Description
1 Overview
For this assignment you are required to write a Java program that plays (n, k)-tic-tac-toe; (n, k)-tictac-toe is played on a board of size n × n and to win the game a player needs to put k symbols on
adjacent positions of the same row, column, or diagonal. The program will play against a human
opponent. You will be given code for displaying the gameboard on the screen.
2 The Algorithm for Playing (n, k)-Tic-Tac-Toe
The human player always starts the game. The human uses ’X’s and the computer uses ’O’s. In
each turn the computer examines all possible plays or moves and selects the best one; to do this,
each possible move is assigned a score. Your program will use the following 4 scores for moves:
• HUMAN WINS = 0: this score is given to a move that will ensure that the human player wins
• UNDECIDED = 1: this score is give to a move for which it is not clear which player will win
• DRAW = 2: this score is given to a move that will lead to a draw (i.e. no player wins the game)
• COMPUTER WINS = 3: this score is given to a move that will ensure that the computer wins
For example, suppose that the gameboard looks like the one in Figure 1(a). If the computer plays
in position 8, the game will end in a draw (see Figure 1(b)), so the score for the computer playing in
position 8 is 2 (DRAW). Similarly, the score for the computer playing in position 9 is 0 (HUMAN WINS).
A configuration is the positioning of the symbols on the gameboard. For example, the configuration shown in Figure 1(b) corresponds to a game that ends up in a draw. The configuration in
Figure 1(c) corresponds to a game won by the human player. Each configuration is also assigned one
of the 4 above scores. So, the configuration in Figure 1(b) gets a score of 2 and the configuration in
Figure 1(c) gets a score of 0.
1 2 3
4 5 6
7 8 9
X
X X
X
O
O
O
(b)
1 2 3
4 5 6
7 8 9
X
X X
X
O
O
O
(c)
1 2 3
4 5 6
7 8 9
X
X X
X
O
O
O
(a)
O O X X
Figure 1: Board configurations.
Note. You can skip the rest of this section and Section 2.1, as you do not need this to complete
the assignment. However, if you are interested in knowing how a computer program that plays a
2-player game works, then keep on reading.
To compute scores, your program will use a recursive algorithm that repeatedly simulates a move
from the computer followed by a move from the human, until an outcome for the game has been
decided. This recursive algorithm will implicitly create a tree formed by all the moves that the
players can make starting at the current configuration. This tree is called a game tree. An example
of a game tree is shown in Figure 2.
Assume that after several moves the gameboard is as the one shown at the top of Figure 2 and
suppose that it is the computer’s turn to move. The algorithm for computing scores will first try all
1 2 3
4 5 6
7 8 9
1 2 3
4 5 6
7 8 9
1 2 3
4 5 6
7 8 9
1 2 3
4 5 6
7 8 9
1 2 3
4 5 6
7 8 9
O
O
O
X
X
X
1 2 3
4 5 6
7 8 9
1 2 3
4 5 6
7 8 9
O
O
O
X
X
O X
1 2 3
4 5 6
7 8 9
1 2 3
4 5 6
7 8 9
O
O
O
X
X
X O
1 2 3
4 5 6
7 8 9
1 2 3
4 5 6
7 8 9
O
O
O
X
X
X
O
1 2 3
4 5 6
7 8 9
1 2 3
4 5 6
7 8 9
O X
O
O
O
X
X
X
1 2 3
4 5 6
7 8 9
1 2 3
4 5 6
