Back to case study
Live Analysis

ε-tx privacy scanner

Paste a Bitcoin address. The scanner runs 10 analysis phases against the real blockchain and shows you exactly how much an adversary can learn — in bits. Every number is computed live, not pre-generated.

Adversary model:
Showing pre-computed analysis of Satoshi's genesis address. Paste your own address above for a live scan.
01
Fetching history
Transactions
0
Balance
72.16259319 BTC
First seen
3 Jan 2009
Last seen
21 Mar 2024

Found 100 transactions spanning 15 years. This is the genesis address — it has received thousands of donations but the original 50 BTC coinbase reward is unspendable due to a quirk in Bitcoin's code.

02
Building co-spend graph
No co-spend clustering detected. Single-input transactions only.

This address doesn't share inputs with other addresses. No clustering leakage detected. The genesis address has never sent a transaction, so there are no co-spend links to follow.

03
Fingerprinting wallet
Likely wallet: legacy
62%
Input orderingunknown
Output orderingunknown
Script typesp2pkh
Mixed scriptsno
Change positionunknown
Anti-fee-snipingdisabled
Fee rateprecise
CoinJoin structureno
Fingerprint leakage: 0.45 bits

The address uses P2PKH (legacy format), which narrows the wallet type. But since it's receive-only, the fingerprint is weak — we can't analyse input ordering, change position, or fee rates. Classified as legacy with moderate confidence.

04
Analysing timing
00
06
12
18
UTC hour
Activity: 0:00\u201324:00
Hourly entropy: 4.48 bits

Timing patterns are weak. Donations to this address come from all over the world at all hours, producing near-maximum entropy (4.48 of 4.58 possible bits). No timezone inference, no periodicity. This is what high timing privacy looks like.

05
Amount patterns
Average output entropy0.34 bits
0 (exposed)1.0 (ambiguous)
Change detected: 0 / 100 txs
Round amounts: 67 txs
Correlations: 0 pairs

67 of 100 transactions use round amounts (0.001 BTC, 0.01 BTC, etc.). These are clearly donations, not economic transactions. The round amounts are easily identifiable but since there's no spending, there's no change detection or amount correlation.

06
Cross-chain scan
No cross-chain bridge transfers detected.

No cross-chain bridge transfers detected for this address. Privacy across chains isn't a concern here.

07
CoinJoin detection
No CoinJoin-like transactions detected.

No CoinJoin-like transactions detected. The address uses standard transaction patterns.

08
Chain analysis view
normal-payment
72%
Feature importance
output-count
35%
input-count
22%
equal-output-ratio
18%

A chain analysis classifier would label this as "normal-payment" with 72% confidence. Very low suspiciousness (2%). The simple transaction structure (1-2 outputs, single inputs) drives the classification toward normal payments.

09
Fusing evidence
1.2bits total leakagelow
Effective anonymity set: ~543 (adversary narrows identity to 1 in 543)
Bel(exposed)Pl(exposed)
0.180.52
0 (private) uncertainty 1 (exposed)
Heuristic conflict:K=0.060(low — signals agree)
amount-analysis
0.66b
wallet-fingerprint
0.45b
timing
0.18b

After fusing all evidence sources via Dempster-Shafer, the overall privacy score is 1.2 bits (low risk). Conflict between heuristics is low (K=0.06) — the signals agree that this address has minimal exposure. The wide Bel-Pl interval (0.18 to 0.52) reflects high uncertainty from limited spending data.

10
Adversary models
👁
casual
0.4 bits
low
amount-analysis0.3b
🏦
exchange
0.9 bits
low
amount-analysis0.5b
wallet-fingerprint0.2b
🛡
law enforcement
1.1 bits
low
amount-analysis0.5b
wallet-fingerprint0.3b
🌐
nation state
1.2 bits
low
amount-analysis0.6b
wallet-fingerprint0.4b

Even a nation-state adversary learns only 1.2 bits from this address. The genesis address is unusually private because it has never spent — no co-spend clustering, no wallet fingerprinting from transaction structure, no timing correlation from spending patterns. The only leakage comes from amount patterns (round donations) and the legacy P2PKH format.

Want the full analysis?

This demo runs 5 of 8 attack surfaces with a 100-transaction cap. The CLI analyses your complete history, adds Monero ring analysis, Lightning routing privacy, batch wallet timelines, and computes how your privacy score changes over every transaction.

# install
npx epsilon-tx analyse 1A1zP1eP5QGefi2D...
# full history + all adversary models
etx batch 1A1zP1eP5QGefi2D... --timeline --adversary nation-state