Games development

It is a extremely complicated endeavor to create a popular game. The process involves a lot of hours of highly trained specialists as well as wide array of technologies. Most game development companies already use: game engines, compilers, source control and task tracking software. Up until recently, AI was used to control in-game characters, but the game development itself can benefit greatly from cognitive computing innovations.


Gameplay optimization

Modern computer games have highly complicated game rules. Consider for example StarCraft game or DOTA2. It requires many hours for a player to learn even the basics of the game. Balancing such design (all the parameters, variations and possibilities) is a colossal work. Typical balancing process if following:

  1. Create a build for testing.
  2. Play the game for a while.
  3. Identify a problem of game balance.
  4. Adjust some parameters.
  5. Repeat the process until quality is acceptable.

This is slow, tedious and inefficient process. Even top companies with considerable development budgets and top developers at their disposal, often release games with imperfect game balance...


Imagine the convenience if it was possible to connect your game to Artificial Intelligence machine, define parameters for optimization, set clear goal metrics and then - allow it do the actual balancing work! Results would be objective, measurable and much faster.

Diablo3 difficulty window
Diablo3 difficulty meaning


Game development process is already complicated as it is, but adjusting it for every possible player type - is almost impossible. Developers are looking for practical ways to make the game appealing to different groups of players, but the solutions are very rough and inconvenient. It is common to see following choice in a game:

  • Easy
  • Normal
  • Hard

How can a player know in advance, which choice will bring best experience?

Also consider the diversity of players. Some have great aiming (shooting) skills, but poor puzzle solving (or navigational skills), while others can be a complete opposite. All players want a challenge as too easy games are boring while too high difficulty is frustrating.


Imagine the convenience if it was possible to allow Artificial Intelligence to observe a player in real-time, learn particular skill-set and adjust the game to bring optimal experience without forcing the player to make explicit choices.