7 8 9
1 2 3
4 5 6
7 8 9
X
O
O
O
O
X
X
X
1 2 3
4 5 6
7 8 9
1 2 3
4 5 6
7 8 9
1 2 3
4 5 6
7 8 9
1 2 3
4 5 6
7 8 9
O
O
O
X
X
X X O
1 2 3
4 5 6
7 8 9
1 2 3
4 5 6
7 8 9
1 2 3
4 5 6
7 8 9
O
O
O
X
X
X O
X
1 2 3
4 5 6
7 8 9
1 2 3
4 5 6
7 8 9
1 2 3
4 5 6
7 8 9
O
O
O
X
X
X
O
X
1 2 3
4 5 6
7 8 9
1 2 3
4 5 6
7 8 9
1 2 3
4 5 6
7 8 9
O
O
O
X
X
X
O
X
1 2 3
4 5 6
7 8 9
1 2 3
4 5 6
7 8 9
1 2 3
4 5 6
7 8 9
O
O
O
X
X
X X O
1 2 3
4 5 6
7 8 9
1 2 3
4 5 6
7 8 9
1 2 3
4 5 6
7 8 9
O
O
O
X
X
X O
X
1 2 3
4 5 6
7 8 9
1 2 3
4 5 6
7 8 9
1 2 3
4 5 6
7 8 9
O
O
O
X
X
X
O
X
1 2 3
4 5 6
7 8 9
1 2 3
4 5 6
7 8 9
1 2 3
4 5 6
7 8 9
O
O
O
X
X
X
O
X
11
O X
O
O
O
X
X
X
1 2 3
4 5 6
7 8 9
1 2 3
4 5 6
7 8 9
1 2 3
4 5 6
7 8 9
1 2 3
4 5 6
7 8 9
X
O
O
O
O
X
X
X
1 2 3
4 5 6
7 8 9
1 2 3
4 5 6
7 8 9
1 2 3
4 5 6
7 8 9
1 2 3
4 5 6
7 8 9
O
O
O
O O O
Level 0
Computer
Level 1
Human
Level 2
Computer
Level 3
Human
3 2 0 2 3 0
3 2 0 3 0
2
2
A B C
2 0 0
A1 A2 B1 B2 1 2
A A22 B11 B22 C11 C22
C C
Figure 2: A game tree.
possible moves that the computer can make: A (play at position 7), B (play at position 9), and C
(play at position 3). For each one of these moves, the algorithm will then consider all possible moves
by the human player: A1 and A2 (if the computer plays as in configuration A), B1 and B2 (if the
computer plays as in configuration B), and C1 and C2 (if the computer plays as in configuration C).
Then, all possible responses by the computer are attempted, and so on until the outcome of each
possible sequence of plays is determined.
In Figure 2 each level of the tree is labelled by the player whose turn is next. So levels 0 and 2 are
labelled “computer” and the other 2 levels are labelled “human”. After reaching final configurations
A11, A22, B11, B22, C11, and C22, the algorithm computes a score for each one of them depending
on whether the computer wins, the human wins, or the game is a draw. These scores are propagated
upwards as follows:
• For a configuration b on a level labelled “computer”, the highest score of the adjacent configurations in the next level is selected as the score for b. This is because the higher the score is,
the better the outcome is for the computer.
• For a configuration b on a level labelled “human”, the score of b is equal to the minimum score
of the adjacent configurations in the next level, because the lower the score is, the better the
outcome is for the human player.
The scores for the configurations in Figure 2 are the numbers in boldface. For example, for the
configuration at the top of Figure 2, putting an ’O’ in position 7 yields the configuration with the
highest score (2), hence the computer will choose to play in position 7. Similarly, for configuration
A in Figure 2, placing an ’X’ in position 3 yields the configuration with the smallest score (2), so the
human will choose to play in position 3.
We give below the algorithm for computing scores and for selecting the best available move. The
algorithm is given in Java, but we have omitted variable declarations and some initialization steps. A
full version of the algorithm can be found inside class Play nk TTT.java, which can be downloaded
from the course’s website.
