticated money laundering—split, mix, consolidate, cash out.
Moderate-Risk Pattern
- Receives funds → waits 2 weeks
- Single transfer to new address
- New address sends to exchange after 1 day
Analysis: Simple layering with time delay—less sophisticated but still deliberate.
Low-Risk Pattern
- Receives funds
- Immediate direct transfer to known exchange
- No mixers, splits, or delays
Analysis: Legitimate user or unsophisticated actor.
Investigation Decision Trees
When Split is Detected
IF outputs count == 2:
One is likely payment, one is change → Follow non-change
ELSE IF outputs count between 3-9:
Trackable split → Investigate all paths (prioritize largest)
ELSE IF outputs count >= 10:
Dispersal → End investigation (too many paths)
When Consolidation is Detected
Consolidation detected:
1. Note: All input addresses likely same entity
2. Follow consolidated output (main destination)
3. Check if output goes to exchange (common pattern)
4. Document: Consolidation often precedes important action
Pattern-Based Confidence
| Pattern | Confidence Boost | Reasoning |
|---|---|---|
| Consolidation → Exchange | +High | Common legitimate pattern |
| Split → Multiple mixers | +Very High | Clear obfuscation intent |
| Peeling to exchanges | +Moderate | Could be business or cashing out |
| Immediate mixer usage | +High | Suggests guilt consciousness |
| Long dormancy then burst | +Moderate | Suspicious timing |
Next Steps
- See patterns in action: Reading Reports
- Understand detection confidence: Confidence Scores
- Learn investigation best practices: Best Practices