Message us t.me/qshepel
Poker integrity • Bot detection • Collusion & chip dumping • Investigation-ready evidence

Security & integrity analytics for online poker networks.

PPPlatform Software helps poker operators detect automated play, multiaccounting, and coordinated abuse using statistical and machine-learning analysis of hand histories—supported by expert review to build a clear, investigation-ready evidence base.

Designed for real operations: actionable flags, explainable reasoning, and workflows your security team can use.

Services built for poker security teams

We combine statistical signals, pattern analysis, and expert validation to help operators detect and act on automated play and coordinated abuse—while keeping results explainable and review-friendly.

Bot Detection (Stats + ML)

Hierarchical clustering on normalized multi-stat distance, plus confidence interval and “range” analysis to reveal algorithmic consistency that humans rarely match.

Bet Sizing Similarity

Visual pattern dynamics across players and groups to detect abnormal bet sizing behavior—often a strong signal when combined with stats and schedule.

Interaction-Based Detection

Advanced signatures focused on automated state reading and action input. Known signatures can be identified quickly; new ones may be detected after ~1000 hands in some environments.

Multiaccounting Detection

Multiaccounters become obvious with enough hands: near-identical statistics + overlapping schedules. We use the same clustering foundation enhanced with schedule analysis to surface related accounts.

Collusion & Chip Dumping Investigation

Automated penalty scoring (e.g., strong hand folds, suspicious 3-bets), threshold-based flagging of pairs, and money-stream summaries—built to accelerate manual review.

How the analysis works

The workflow is designed to be reproducible and explainable: compute features, detect anomalies, form groups, and build evidence artifacts that investigators can validate quickly.

Detection primitives

Clustering on multi-stat distance Distance matrix across many normalized poker statistics, then hierarchical clustering to discover suspicious groups.
Confidence interval + range analysis Detect groups whose stat values are unusually narrow and synchronized—often a sign of automated decision policies.
Bet sizing patterns Compare sizing dynamics against typical winning player behavior for the network and game type.
Schedule fingerprints Spot pre-programmed patterns and cross-account similarities suggesting coordination or shared control.

End-to-end delivery

01

Ingest & feature build

Parse hand histories, compute a broad set of poker statistics, plus bet sizing and schedule features.

02

Automated detection

Find suspicious outliers, synchronized changes, and tight-value clusters not typical of human play.

03

Expert validation

Review signals and build an evidence bundle: why it’s suspicious, what patterns recur, and where.

04

Actionable output

Flagged accounts/pairs, summary tables, charts, and investigation notes—ready for enforcement workflows.

Outliers Similar stats Bet sizing Schedule Penalty scoring Money streams

Evidence examples from sample reports

Below are example scope numbers and evidence patterns reflected in the provided PDF samples.

Large-scale processing

Demo scope example: 113,068,657 processed hands across 32,446 players, with 63 identified as suspicious in the sample.

Group-level similarity

Reports highlight groups with many synchronized statistics—sometimes nearly identical values across many dimensions, which is highly improbable for independent human play.

Expert reasoning

Expert validation emphasizes combinations of signals and recurring signatures—not single metrics in isolation.

Real case narrative

The case-study sample describes an operator engagement where a suspicious group was identified by algorithms, validated by experts, then accounts were closed and funds confiscated for AI use (as described in the sample).

Investigator-friendly outputs

Collusion/chip dumping modules use penalty points, thresholds, date grouping (moving toward session-based grouping), and can produce money-stream visuals and investigator reports.

Note: these are sample documents; thresholds and modules can be tuned to your network’s games, player pool, and enforcement policy.

PDF report samples (embedded)

Put the PDFs in the same folder as this index.html. The viewer below will load them locally. Use the buttons to switch between documents.
Security Services Overview (2020)
If the embedded viewer doesn’t work in your browser, use “Download PDF” and open locally.

Ready to reduce botting and abuse?

Message us on Telegram and share: game types (NL/PLO), table sizes, timeframe of hand histories, and your enforcement goals. We’ll reply with an engagement plan and expected deliverables.

Bot detection Multiaccounting Collusion Chip dumping Reports & dashboards
Direct contact
Telegram: t.me/qshepel
Start chat
What to send: links to sample hand histories (or export format), plus your preferred reporting cadence (one-off report vs recurring monitoring).
We adapt thresholds and outputs to your policy. Deliverables can include flagged accounts/pairs, summary tables, and visual evidence to accelerate manual review.