Tactics and strategy
The first thing that I will be
talking about in this final section of my assignment, under the topic of AI in
games, is strategy and tactics. I will start with strategy then I will move on
to tactics. So, in strategy, each character of the enemy group normally have
their own decisions to make. Also they have their own algorithm that
affects the way they move around the environment. But the enemy groups overall
decision will normally be made by the group strategy, this can be that they
team up together to overwhelm the player in order to try and eliminate
them (Gallear, 2017). Also it gives the player a challenge. A game example of this would be
Half-Life in which the enemies form a team to take the player out (Gallear, 2017). Now
I will talk about tactics, this can be things like the enemies hiding behind
cover to surprise attack the player or they could use the shadows to hide
in and surprise the player. They might even use the shadows as a sniper point
to take out the player from a distance or the sniper point can be behind
cover for a surprise attack from afar. So the advantage of strategy
and tactics from a game point of view is that it makes the AI characters
unpredictable and therefore makes them more realistic and again helps to
increase the player experience. The disadvantage is that if they do make a team
the overwhelming number of enemies attacking can be impossible to complete the
game making it annoying and can damage the player experiences.
Learning techniques
The second thing that I will talk about in this final
section of my assignment, under the topic of AI in games, is Artificial
Neural Network and Q Learning. I will start off with Artificial Neural Network
then I will move one to Q Learning. Artificial neural network is,
in simple terms, a program that is designed to simulate the human
brain. It learns based on what it has experience before and how the player
plays, adapting based on that, to counter the players playing style and
take down the player, increasing immersion and giving the player a challenge. Also ANN
has neuron nodes that connect altogether creating a web, they are
similar to the hundred billion neuron cells found in the human
brain, so in many ways it work like a human brain does (tutorialspoint, 2017). An AI character
using this program will try something such as attack the player from the
front, which isn't the best idea so it tries and fails but then it learns that
the player favours the front which means it will try something to counter
this like a sneak attack from behind. Then there is Q learning which is another
learning techniques that is used in AI. In this
learning technique the agent, IA Entity or even bots look back at the
history of interactions with the environment to find the best solution to get
to the goal required (Artificial Intelligence , 2010). For example there could be five rooms in a
building and the outside of the building is the goal, so to get to
this the agent will look at the history of the route it has already taken
to get to the goal and find the best one. So I will put the advantages and
disadvantages of these two learning techniques. The advantages of these
learning techniques is that they give the AI character the ability to lean
which again make the AI character feel human and makes it more realistic,
helping to increase the player experience and the immersion. On the other hand
as they are learning from their experiences or history this makes them more
effective opponents and therefore it can be really hard on the player which
could drop that player experience (mnemstudio, 2012).
 |
| (mnemstudio, 2012) |
Comments
Post a Comment