Subscribe to Dr. Granville's Weekly Digest

Full Tilt Poker: Big Data Fraud Detection and Prevention

Full Tilt Poker is the second largest online poker room; it belongs to Rational FT which is a computer services company that develops software, business computer maintenance and performs operations based digital marketing. http://www.rfts.com/  


Two fundamental points of running a poker room online are the integrity of the site and the trust of customers. Players must be completely sure that they are protected against fraud. Negative publicity on this subject has a direct and immediate impact on customer loyalty, the ability to attract new customers and sales. The difficulty here is the volume of data to process, the complexity of the analysis, the difficulty of detecting all forms of fraud and costs. Depending on the activity, the rapid increase in data volumes and analytical requirements may involve increased infrastructure costs and lower overall margin.


Frauds are taking place in the online poker rooms and are not detected due to low analytical capabilities. The magnitude of the problem is often unknown. It is difficult or impossible to identify many types of fraud, collusion between players, the practice of multi-accounts, the bypass of the system to prevent unwanted disconnects, the use of assistive software, the physical player changing (so changing the playing style), using a Trojan horse to see the hole cards of opponents, etc.. ...

 
Fraud detection process requires a lot of data and time; we must work per iteration of question and answer with very long cycles to complete. In the end it is difficult to track down sophisticated frauds and many fraudulent actions are not identified. However, it is possible to identify players who often play together and are likely to be conniving, who have already made disconnections, who have played more than 500 hands, etc.


Using an analytical Aster Data platform, Full Tilt Poker had significantly improved its fraud tracking, particularly in being able to take into account in the analyzes detailed hands played. Earlier treatment of these data were long, transfer from one server to another, decompression via batch processing and final analysis took seven days to analyze a week of frauds at the rate of 1200 hands per second. With the Aster Data solution, transfers are removed, the treatments are massively parallelized on a set of low-cost servers and analysis of seven days of data takes 15 minutes at a rate of 140,000 hands per second.


More, the Aster Data analytic platform is able to run sophisticated algorithms and detect fraudulent actions in very large sets of events. Full Tilt Poker has observed that queries that took 90 ' take now 90''. Queries that usually do send back any result are now producing valuable responses in less than 1 hour. These capabilities have helped to provide access to data, via SQL queries, to all the analysts of the society.

 

In summary, the Aster Big Data solution identifies forms of fraud that were previously undetected, significantly reduce the analysis cycle time, shows extraordinary performance especially for the treatment of hands, allows deeper analysis of sophisticated fraud, better meets money laundering tracking requirements, provides players with new customization services benefits, improves internal reporting on activity poker room , while reducing the TCO of the system (hardware, maintenance, SAN, personnel, network).


Views: 1247

Tags: Big, Business, CRM, Data, Intelligence, Marketing, Media, Mining, Science, Social, More…Teradata, Warehouse

Comment

You need to be a member of Data Science Central to add comments!

Join Data Science Central

Follow Us

Videos

  • Add Videos
  • View All

© 2014   Data Science Central

Badges  |  Report an Issue  |  Terms of Service