Around 40 years worth of Cricket World Cup data is being mined to produce insights that enhance the viewer's experience at this year's tournament.
One of the International Cricket Council's (ICC) main objectives is to deliver real-time, interesting, storytelling stats to fans through the Cricket World Cup app or website.
The ICC is using the SAP HANA Cloud Platform and SAP Lumira software to analyse statistics on scores, player performance, player profiles and more. This information is updated every 20 seconds.
Players are what fans obsess most about so churning out information on each player's performance is a big priority for ICC. In one data analysis exercise, the ICC ranked the top 50 batsmen and bowlers. Nine data elements, or criteria, were created for assessing players.
These included the total runs scored by each player, how quickly they scored the runs, the total 4s and 6s they hit, and how many times a player scored less than 10 runs.
“During this World Cup we've seen an increase in big scores - we've seen a world record or a first time in World Cup history, a double century scored by Chris Gayle. And not just individual players, but team totals,” said former Australian cricketer, Adam Gilchrist.
The ICC and SAP have analysed the data to see how important a team's bowler is in winning games of cricket. South Africa, New Zealand and India have dismissed teams the most number of times in this year's tournament. This is what makes them tough competition and likely to win the World Cup, said Jenni Lewis, a data scientist at SAP.
"I think the team that will win this World Cup will be the team with the most well-equipped bowling line up, not [necessarily the team with] the power-packed batting [lineup]," said Gilchrist.
Gilchrist said the bowling line up that can take wickets, particularly in the last 10 overs, to prevent teams from mounting big 400 plus scores, are doing well.
"It's not the flying starts, it's the flying finishes," Gilchrist said.
The ICC and SAP also used data to find similarities between skilled players using seven different characteristics. In this year's World Cup, it was discovered that countries which are not usually recognised as dominant cricket nations, such as the UAE and Ireland, were found to have strong performing players with similar characteristics.
The ICC and SAP want to not only use this data for delivering interesting stats to fans but also improve team performance and strategies out on the field, to ultimately increase a particular team's chances of winning. Predictive analytics is key to this as it forecasts particular outcomes for events that take place within the game, looking at past patterns in the data and factoring in many variables.
For example, is a particular player more likely to score more runs at the beginning of his innings than if he takes some time to settle in? Knowing the answer to this question gives a team insight to rethink their strategy for the next game.
The ICC also plans to mine social media data in the future to gain insights into what fans are most interested in when tweeting about a game, popular players and teams.
Rebecca Merrett travelled to Melbourne as a guest of SAP.