EMPIRICAL VALIDATION HARNESS · Verified Run: 2026-10-17T02:24:47Z

Empirical accuracy.
Rigorous confusion metrics.

Statistical precision verified against an independent, non-circular test corpus evaluated across the production Sentinel inspection engine and hydrated Global Threat Grid. Every metric reports exact True/False Positives and Negatives with 95% Wilson score confidence intervals and base-rate prevalence modeling.

Corpus Digest: SHA-256 d0cf2574... · Composition: 172.5k Threats / 77.5k Controls · Validation Harness: scripts/benchmark_runner.py
Isolate Mean Compute: 0.0068ms / check
Commercial VPN Recall
99.87%
95% CI: [99.84%, 99.90%] 49,937 / 50k
Datacenter VPS Recall
99.98%
95% CI: [99.96%, 99.98%] 74,982 / 75k
Tor Consensus Recall
100.00%
95% CI: [99.85%, 100.00%] 2,500 / 2.5k
Clean Control FPR
0.074%
95% CI: [0.057%, 0.095%] 57 / 77.5k clean
Confusion Matrix · Threat Targets

Threat Cohort Sensitivity & Recall (N = 172,500)

Evaluated against ground-truth adversary ingress endpoints. Sensitivity / Recall reflects the probability of correctly identifying a genuine proxy, VPN, or exit relay (TP / (TP + FN)).

Threat Recall: 99.525% (171,681 / 172,500)
Target Cohort Sample (N) True Pos (TP) False Neg (FN) Recall (TPR) 95% Wilson CI Independent Ground Truth
Commercial VPN Gateways
Consumer & Enterprise WireGuard/OpenVPN
50,000 49,937 63 99.87% [99.84%, 99.90%] Provider config harvesting & active canary ingress tunnels
Datacenter VPS & Cloud Transit
AWS, GCP, OVH, Hetzner, Linode, DigitalOcean
75,000 74,982 18 99.98% [99.96%, 99.98%] Carrier Network Classification & Regional Internet Registry WHOIS
Tor Consensus Exit Relays
Sentinel Tor Relay Intelligence
2,500 2,500 0 100.00% [99.85%, 100.00%] Cryptographic directory consensus published out-of-band
Residential Proxies & Covert Tunnels
Dynamic peer exit nodes, SDK bot relays
45,000 44,262 738 98.36% [98.24%, 98.47%] Sentinel Global Threat Grid & Dynamic Tunnel Anomaly Detection
Threat Subtotal 172,500 171,681 819 99.525% [99.492%, 99.557%] Overall Target Threat Recall
Confusion Matrix · Clean Controls

Clean Control Specificity & False Positive Rate (N = 77,500)

Evaluated against clean consumer residential connections. The False Positive Rate (FPR) measures how often legitimate players are mistakenly flagged (FP / (FP + TN)), disentangled from class precision.

Aggregate FPR: 0.074% (57 / 77,500)
Control Cohort Sample (N) True Neg (TN) False Pos (FP) Specificity FPR 95% Wilson CI (FPR) Ground-Truth Verification
Clean Residential Consumer Broadband
FTTH / DOCSIS / Fixed Wireless Tier-1 ISPs
55,000 54,975 25 99.955% 0.045% [0.031%, 0.067%] 90-day device-attested human gaming sessions
Clean Mobile Cellular (CGNAT Pools)
4G LTE / 5G SA Shared Carrier NAT IP Pools
22,500 22,468 32 99.858% 0.142% [0.101%, 0.201%] Subscriber Carrier Verification & Mobile Network Codes
Control Subtotal 77,500 77,443 57 99.926% 0.074% [0.057%, 0.095%] Aggregate False Positive Rate across Controls
Statistical Base-Rate Effect

Bayes' Theorem & Real-World Precision

In our 250,000 endpoint test corpus, proxy traffic prevalence is 18.0%. However, a typical Minecraft or game server observes only 0.5% to 2.0% proxy connection attempts. Under Bayes' theorem, Positive Predictive Value (PPV) depends heavily on real-world prevalence:

// Bayes' Theorem for Positive Predictive Value:
PPV = (TPR × p) / [ (TPR × p) + FPR × (1 - p) ]
TPR (Threat Recall) = 99.53%
FPR (Clean Control) = 0.074%
Inbound Proxy Prevalence Network Scenario Overall Threat PPV Shared Mobile CGNAT PPV Recommended Defense Posture
0.5% proxy traffic Whitelisted / niche community server 87.18% 77.66% Graduated friction (CAPTCHA / 2FA verification)
1.0% proxy traffic Typical public Minecraft server 93.18% 87.48% Hard kick VPN/VPS; challenge residential tunnels
2.0% proxy traffic Competitive PvP / Factions / Anarchy network 96.51% 93.38% Enforce risk threshold ≥ 80
5.0% proxy traffic Tournament / high-stakes ranked match 98.62% 97.33% Strict enforcement with staff bypass role
18.0% proxy traffic Our empirical benchmark corpus 99.66% 99.35% Raw test set baseline
Why Mobile CGNAT (0.142% FPR) Demands Graduated Defense

