IOP Conference Series: Materials Science and Engineering

PAPER • OPEN ACCESS

Development of Tic-Tac-Toe Game Using Heuristic Search

To cite this article: AM Zain et al 2020 IOP Conf. Ser.: Mater. Sci. Eng. 864 012090

View the article online for updates and enhancements.

This content was downloaded from IP address 178.171.1.217 on 10/07/2020 at 18:12

 

2nd Joint Conference on Green Engineering Technology & Applied Computing 2020 IOP Publishing

IOP Conf.
 Series: Materials Science and Engineering 864 (2020) 012090 doi:10.1088/1757-899X/864/1/012090



Development of Tic-Tac-Toe Game Using Heuristic Search

Zain AM1,2, Chai CW1, Goh CC1, Lim BJ1, Low CJ1, Tan SJ1

1
School of Computing, Faculty of Engineering, Universiti Teknologi Malaysia, Skudai
Johor, Malaysia
2
UTM Big Data Centre, Ibnu Sina Institute for Scientific and Industrial Research,
Universiti Teknologi Malaysia, Skudai Johor, Malaysia

E-mail: [email protected]

Abstract: Tic-tac-toe is a fun game played by two players. Before starting the game, a 3x3
square grid is formed using two vertical and two horizontal lines. The player who succeeds in
placing three of their marks in a horizontal, vertical, or diagonal row is the winner. In this study,
we develop a computer program where a player can challenge artificial intelligent heuristic
approach (opponent) to win the game. We want to prove that artificial intelligent heuristic
approach enables in giving a great performance for development of tic-tac-toe game.

1. Introduction
The tic-tac-toe is a unique game as illustrated in Figure 1. A 3x3 grid is formed by using two vertical
and two horizontal lines before the game starts. The players can fill the nine spaces with any two
different sign normally crosses (‘X’) and noughts (‘O’) symbols [10]. In this study, we use Visual Studio
2019 software to develop the tic-tac-toe program based on artificial intelligent (AI) heuristic approach,
min-max algorithm and alpha-beta pruning algorithm. When there are no player or less number of
players be able to win the game, there is an evidence that the AI approach is effective for tic-tac-toe
game development.

Figure 1. Tic-Tac-Toe game

AI is the simulation of human intelligence processes by machines, especially computer systems.

Artificial Intelligence exists when a machine can have human-based skills such as learning, reasoning,
and solving problems [1]. Nowadays, there is a lot of applications of AI in different fields. In astronomy,
AI can be very useful in solving complex universe problems. In healthcare, AI could give a better and
faster diagnosis than humans can be done. AI can also be used for gaming purposes, where AI machines
can be used to play strategy games like chess, where the machine can perform the best move in different

Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution
of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
Published under licence by IOP Publishing Ltd 1
 

2nd Joint Conference on Green Engineering Technology & Applied Computing 2020 IOP Publishing

IOP Conf.
 Series: Materials Science and Engineering 864 (2020) 012090 doi:10.1088/1757-899X/864/1/012090



situations. In this study, AI is proposed for computer gaming where it acts as a challenger (opponent) in
tic-tac-toe game.

2. Heuristics in Games
Heuristic search in AI is a technique to solve a problem faster than classic methods or to find an
approximate solution when classic methods cannot. This is achieved by trading optimality,
completeness, accuracy, or precision for speed. The heuristic search technique can evaluate the available
information and makes a decision on which branch to follow. The heuristic technique is capable to
produce a solution that is good enough for the problem. [2].
Nowadays, many effective heuristics methods has been successfully applied in various problem
domains. The heuristic methods are possible to be used in game development as they can help to inspire
a creative player experience. In computer games, player want to enjoy the games, and programmer needs
to have heuristic skills to guide the software for winning the game. Thus, the program needs to be
intelligent in order to make user (player) more exciting with the game.

3. Methodology
We firstly create the rules needs for Tic-Tac-Toe game. Then we study the theory of Minimax Search
and Alpha-Beta Pruning for tic-tac-toe game, logic for game development.

3.1 Rule-based Strategy

In the Tic-Tac-Toe game, the player is required to mark the nine spaces with any two different symbols
normally crosses (‘X’) and noughts (‘O’). The five rule-based strategy used in tac-tac-toe for player
winning the games is given in Table 1 [3].