2
private PosPlay computerPlay(char symbol, int highestScore, int lowestScore, int level) {
if (level == 0) configurations = t.createDictionary();
if (symbol == ’X’) opponent = ’O’; else opponent = ’X’;
for(int row = 0; row < board size; row++)
for(int column = 0; column < board size; column++) if(t.squareIsEmpty(row,column)) { // Empty position found t.storePlay(row,column,symbol); if (t.wins(symbol)||t.isDraw()||(level == max level)) reply = new PosPlay(t.evalBoard(),row,column); else { (*) lookupVal = t.repeatedConfig(configurations); (*) if (lookupVal != -1) reply = new PosPlay(lookupVal,row,column); } else { reply = computerPlay(opponent, highestScore, lowestScore, level + 1); t.insertConfig(configurations,reply.getScore()); } } t.storePlay(row,column,’ ’); if((symbol == COMPUTER && reply.getVal() > value) || // A better play was found
(symbol == HUMAN && reply.getVal() < value)) { bestRow = row; bestColumn = column; value = reply.getVal(); if (symbol == COMPUTER && value > highestScore) highestScore = value;
else if (symbol == HUMAN && value < lowestScore) lowestScore = value; if (highestScore >= lowestScore) /* alpha/beta cut */
return new PosPlay(value, bestRow, bestColumn);
}
}
return new PosPlay(value, bestRow, bestColumn);
}
The fist parameter of the algorithm is the symbol (either ’X’ or ’O’) of the player whose turn is
next. The second and third parameters are the highest and lowest scores for the board positions that
have been examined so far. The last parameter is used to bound the maximum number of levels of
the game tree that the algorithm will consider. Since the number of configurations in the game tree
could be very large, to speed up the algorithm the value of the last parameter specifies the highest
level of the game tree that will be explored. Note that the smaller the value of this parameter is, the
faster the algorithm will be, but the worse it will play.
Also note that if we bound the number of levels of the game tree, it might not be possible to
determine the outcome of the game for some of the configurations in the lowest level of the tree.
For example, if in the game tree of Figure 2 we set the maximum level to 2, then the algorithm will
explore only levels 0, 1, and 2. At the bottom of the tree will appear configurations A1, A2, B1,
B2, C1, and C2. Among these configurations, the scores for B1 and C2 are 0, as the human player
wins in those cases; however, the scores for the remaining configurations are not known as in none
of these configurations any player has won, and the configurations still include empty positions, so
they do not denote game draws. In this case, configurations A1, A2, B2, and C1 will receive a score
of UNDECIDED = 1.
3
2.1 Speeding-up the Algorithm with a Dictionary
The above algorithm includes several tests that allow it to reduce the number of configurations that
need to be examined in the game tree. For this assignment, the most important test used to speed-up
the program is the one marked (*). Every time that the score of a board configuration is computed,
the configuration and its score are stored in a dictionary, that you will implement using a hash table.
Then, when algorithm computerPlay is exploring the game tree trying to determine the computer’s
best move, before it expands a configuration b it will look it up in the dictionary. If b is in the
dictionary, then its score is simply extracted from the dictionary, instead of exploring the part of the
game tree below b.
For example, consider the game tree in Figure 3. The algorithm examines first the left branch
of the game tree, including configuration D and all the configurations that appear below it. After
exploring the configurations below D, the algorithm computes the score for D and then it stores D
and its score in the dictionary. When later the algorithm explores the right branch of the game tree,
configuration D will be found again, but this time its score is simply obtained from the dictionary
instead of exploring all configurations below D, thus reducing the running time of the algorithm.
1 2 3
4 5 6
7 8 9
1 2 3
4 5 6
7 8 9
X
O
1 2 3
4 5 6
7 8 9
X X
O
1 2 3
4 5 6
7 8 9
X O
1 2 3
4 5 6
7 8 9
X
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
X X
O
O
1 2 3
4 5 6
7 8 9
. . .
…
. .. .. .
…
D
X X O
O
X X O
1 2 3
4 5 6
7 8 9
D
Figure 3: Detecting repeated configurations.
3 Classes to Implement
You are to implement at least 3 Java classes: Dictionary.java, nk TicTacToe.java, and
Record.java. You can implement more classes if you need to. You must write all the code yourself.
