Banks, retailers, Google, the NSA, they all lover their data. But one of the more intriguing use cases for Big Data involves the gambling industry, a sector that’s so reliant on number crunching and intelligent prediction that it’s been left with little choice to grab a hold of its masses of data and exploit it to the max.

The first adopters of Big Data in the betting game were, unsurprisingly, the bookies themselves. UK bookmaker Paddy Power‘s experience is one that mirrors that of many leading betting sites, when the rise of online gambling meant that it was accumulating hundreds of terabytes of data that it had little idea what to do with.

“We had a start-up culture. [In BI], we managed more than we thrived,” Conor McMenamin, Paddy Power’s business intelligence manager, explained to ComputerWeekly.

The problem with online gambling is that it’s a hugely dynamic business, with odds constantly evolving in real-time. This makes standard data warehousing techniques redundant, as bookies need to be able to access their data instantaneously. Paddy Power got around the problem by adopting QlikView’s business intelligence solution for live data based on Microsoft SQL Server and SQL Server Integration Services with Excel-based clients.

“It was about putting in place a scalable platform that we could develop on quickly, and not be constrained by legacy issues,” McMenamin adds.

Paddy Power’s self-service BI model meant that its decision makers had wider access to the company’s data in way that keeps its IT managers in control of governance and data structure, whilst affording managers more time to focus on their strategic goals.

Betting firms aren’t the only one’s using Big Data in the gambling sector. Some services are also looking at ways in which data can be delivered to the gamers themselves, providing insights that enticing them to stick around and, presumably, increase their chances of cashing out. This is the promise of poker stats firm SharkScope, which recently deployed the TokuDB’s performance engine to optimize its database performance.

SharkScope claims to be the largest online poker tournament database in the world, and as such it accumulates vast amounts of data from millions of poker games played online every single day. The idea is that players can use SharkScopre to track their poker stats and improve their gameplay whilst avoiding the ‘sharks’ who’ll quickly eat up their cash. Until recently it got by with its old MySQL database, but in the fast paced world of online poker data needs to be accessible almost instantaneously. With some queries taking up to 20 seconds to process, this was clearly far too long, leaving SharkScope with choice but to find a faster solution.

Using TokuDB, Sharkscope’s MySQL database is now capable of ingesting data from over a million games every day, spewing out insights for a diverse set of requests in something that’s about as close to real-time as it’s possible to get – just under two seconds, according to TokuDB.

“TokuDB gave us faster search results, a much smaller database and fewer servers to manage,” said Steve Mayes, CEO SharkScope. “Because of Tokutek, SharkScope can continue to add new subscribers for the foreseeable future.”

Okay, so what if you could accurately predict the outcome in nine out of every ten English Premier League soccer matches? You’ll probably think you’re on the cusp of striking it very, very rich indeed, if only it were true.

Well, it might just be that your dreams really can come true if you’re a devoted soccer betting fan, with a new startup going by the name of Betegy confidently boasting that its capable of doing just that. Betegy’s site says that its able to successfully predict the outcomes of 90% of all English Premier League matches via its complex algorithm that considers every possible factor that might affect the game, from the coach’s birthday to the weather on the big data.

Based in Poland, Betegy actually offers forecasts for 21 different soccer leagues around the world, and claims a 76% success rate in its predictions. There are some leagues in which it does notably better however – such as England where this rises to 90%. In other leagues, its strike rate might be as low as 50%.

How does it do it? According to CEO Alex Kornilov, betegy uses two layers of data to predict the outcome of soccer matches. The first layer includes basic stats such as recent performances, history between the two teams in question, form and the average number of goals. The second layer goes deeper into it, taking into account factors like the weather and other details that may affect player’s motivation.

Betegy doesn’t claim to be perfect, but with only three possible outcomes in soccer – win, lose and draw – it’s 90% success rate stacks looks pretty tempting seeing as choosing an outcome at random would normally only result successful predictions 33% of the time.

So there’s every reason to play with Big Data if making money is your number one priority.