Table 1. Rule in Tic -Tac –Toe

Rules Description

Rule 1 If the player has a winning move, take it.

Rule 2 If the opponent has a winning move, block it.

Rule 3 If the player can create a fork (two winning

ways) after this move, take it.

Rule 4 Do not let the opponent create a fork after the

player’s move.

Rule 5 Move in a way such as the player may win the

most number of possible ways.

3.2 Minimax Search Algorithm

Minimax algorithm is a kind of backtracking algorithm that is used in decision theory and game theory
[4]. It uses game theory, decision theory, statistics and philosophy to find the optimal move for a player,
assuming that the opponent also plays optimally. It is commonly being used in two-player turn-based
games such as Tic-Tac-Toe, Chess, etc. [11]
The player needs to fulfil two conditions in order to win a game. Firstly, the player needs to
maximize the chance to win the game. Secondly, player needs to minimize the opponent’s winning
chance. The principle of the minimax search algorithm is to find the optimal path to minimize the
maximum possible loss. Two possible results, + shows for computer wins, and - for computer loses.
The steps of the Minimax search algorithm are summarized below.
i. Construct the complete game tree
ii. Evaluate scores for leaves using the evaluation function
iii. Back-up scores from leaves to root, considering the player type:
x For max player, select the child with the maximum score

2
 

2nd Joint Conference on Green Engineering Technology & Applied Computing 2020 IOP Publishing

IOP Conf.
 Series: Materials Science and Engineering 864 (2020) 012090 doi:10.1088/1757-899X/864/1/012090



x For min player, select the child with the minimum score
iv. At the root node, choose the node with maximum value and perform the corresponding move

Figure 2 illustrates the minimax algorithm, where it evaluates the leaf nodes (terminating nodes
or at a maximum depth of 4) using the heuristic evaluation function. The algorithm continues evaluating
the maximum and minimum values of the child nodes alternately until it reaches the root node, where it
chooses the maximum value.

Figure 2. Illustration of the minimax algorithm

3.3 Alpha-Beta Pruning

Alpa-Beta Pruning algorithm is an optimization algorithm for the minimax algorithm. It reduces the
computational time by a huge factor. It allows faster search and even goes into deeper levels in the game
tree. It will cut off branches of game trees that does not require to be searched when there is a better
movement exists [5].
Alpha-beta pruning seeks to reduce the number of nodes that needs to be evaluated in the search
tree by the minimax algorithm. The alpha cut-off process shown in Figure 3, node C (MIN) cannot be
more than 1 since node D returns 1. Node B with value of 4 will not search the remaining children of
node C, as node A will certainly pick node B over node C for the max node. The remaining children can
be aborted if alpha  beta, for both the alpha cut-off and beta cut-off [12]. The steps of the Alpha-Beta
Pruning algorithm fundamentally is summarized as below.
i. Search down the tree to the given depth.
ii. Once reaching the bottom, calculate the evaluation for this node.(i.e. it's utility)
iii. Backtrack, propagating values and paths
iv. Attain the minimum score of the Alpha.

Figure 3. Illustration of the Alpha-Beta Pruning Algorithm

3
 

2nd Joint Conference on Green Engineering Technology & Applied Computing 2020 IOP Publishing

IOP Conf.
 Series: Materials Science and Engineering 864 (2020) 012090 doi:10.1088/1757-899X/864/1/012090



4. Development
This section discuss the development of the tic-tac-toe games using artificial intelligence heuristics as
illustrated in Figure 4. Firstly, we perform pre-production process include rules-based study, game
prototype and prototype testing. Secondly, we perform production process including programming and
game design. Finally, we perform post-production process including maintenance and testing [6,7,9].

Figure 4. Game development process

In the first phase, pre-production, we produce a proper prototype for game interface including
gameplay ideas and features. It acts as a proof of concept and to test ideas, by modifying some of the
game features. Later, we take the prototype testing result to produce the real game interface. This process
allows us to gain more understanding about the final interface and the features to let users be more
attracted by our game.
In the second phase, production, we focused on two parts: implementation of heuristic concept in
the source code and provide more detailed design for the game. In the programming part, we provide a
proper AI for perfect game [12,13]. Computer (opponent of game) needs to recognize game’s difficulty
level for all possible solution [14]. Then, we produce a design that enables the program fix with game
rules. We develop a design for sound effects where it is important to give impression for the game’s
delivery [8]. Then, we design attractive theme for the game and find right elements to suit the game
theme. We also apply background image for each of the scene. Example of game interface is illustrated
in Figure 5.
In the third phase, we perform post-production development process including maintenance and
testing. Maintenance is very important next to game production, where it is necessary for programmer
to recheck the source code and fix all the bugs produced after executing the game design. Then, testing
is a process to test the efficiency of the game discussed in next section.