You cannot use code from the textbook, the Internet, or any other sources. You cannot use the standard Java classes Hashtable, HashMap, HashSet or any other Java provided class that implements a
hash table. You cannot use the hashCode() method.
3.1 Record
This class represents an entry in the dictionary, associating a configuration with its integer score.
Each board configuration will be represented as a string as follows: concatenate all characters placed
4
in the board starting at the top left position and moving from left to right and from top to bottom. For example, for the configurations in Figure 1, their string representations are “OXXXXOO ”
1
,
“OXXXXOOOX”, and “OXXXXOOXO”.
For this class, you must implement all and only the following public methods:
• public Record(String config, int score): A constructor which returns a new Record
with the specified configuration and score. The string config will be used as the key attribute
for every Record object.
• public String getConfig(): Returns the configuration stored in this Record.
• public int getScore(): Returns the score in this Record.
You can implement any other methods that you want to in this class, but they must be declared as
private methods (i.e. not accessible to other classes).
3.2 Dictionary
This class implements a dictionary. You must implement the dictionary using a hash table with
separate chaining. You will decide on the size of the table, keeping in mind that the size of the table
must be a prime number. A table of size between 6000-8000, should work well.
You must design your hash function so that it produces few collisions. (A bad hash function
that induces many collisions will result in a lower mark.) Collisions must be resolved using separate
chaining.
For this class, you must implement all the public methods in the following interface:
public interface DictionaryADT {
public int insert (Record pair) throws DictionaryException;
public void remove (String config) throws DictionaryException;
public int get (String config);
public int numElements();
}
The description of these methods follows:
• public int insert(Record pair) throws DictionaryException: Inserts the given Record
pair in the dictionary. This method must throw a DictionaryException (see below) if
pair.getConfig() is already in the dictionary.
You are required to implement the dictionary using a hash table with separate chaining. To
determine how good your design is, we will count the number of collisions produced by your
hash function.
Method insert must return the value 1 if the insertion of pair produces a collision, and it will
return the value 0 otherwise. In other words, if for example your hash function is h(key) and
the name of your table is T, this method will return the value 1 if T[h(pair.getConfig())]
already stores at least one element; it will return 0 if T[h(pair.getConfig())] was empty
before the insertion.
• public void remove(String config) throws DictionaryException: Removes the entry
with the given config from the dictionary. Must throw a DictionaryException (see below)
if the configuration is not in the dictionary.
1
note that there are two blank spaces at the end of the string representing the two empty places in the board of
Figure 1(a)
5
• public int get(String config): A method which returns the score stored in the dictionary
for the given configuration, or -1 if the configuration is not in the dictionary.
• public int numElements(): A method that returns the number of Record objects stored in
the dictionary.
Since your Dictionary class must implement all the methods of the DictionaryADT interface,
the declaration of your method should be as follows:
public class Dictionary implements DictionaryADT
You can download the file DictionaryADT.java from the course’s website. The only other public
method that you can implement in the Dictionary class is the constructor method, which must be
declared as follows
public Dictionary(int size)
this returns an empty dictionary of the specified size.
You can implement any other methods that you want to in this class, but they must be declared
as private methods (i.e. not accessible to other classes).
3.3 nk TicTacToe
This class implements all the methods needed by algorithm computerPlay, which are described
below. The constructor for this class must be declared as follows
public nk TicTacToe (int board size, int inline, int max levels)
The first parameter specifies the size of the board, the second is the number of symbols in-line needed
to win the game, and the third is the maximum level of the game tree that will be explored by the
program. So, for example, to play the usual (3, 3)-tic-tac-toe, the first two parameters will have value
3.
This class must have an instance variable gameBoard of type char[][] to store the gameboard.
This variable is initialized inside the above constructor method so that every entry of gameBoard
stores a space ’ ’. Every entry of gameBoard stores one of the characters ’X’, ’O’, or ’ ’. This class
must also implement the following public methods.
