MinimaxRuleGames II 2007

Game-playing AIs
Part 2
CIS 391
Fall 2007
Games: Outline of Unit
Part II
 The Minimax Rule
 Alpha-Beta Pruning
 Game-playing AI successes
CSE 391 - Intro to AI

2
`Don’t play hope chess’: The Minimax Rule
• Idea: make the move for player MAX which has the most benefit
assuming that MIN makes the best move for MIN in response
• This is computed by a recursive process

• The backed-up value of each node in the tree is determined by
the values of its children
• For a MAX node, the backed-up value is the maximum of the
values of its children (i.e. the best for MAX)
• For a MIN node, the backed-up value is the minimum of the
values of its children (i.e. the best for MIN)

3
The Minimax Procedure
1. Start with the current position as a MAX node.
2. Expand the game tree a fixed number of ply (half-moves).
3. Apply the evaluation function to the leaf positions.
4. Calculate back-up up values bottom-up.
5. Pick the move which was chosen to give the MAX value at the root.

4
2-ply Minimax Example
2
2 1
2 1
2 7 1 8 2 7 1 8
2 7 1 8
This is the move 2
Evaluation function selected by minimax
value
2 1
MAX
MIN 2 7 1 8
5
What if MIN does not play optimally?
 Definition of optimal play for MAX assumes MIN

plays optimally:
• maximizes worst-case outcome for MAX.
 But if MIN does not play optimally, MAX will do
even better. [Theorem-not hard to prove]

6
Minimax Algorithm
function MINIMAX-DECISION(state) returns an action
inputs: state, current state in game
vMAX-VALUE(state)
return the action in SUCCESSORS(state) with value v
function MAX-VALUE(state) returns a utility value
if TERMINAL-TEST(state) then return UTILITY(state)
v  -∞
for a,s in SUCCESSORS(state) do
v  MAX(v, MIN-VALUE(s) )
return v
function MIN-VALUE(state) returns a utility value
v∞
v  MIN(v, MAX-VALUE(s) )
return v
7
Comments on Minimax Search
 Depth-first search with fixed number of ply as the limit.
• O(bm) time complexity – Ooops!
• O(bm) space complexity
 Performance will depend on
• the quality of the evaluation function (expert knowledge)
• depth of search (computing power and search algorithm)
 Differences from normal state space search
• Looking for one move only
• No cost on arcs
• MAX can’t be sure how MIN will respond to his moves
 Minimax rule forms the basis for other game tree search
algorithms.

8
Alpha-Beta Pruning
Slides of example from screenshots by

Mikael Bodén, Halmstad University, Sweden
found at
http://www.emunix.emich.edu/~evett/AI/AlphaBeta_movie/sld001.htm
Alpha-Beta Pruning
 A way to improve the performance of the Minimax
Procedure
 Basic idea: “If you have an idea which is surely
bad, don’t take the time to see how truly awful it
is” ~ Pat Winston
>=2
• We don’t need to compute
=2 <=1 the value at this node.
• No matter what it is it can’t
effect the value of the root
node.
2 7 1 ?
10
Alpha-Beta Pruning
 Traverse the search tree in depth-first order
 For each MAX node n, α(n)=maximum child value found so
far
• Starts with –
• Increases if a child returns a value greater than the current α(n)
• Lower-bound on the final value
 For each MIN node n, β(n)=minimum child value found so
far
• Starts with +
• Decreases if a child returns a value less than the current β(n)
• Upper-bound on the final value
 MAX cutoff rule: At a MAX node n, cut off search if
α(n)>=β(n)
 MIN cutoff rule: At a MIN node n, cut off search if β(n)<=α(n)
 Carry α and β values down in search

11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Alpha-Beta Algorithm I
function ALPHA-BETA-SEARCH(state) returns an action

inputs: state, current state in game
vMAX-VALUE(state, - ∞ , +∞)
return the action in SUCCESSORS(state) with value v
function MAX-VALUE(state, , ) returns a utility value

v-∞
v  MAX(v,MIN-VALUE(s,  , ))
if v ≥  then return v
  MAX( ,v)
return v

30
Alpha-Beta Algorithm II
function MIN-VALUE(state,  , ) returns a utility value

v+∞
v  MIN(v,MAX-VALUE(s,  , ))
if v ≤  then return v
  MIN( ,v)
return v

31
Effectiveness of Alpha-Beta Pruning
 Guaranteed to compute same root value as
Minimax
 Worst case: no pruning, same as Minimax (O(bd))
 Best case: when each player’s best move is the
first option examined, you examine only O(bd/2)
nodes, allowing you to search twice as deep!
 For Deep Blue, alpha-beta pruning reduced the
average branching factor from 35-40 to 6.

32
Chinook and Deep Blue
 Chinook
• the World Man-Made Checkers Champion, developed at the University of
Alberta.
• Competed in human tournaments, earning the right to play for the human
world championship, and defeated the best players in the world.
• Play Chinook at http://www.cs.ualberta.ca/~chinook
 Deep Blue
• Defeated world champion Gary Kasparov 3.5-2.5 in 1997 after losing 4-2
in 1996.
• Uses a parallel array of 256 special chess-specific processors
• Evaluates 200 billion moves every 3 minutes; 12-ply search depth
• Expert knowledge from an international grandmaster.
• 8000 factor evaluation function tuned from hundreds of thousands of
grandmaster games
• Tends to play for tiny positional advantages.
33

MinimaxRuleGames II 2007

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

MinimaxRuleGames II 2007

Uploaded by

Copyright:

Available Formats

Game-playing AIs

CSE 391 - Intro to AI

• This is computed by a recursive process

CSE 391 - Intro to AI

2. Expand the game tree a fixed number of ply (half-moves).

3. Apply the evaluation function to the leaf positions.

4. Calculate back-up up values bottom-up.

CSE 391 - Intro to AI

 Definition of optimal play for MAX assumes MIN

CSE 391 - Intro to AI

CSE 391 - Intro to AI

Slides of example from screenshots by

CSE 391 - Intro to AI

function ALPHA-BETA-SEARCH(state) returns an action

function MAX-VALUE(state, , ) returns a utility value

CSE 391 - Intro to AI

function MIN-VALUE(state,  , ) returns a utility value

CSE 391 - Intro to AI

CSE 391 - Intro to AI

You might also like