The games we play and the data that drives them
Successful video games nowadays go far beyond an engaging storyline and fancy graphics. They are shaped by a universe of player data such as interaction time, quitting points, average revenue per user and more. These packets of intelligence can be harnessed to tweak game design and development, enrich the customer experience and keep players engaged. They can also be used to keep players and their information safe by detecting and preventing fraud, as well as drive increased revenue for games companies.
But, in order to be effective, companies must have the ability to collect, analyze and make sense of a massive amount of data, which requires the right formula and analytics tools. This is especially true when it comes to using that intelligence to enhance the player experience.
Putting data to work
One of the earliest reasons for massive amounts of data collection spun out of the need to identify “farming” points, explains Dr. Christian Sebastian Loh, a professor of learning system design and technology and the director of the Virtual Environment Lab at Southern Illinois University.
Farming is a tactic, often used in MMOG (massively multiplayer online games), where players congregate at a certain area on a game map that continually generates enemies or items. Those players can then repetitiously gather points or in-game currency. It’s not always a negative thing, says Loh, but game creators still like to keep tabs on these areas of interest in the game. They do so by tracking the number of players that gather there and analyze whether or not excessive farming is causing an unfair advantage or dissatisfaction among gamers. In addition to highlighting behavioral patterns, data is also helpful in identifying bottlenecks. For example, big data blog Datanami writes that King Digital Entertainment, the makers of Candy Crush Saga, turned to data analysts after it found a mass exodus at level 65 in the game. It was a major problem for a game with 725 levels. After King identified the specific gaming element preventing players from clearing the level and deleted it from play, customer retention instantly saw a spike.
But game and player data isn’t just useful for games companies, it can also benefit players. Take the latest iteration of Mario Kart for Nintendo Switch. The game allows users to not only pick a character, but also choose their vehicle and tires. A recent article in Medium showcased how data analysis helped identify the 15 most optimal combinations, from a total of nearly 150,000 possibilities. For the Mario Kart fanatic, this type of big data analysis is a player’s dream!
Powering up revenue per player
One of the key applications of big data for gaming companies is finding ways to increase revenue. Boosts (e.g. speed boosts, paid power-ups, outfit upgrades, etc.) are a prime example. Data can help point out which boosts are best sellers, how they’re being used and can then model future power-ups or add-ons for the game.
According to Datanami, when Farmville maker Zynga mined its player data, it discovered that they were willing to pay for animals that originally started off as purely decorative items, using either real money or virtual coins earned in the game. The company even added a scarcity by creating “rare species” to further pique players’ investments.
The ability to use data to understand players’ preferences helps games companies tailor the experience, create more effective in-game advertisements and know what price structures — freemium, pay-to-play, free-to-play — will draw the most interest for the game.
Rooting out fraud in gaming
With the increased monetization of games comes an increased risk of fraud. According to Loh, it’s not unusual for players to suffer the consequences of in-game scams and hackers. He points out that when it comes to MMOGs, there is a whole sub-industry of fake players (sometimes bots, he says) that will “farm” a certain item or resource, then turn around and sell it for real money.
When there are millions of players playing a game, odd behavior from a small minority of people can be hard to detect quickly. And, because games companies process such high volumes of tiny microtransactions, payments fraud also has the potential to go undetected for a long time. These are two prime examples why the games industry leans heavily on the speed and accuracy of machine learning and other AI and big data solutions to help in identifying suspicious trends.
Completing the picture
The way a player interacts within a game is only one part of the data equation. External information and insight from outside sources — everything from linked social media accounts to comments under streams or forums like Reddit and Discord — work together to provide game developers with a 360-view of players’ likes, dislikes and preferences about a game.
For example, the gaming community managers at Bungie — the developer behind uber-popular online shooter game Destiny — often take to social media to address concerns about their game and make adjustments based on that feedback. Having that one-to-one link with players allows games companies to collect anecdotal information on how they feel about certain aspects of the game. That depth of insight can’t be collected through big data, but must be done by high-quality player support and customer service teams that complement a data-driven solution.
This type of high-tech, high touch approach paints a more complete picture of the player, opening the door to stronger player engagement and better overall customer experience.