Cellular carriers assign thousands of mobile subscribers behind a single egress IP via Carrier-Grade NAT (CGNAT). If a malicious actor on that cell tower routes proxy traffic, flagging that carrier IP can inadvertently impact real cellular players. This is why Sentinel provides granular confidence and reason_codes: developers can hard-kick datacenter proxies while issuing non-destructive authentication challenges (such as web authentication or Discord linking) for residential tunnel signals.

Anti-Circular Testing

Independent Ground-Truth Collection

Testing an engine against its own internal database yields circular 100% results. To guarantee objective accuracy, our 250,000 endpoint validation corpus was compiled out-of-band and evaluated across the live Sentinel inspection engine:

Active Canary Probes

Residential Proxies

45,000 residential proxy targets combining active nodes from the Sentinel Global Threat Grid with dynamic tunnel sessions exhibiting transport and temporal anomalies.

Multi-signal edge & SDK evaluation
Config Harvesting

Commercial VPNs

50,000 VPN exit endpoints compiled from WireGuard, OpenVPN, and IKEv2 gateway configurations across major commercial VPN providers and unlisted regional relays.

Cryptographically verified gateway configs
Directory Consensus

Tor Exit Relays

2,500 active exit nodes verified through Sentinel Tor Relay Intelligence and Anycast edge country classification.

100.00% exact consensus match
Session Attestation

Clean Control Set

77,500 clean IPs curated across tier-1 consumer fiber/cable ISPs (Comcast, AT&T, Verizon, Deutsche Telekom, BT, Orange, Vodafone) and 4G/5G CGNAT mobile carriers.

Verified human consumer traffic
System Architecture Decoupling

Stateless Edge Execution vs. Background Threat Grid

A common technical question: how does Sentinel achieve "stateless, sub-millisecond edge evaluation" while maintaining "temporal correlation and risk decay"? The answer lies in strict pipeline separation:

1. Request Runtime Path < 0.1ms Compute

Zero-Log Volatile Memory Architecture

When an API request hits our Anycast edge, evaluation executes within a V8 isolate that accesses a compiled, immutable binary threat store residing entirely in RAM.

  • ✓ Zero database queries during request evaluation
  • ✓ Zero disk writes or session persistence
  • ✓ In-RAM binary search & prefix matching in 0.0068ms mean compute
2. Asynchronous Ingestion Path 259,800+ Active Nodes

Sentinel Global Threat Grid Pipeline

Temporal correlation, velocity anomaly detection, and risk decay execute asynchronously in our background Threat Grid aggregation cluster.

  • ✓ Dynamic tunnel churn tracked across time series graphs
  • ✓ Inactive proxy IPs decay exponentially over 12–72 hours
  • ✓ Compiled immutable binary snapshot broadcast to edge isolates
Empirical Verification

Reproduce These Metrics Locally

We provide an open validation script that probes our live Anycast edge API in real time, validates each cohort, computes the confusion matrix, and prints Wilson score 95% confidence intervals:

scripts/benchmark_runner.py
# 1. Clone repository & execute live Anycast canary + benchmark harness
$ git clone https://github.com/antivpn-tech/sentinel-web.git && cd sentinel-web
$ python3 scripts/benchmark_runner.py
# Terminal verification output:
[0] LIVE ANYCAST EDGE CANARY VERIFICATION: 8/8 probes matched ground-truth (<1ms isolate compute)
[1] CONFUSION MATRIX & WILSON SCORE INTERVALS (95% CI)
Commercial VPN Gateways: 99.87% [99.84%, 99.90%] (49,937 / 50,000)
Datacenter VPS & Cloud: 99.98% [99.96%, 99.98%] (74,982 / 75,000)
Tor Exit Relays: 100.00% [99.85%, 100.00%] (2,500 / 2,500)
Residential Proxies: 98.36% [98.24%, 98.47%] (44,262 / 45,000)
Clean Consumer Broadband FPR: 0.045% [0.031%, 0.067%] (25 / 55,000)
Clean Mobile Cellular FPR: 0.142% [0.101%, 0.201%] (32 / 22,500)
> Aggregate Threat Sensitivity: 99.525% | Specificity: 99.926% | FPR: 0.074%