• public Dictionary createDictionary(): returns an empty Dictionary of the size that you
have selected.
• public int repeatedConfig(Dictionary configurations): This method first represents
the content of gameBoard as a string as described above; then it checks whether the string
representing the gameBoard is in the configurations dictionary: If it is in the dictionary this
method returns its associated score, otherwise it returns the value -1.
• public void insertConfig(Dictionary configurations, int score): This method first
represents the content of gameBoard as a string as described above; then it inserts this string
and score in the configurations dictionary.
• public void storePlay(int row, int col, char symbol): Stores symbol in
gameBoard[row][col].
• public boolean squareIsEmpty (int row, int col): Returns true if
gameBoard[row][col] is ’ ’; otherwise it returns false.
• public boolean wins (char symbol): Returns true if there are k adjacent occurrences of
symbol in the same row, column, or diagonal of gameBoard, where k is the number of required
symbols in-line needed to win the game.
6
• public boolean isDraw(): Returns true if gameBoard has no empty positions left and no
player has won the game.
• public int evalBoard(): Returns one of the following values:
– 3, if the computer has won, i.e. there are k adjacent ’O’s in the same row, column, or
diagonal of gameBoard;
– 0, if the human player has won.
– 2, if the game is a draw, i.e. there are no empty positions in gameBoard and no player
has won.
– 1, if the game is still undecided, i.e. there are still empty positions in gameBoard and no
player has won.
4 Classes Provided
You can download classes DictionaryADT.java, PosPlay.java, Play nk TTT.java and
DictionaryException.java from the course’s website. Class PosPlay is an auxiliary class used
by Play nk TTT to represent plays. Class Play nk TTT has the main method for your program, so
the program will be executed by typing
java Play nk TTT size in line max levels
where size is the size of the gameboard, in line is the number of symbols that need to be placed
in-line to win the game, and max levels is the maximum number of levels of the game tree that the
program will explore.
Class Play nk TTT also contains methods for displaying the gameboard on the screen and for
entering the moves of the human player.
5 Testing your Program
We will perform two kinds of tests on your program: (1) tests for your implementation of the
dictionary, and (2) tests for your implementation of the program to play (n, k)-tic-tac-toe. For testing
the dictionary we will run a test program called TestDict which verifies whether your dictionary
works as specified. We will supply you with a copy of TestDict so you can use it to test your
implementation.
6 Coding Style
Your mark will be based partly on your coding style.
• Variable and method names should be chosen to reflect their purpose in the program.
• Comments, indenting, and white spaces should be used to improve readability.
• No variable declarations should appear outside methods (“instance variables”) unless they
contain data which is to be maintained in the object from call to call. In other words, variables
which are needed only inside methods, whose value does not have to be remembered until the
next method call, should be declared inside those methods.
• All variables declared outside methods (“instance variables”) should be declared private (not
protected), to maximize information hiding. Any access to these variables should be done
with accessor methods (like getScore() for Record).
7
7 Marking
Your mark will be computed as follows.
• Program compiles, produces meaningful output: 2 marks.
• Dictionary tests pass: 4 marks.
• nk TicTacToe tests pass: 4 marks.
• Coding style: 2 marks.
• Record and Dictionary implementation: 4 marks.
• nk TicTacToe program implementation: 4 marks.
8 Handing In Your Program
You must submit an electronic copy of your program. To submit your program, login to
OWL and submit your java files from there. Please do not put your code in sub-directories. Do
not compress your files or submit a .zip, .rar, .gzip, or any other compressed file. Only your .java
files should be submitted. Remember that the TA’s will test your program on the computers of the
Department.
When you submit your program, we will receive a copy of it with a datestamp and timestamp.
If you re-submit your program after the due date, please send me an email message to ensure that
the TA’s mark your latest submission. We will take the last program submitted as the final version,
and will deduct marks accordingly if it is late.
It is your responsibility to ensure that your assignment was received by the system.
8
