AI Analytics: mature approach to kids’ gaming
Developing games for children is challenging per se because the developers of the product are not its target audience. This is one of the reasons why not all projects for kids succeed, most of them hardly pay off. This can be easily fixed using AI for GameDev analytics. Today we’re going to tell you how.
All children love playing games and they actually do not care much about the gadgets they use for that. They would eventually play if there is a chance to do that and if the parents don’t mind. However, the market of computer games for kids has a lot of peculiarities. According to the experienced developers, there are three main hurdles:
- Lack of monetization at first — developers hardly ever get paid for the first levels of the game as parents need time to see whether the child is interested and willing to continue the game.
- Close attention to the gameplay — it’s not only children who matter but also their parents, various communities and, for sure, the authorities. For instance, it is mandatory to comply with GDPR and that is a source of many complications for the developers.
- Audience diversity — If adult players can be easily divided into age groups 20–25, 25–30 etc., children’s behavior varies from year to year. A three-year-old and a five-year-old would have totally different experience expectations from the game.
The role of artificial intelligence in the analytics of games for children ends up being more important than its role in universal projects and adult games. The trick is that AI has no personal experience and it doesn’t perceive the game as a person with their own opinion. A neural network uses specially developed algorithms for matching, revealing patterns and boosting the developer’s business.
Understand what the child wants
While creating games we use hypotheses. We assume that a child would be interested in collecting mushrooms, opening chests, catching some cunning mice etc. However, not all the hypotheses turn out to be right, and the sooner we realize that, the better it is for the business. AI allows us to track kids’ response to the new game levels and features, as well as to determine how interested in them our users are.
Moreover, such a personalized approach helps you determine areas of interest depending on different groups, such as countries, gender, age and others. Thus, working with neural networks, the game developer collects a whole database for personal game experience preparation for every little player in particular.
Get the feedback, which we don’t talk about
However, mistakes can occur to anyone and sometimes there are controversial points in the game.
For instance, gamers struggle with getting the crystals, despite their expectations in some episodes they have to wait too long or the boss of the third level is too strong. An adult player can write to the customer service, leave feedback on the market or do something else. Children hardly ever have the courage to do so and it is hard for them to express their thoughts clearly. In order to get feedback from them, you need to set up predictive and prescriptive analytics as well as a neural network to constantly monitor users’ behavior.
If you get to understand what every player is feeling at every moment of time (at Smartcoders we can do that), you are able to assess the attractiveness of any game moment and to fix bottlenecks if they appear.
Sum up the success
Children’s perception is different from adults’ perception, and often develeopers have no idea why one game was a success in New Zealand while German kids didn’t like it at all. Another game, which is popular with toddlers in Germany, is a failure in Great Britain. Supposingly, it’s all about some tiny details, but which ones? Should we change the character color to green? Should we switch the form of buttons and make them round instead of square or change the background color? AI allows us to test these hypotheses at minimal cost, and in case of the process automation we could even provide A/B-testing and collect real data on audience response.
On the way to monetization
Monetization in games for children is a challenging task. Firstly, children usually don’t have their own money, i.e. they need to persuade the parents to pay. Secondly, a child might not want to do this and just close the game. Thirdly, parents also need some strong arguments to buy something in the game. Commercial is not an option because it scares children off and many of them often close the games when they see some ad offers and videos.
AI is a tool that picks the most suitable moment for offering the user to purchase something as well as defines which content should be paid and which one should be offered for free for every child in particular. For instance, at Smartcoders we set up an analytics system which is in charge of predicting potential responses to buying offers for each gamer. Under close integration of the neural network with the game engine it is also possible to automate the choice of objects, levels and other content to sell or give to the gamer.
Retain the users
Since it is impossible to quickly monetize the game, in the field of video games for children we pay special attention to retaining the audience. According to the statistics, an average game session for children lasts from 3 to 15 minutes, less likely — half an hour, as children get tired faster. So it is crucial for us that the child comes back and continues the game. In order to obtain this result we apply push-notifications, special offers and other methods based on personalized recommendations. The better AI is integrated into the gameplay, the more precise the retention parameters are.
Ensure complying with GDPR
Not in all the countries parents would pay even for high quality content for their children, therefore many developers aim to promote their games in Europe. However, in that case it is required to ensure the compliance with the requirements of the new EU law on personal data protection — GDPR. According to the GDPR, the requirements to children’s data treatment are even more strict than for adults, and this category includes children up to the age of 16.
So here is a question: can we use AI to work with young gamers in the EU? Sure we can! There are two possible ways here. Firstly, you can request for the parents’ permission but this is not always convenient. Secondly, at Smartcoders we are able to work with anonymized information only, i.e. a gamer’s profile can contain only ID without any personal data. This approach helps you avoid any law violations (and considerable fines accordingly) ensuring personalized experience for the gamers at the same time.
Game engine integration with a neural network
All these approaches turn out to be the most efficient when the game analytics and the game design process itself are integrated with the neural network. AI helps us assess the efficiency of each and every action as well as evaluates users’ response to updates and suggests monetization methods for every player. At Smartcoders we aim to establish relationships with the game developers and take part in the projects not only as solutions providers in the field of neural networks but also as partners. So you can try our analytics right now, without any direct costs. Get our demo version at smartcoders.ai.