Figure 5. Game interface

4
 

2nd Joint Conference on Green Engineering Technology & Applied Computing 2020 IOP Publishing

IOP Conf.
 Series: Materials Science and Engineering 864 (2020) 012090 doi:10.1088/1757-899X/864/1/012090



5. Testing
We recruit ten players in order to test the efficiency of tic-tac-toe game that we developed. Each game
consists of 3 rounds and the results are recorded in Table 2. Then, we generate a bar chart in Figure 6 to
demonstrate the winner of the game. We observed that there is no player won the game against the
computer (opponent). The results showed the results of 30 games (three trials given to each player), and
it was found that the players win, lose, and draw the games against the computer are 0, 5, and 25
respectively.

Table 2. Result of the testing

Player # Win Lose Draw No. of

games

1 0 1 2 3

2 0 0 3 3

3 0 2 1 3

4 0 0 3 3

5 0 0 3 3

6 0 1 2 3

7 0 0 3 3

8 0 0 3 3

9 0 0 3 3

10 0 1 2 3

Total 0 5 25 30

Figure 6. Bar chart for players’ results against AI

5
 

2nd Joint Conference on Green Engineering Technology & Applied Computing 2020 IOP Publishing

IOP Conf.
 Series: Materials Science and Engineering 864 (2020) 012090 doi:10.1088/1757-899X/864/1/012090



6. Conclusion
This paper has discussed and developed the tic tac toe game using artificial intelligence heuristic
approach. It was proved that min-max and alpha-beta pruning algorithms have extraordinary talent in
giving a great performance for tic-tac-toe game.

Acknowledgement
Special appreciations to editor and all reviewers on the useful advices and comments provided. The
authors greatly acknowledge the Research Management Centre, Universiti Teknologi Malaysia (UTM)
(UTM-TDR - vot. No. Q.J130000.3551.05G41).

References
[1] Russel S, Norvig P. Artificial intelligence: a modern approach. Pearson Education Limited; 2013.
[2] Pearl J. Intelligent search strategies for computer problem solving. Addision Wesley. 1984.
[3] Stuart R, Peter N. Artificial intelligence: a modern approach. 2003
[4] Moret BM. Decision trees and diagrams. ACM Computing Surveys (CSUR). 1982 Dec
1;14(4):593-623.
[5] Fullerton T. Game design workshop: a playcentric approach to creating innovative games. CRC
press; 2014 Mar 5.
[6] Brathwaite B, Schreiber I. Challenges for game designers. Nelson Education; 2009.
[7] Bethke E. Game development and production. Wordware Publishing, Inc.; 2003.
[8] Prasad B, Reddy MS, Reddy VS, Kishore RR. Game Development-Bounty Rescuestep. Empirical
Research Press Ltd.. 2017.
[9] Siegel AN. Combinatorial game theory. American Mathematical Soc.; 2013 Aug 1.
[10] Motreanu D, Panagiotopoulos PD. Minimax theorems and qualitative properties of the solutions
of hemivariational inequalities. Springer Science & Business Media; 2013 Dec 1.
[11] Knuth DE, Moore RW. An analysis of alpha-beta pruning. Artificial intelligence. 1975 Dec
1;6(4):293-326.
[12] Sarkheyli A, Zain AM, Sharif S. A multi-performance prediction model based on ANFIS and new
modified-GA for machining processes. Journal of intelligent manufacturing. 2015 Aug
1;26(4):703-16.
[13] Zhou KQ, Zain AM, Mo LP. A decomposition algorithm of fuzzy Petri net using an index function
and incidence matrix. Expert Systems with Applications. 2015 May 15;42(8):3980-90.
[14] Deris AM, Zain AM, Sallehuddin R. Overview of support vector machine in modeling machining
performances. Procedia Engineering. 2011 Jan 1;24:308-12.