IPTV SLEUTH — PROVIDER INTELLIGENCE — v1.0
PROVIDER ANALYSIS SYSTEM
SYS ONLINE

How to Use IPTV Sleuth

1. Submit a Provider

Go to the Submit & Compare tab on the home page. Fill in the provider name and as many stream/category counts as you can. These numbers are typically visible in your IPTV app's dashboard or playlist info screen.

2. Review Matches

After submission, IPTV Sleuth compares your provider against every public provider in the database. Matches above 75% similarity are shown. If a match scores 95% or higher, it is flagged as a likely reseller of the same source panel.

3. Check Grouped Matches

The Grouped Matches tab shows clusters of providers that share 95%+ similarity. Each group represents providers that likely share the same underlying source panel.

What Signals Are Compared

When Xtream API credentials are provided, IPTV Sleuth builds a multi-signal fingerprint using 22 weighted signals across content, infrastructure, and metadata dimensions. No credentials are stored — only the resulting fingerprint data.

Content Signals

These compare what the provider actually serves. They carry the most weight because resellers share the same content catalog.

SignalMethodWeight
Stream IDsMinHash Jaccard over all live stream IDs15%
Channel NamesMinHash Jaccard over normalised channel names12%
EPG IDsMinHash Jaccard over EPG channel IDs8%
VOD Stream IDsMinHash Jaccard over VOD stream IDs8%
Category IDsExact Jaccard across live, series, and VOD category IDs8%
Series IDsMinHash Jaccard over series IDs4%
Category NamesExact Jaccard across all category names3%
Channel NumbersJaccard over channel number assignments2%
Infrastructure Signals

These identify the underlying server and delivery network. Matching infrastructure strongly indicates a shared source panel.

SignalMethodWeight
Backend Stream IPsJaccard over resolved content delivery IPs7%
Server URL / IPIP match (100%), hostname match (90%), or domain match (50%)5%
SSL CertificateCertificate serial hash match or SAN overlap5%
Port TupleExact match of HTTP, HTTPS, and RTMP port numbers4%
EPG SourceExact match of EPG/XMLTV source domain2%
HTTP HeadersExact match of server software and cache headers2%
Detected BackendPanel software match (XtreamCodes, Stalker, etc.)2%
Stream Token FormatExact match of authentication token type (JWT, MD5, etc.)1%
Temporal & Metadata Signals
SignalMethodWeight
TimestampsJaccard over sampled "added" timestamps4%
Icon DomainsJaccard over stream icon/cover image domains3%
User MessageExact match of server broadcast message2%
Allowed FormatsJaccard over allowed output formats1%
Derived Signals

Computed from the raw data to capture structural patterns.

SignalMethodWeight
Content AgeCosine similarity of weekly content-age histograms1%
Category DistributionCosine similarity of streams-per-category counts1%

When content signals strongly agree (≥90% average), infrastructure mismatches are automatically discounted — different servers are expected for resellers sharing the same source panel.


Fallback: Count-Based Comparison

If no Xtream credentials are provided, the system falls back to a simpler comparison using content counts only:

MetricWeight
Live Streams30%
Live Categories20%
Series20%
Series Categories15%
VOD Categories5%

If none of these are provided, the system falls back to legacy metrics: Channels (60%) and Groups (40%). Count-based comparison is far less accurate than fingerprint comparison.

How Similarity Scoring Works

Fingerprint mode computes a weighted average across all 22 signals. Each signal produces a similarity score between 0 and 1. Signals where both providers have data are included; missing signals are excluded and their weight is redistributed proportionally.

Count-based fallback uses a simpler formula per field:

similarity = 1 - |value_a - value_b| / max(value_a, value_b)

The overall score in both modes is a weighted average expressed as a percentage:

  • 95%+ — Very likely the same panel/source (flagged as match)
  • 75-94% — Possible reseller or similar provider
  • Below 75% — Likely different providers

Privacy

We do not store credentials, login details, or any personal information. Only aggregate metrics (stream counts, category counts) are saved. Your IP address is recorded with submissions solely to prevent abuse and is not shared.