How AI concepts like FSM, MCST and AR are used in gaming?

We are influencers and brand affiliates.  This post contains affiliate links, most which go to Amazon and are Geo-Affiliate links to nearest Amazon store.


The original half life game was among the most talked about news in the gaming world because of its unique AI (artificial intelligence) with exceptional squad behavior. Gamers were quite taken with the way how the SWAT teams used to speak with one another in the game. Therefore you find AI being implemented in whichever type of game you play. It may be a racing game like Need for Speed: Rivals or real time strategy games like Age of Mythology, Caesar 4 or stealth game like Hitman; there are elements of AI embedded in the games.

AI can play multiple roles in the game. To govern the behavior of certain character entities in the game there are general set of rules that have to be followed. Don’t you know that the characters and objects in pre-scripted events follow a certain type of AI? In the second mission in Hitman Codename 47 game the way how the mafia car comes and goes in a particular time is an example of a pre-scripted event. If you want to learn artificial intelligence concepts then Intellipaat’s deep learning course with Tensorflow is there for you. Intellipaat is one of the leading e-learning institute which provides competent training on emerging technologies.

Finite State Machine (FSM)

Regardless of how AI is used it is used to improve the fun in gaming. Controlling non-player characters (NPCs) in the game is one of the most important role of AI in game development. Game developers use various methods which make these NPCs more intelligent. Generally an NPC has to be programmed in such a way that it is able to react to every situation it encounters. Game developers accomplish it using the Finite State Machine (FSM) which was conceptualized way back in 1990. This machine is used to generate a specific reaction for every action it encounters in the game. For every action from the human’s side there would be a pre-programmed reply from the AI built on FSM model. Take hitman games as an example where a guard NPC would remain calm when hitman acts normal and as soon as hitman shows suspicious behavior it starts shooting him. However there are let downs of this approach which is its predictability. When a gamer plays such FSM modeled games for a few times he knows exactly how the AI would react and therefore the surprise and fun would all be lost.

MCST (Monte Carlo Search Tree):

MCST algorithm is more advanced as it uses the random trials to make its move. You might have heard how world chess champions are defeated. Well MCST may be said as the reason for it. Deep Blue modeled on MCST defeated the world chess champion in 1997. In this algorithm the AI would calculate all the possible moves it can make and all the possible moves the human player would make in response and then would see how it should respond to those moves. This set of possible moves spawning another set of moves can be imagined as a search tree. After putting forth a real move the search tree would be processed again based on the available moves. Hence MCST can calculate thousands of moves and pick one which has the most reward.

Augmented Reality (AR)

While virtual reality is placing the user in the system generated 3d environment, augmented reality is where a computer image is superimposed onto the view of the real world as seen through a mobile device typically. Pokemon Go is the best example for this technology where gyroscope, GPS, satellite services and mapping technologies are used which are then combined with location data, Nintendo’s familiar characters and landmarks. Users could catch pokemons with poke balls with the real world picture behind it. AR in gaming will continue to grow and advertisers can leverage on the location-based data as they have seen how Pokemon Go has succeeded. The challenge for companies developing AR games is that they have to allow advertising without compromising on the gaming experience required by the gamers.


Did you know that diabetic people can get great relief playing Pokemon Go all the while having fun? From Super Mario to GTA Vice city, AI is in play to make gaming more fun to the users. Metro kids of this century spend a lot of their time in gaming. Gaming companies are trying to make actions of NPCs and objects less obvious and more reactive as possible. The equation has changed now in the sense previously the AI was used to react to a situation at hand but now they are increasingly being programmed by gaming companies to engage proactively in the game being played. It is possible that in future where AI could essentially provide near humanlike behavior in games. Imagine how wonderful would gaming then be.

Author Bio:

Vaishnavi Agrawal loves pursuing excellence through writing and have a passion for technology. She has successfully managed and run personal technology magazines and websites. She currently writes for, a global training company that provides e-learning and professional certification training.
The courses offered by Intellipaat address the unique needs of working professionals. She is based out of Bangalore and has an experience of 5 years in the field of content writing and blogging. Her work has been published on various sites related to Big Data Online Training, Business Intelligence, Project Management, Cloud Computing, IT, SAP, Project Management and more.

Generally what comes to the mind of people is that computer-controlled characters in games is where AI is used. But that is only one among many ways how AI is used to develop games.

We are influencers and brand affiliates.  This post contains affiliate links, most which go to Amazon and are Geo-Affiliate links to nearest Amazon store.