When the team offensive, for the players, it (temporarily with it) has a behavior guidance in every moment, which is the next step:
Is it a ball to a direction? Or do you pass someone to someone in your teammate? Or is an immediate shot. Where does this guidance come from? How to root
Is it possible to change? Regardless of the situation, make the players can have a relatively reasonable next behavior at any time, which is probably
Simulated Ai playing football.
In fact, the key to football AI simulation is that the player will make the player to analyze the changing competition in the court, and draw one according to this judgment
A more intelligent and reasonable next behavior. A good player will continue to analyze, so that it quickly adjusts its behavior to make the game to be more favorable.
The tendency to develop, this process continues, and a high-level football match for computer AI simulation can be achieved.
The first difficulty is how to teach the AI player to judge the situation on the field, how to generate AI analysis, how to conclude and generate the next step.
The first step is visual judgment
With the ball players, they can see teammates and other behaviors, from it to it, from it, no matter in front of it or even behind it, excellent
Dai's players can perceive. That is, simulation makes the AI player "open vision". These information can be very convenient from the coordinates of the players in the SIM competition.
The action data is to simulate it is not too difficult.
The second step is to analyze
That is to say, the player has an intent in the collection of visual information, so that the competition can make it more developed more. (For example: teammates are not
Ok, I can't wear the ball, AI makes it a intention to continue the ball; when it is not too far from the gates, there is a certain shot empty block, AI makes it a spherical shot
The intention of the door; when there is a teammate's running position, AI makes it a intention to shot a shot to it; when the two side teammates follow into the position, AI makes it
It is the first step in the first step to generate the first step, which is the first step. It is the prototype of tactics.
From these extremely discrete visual acquisition information (that is, those enemy parties players coordinates, speed, behavior, and even ability) have a tactical intent
It is very mysterious and mysterious. As everyone knows, smart human beings can refine the useful use of very much hidden from these discrete and random data.
Competition information! The task of AI simulation is to let COM players learn to find these hidden information! So you can think (use experience rules)
To) two ways to interact with the inference style (the best behavior to derive the best behavior).
Another important factor is to "filter" a lot of hidden information, that is, AI needs to find the most valuable (or right) hidden
Tibet information, there is a random factor here, which also has the right to sort.
The last step is to form a next step according to the tactical intention
This is like a detail process because the next behavior is often changed frequently, and the tactical intent is relatively stable. (complete
It is very much more, and there is a short walk. The next step is the "action engine element" of this SIM football, and reasonable action elements can make the simulation.
The visibility of the competition is more real, but it is more important to cooperate with AI, which can be said to be complementary. FIFA's defects
It doesn't have a reasonable realistic true in its action engine element, which directly leads to its authenticity is not as good as we, and the AI is no longer good.
Summary: AI players find out the hidden information of discrete data through the observation, and have made a tactical intention, and make a movement behavior.
It seems to have completed a TURN. The remaining, Sid Meier has a very useful experience: "I will teach it how to play a round. Then teach it how to play two rounds, and then consider how to play 10 rounds." Of course , SIM football AI may be more complicated, because we only consider
An example of a player AI simulation, but I believe that these three parts are applicable to every situation on the football field. It is only more in-depth analysis.
Experience, feelings.
