INTELLIGENCE, THINKING, AND PERSONALITY

1
INTELLIGENCE, THINKING, AND PERSONALITY

• GAME PLAYING AND EXPERTISE

2
RESEARCH ON GAME PLAYING

• Mainly on how experts play games such as chess,
  draughts (checkers), poker, go
• There are millions of possible games (or more)
• So, very large state-action trees that cannot be
  held in memory
• Well defined (by rules of the game) as are puzzle
  book problems
• Emphasis on expertise and use of long-term
  memory, unlike work on problem solving

3
GAME PLAYING AS PROBLEM SOLVING

• Overall problem is usually taken to be how to win
  current game
• Though could think, for example, of how to
  improve ones game
• (Samuels checker player)
• Local problem is what move to make at current
  point in the game

4
STATE-ACTION ANALYSIS OF CHESS

• Straightforward, but must remember that players
  take turns, and that what is good for one is bad
  for the other (chess, for example, is a zero-sum
  game).
• If using such a representation to select a move,
  a player can rarely look ahead to a winning
  position. So, as in hill climbing, he or she
  must try to force what appears to be the most
  favourable local development of the game.
• And, as in hill climbing, the player must be able
  to evaluate reachable positions (using a
  so-called static evaluation function).

5
STATE-ACTION ANALYSIS OF CHESS - STATIC EVALUATION

• Factors relevant to evaluation of position at end
  of lookahead include
• number and type of the pieces that each player
  has
• which pieces have relative freedom of movement
• which pieces are vulnerable
• which parts of the board are controlled
• where the pawns are located.
• NOTE USE OF CHESS-SPECIFIC KNOWLEDGE - strong
  heuristic method

6
STATE-ACTION REPRESENTATION OF TWO-PERSON GAMES -
MINIMAXING

• Minimaxing means minimising the maximum loss that
  the other player can inflict on you
• Based on the assumption that the other player
  will, in making things good for him- or herself,
  make things as bad as possible for you

7
STATE-ACTION REPRESENTATION OF TWO-PERSON GAMES -
MINIMAXING
S1
Current position Program to move
Backed-up value Program can select best value
-2
Possible positions after next move Opponent to
move
S2
S3
Backed-up values Opponent will select Least
favourable value
-2
-5
Possible positions after opponents Next move
S4
S5
S6
S7
Static evaluations (for program) at end of
lookahead
-2
2
-5
7
8
HUMAN AND MACHINE CHESS

• In machine chess (CRAY BLITZ DEEP THOUGHT) the
  evaluated positions may be selected by brute
  force (because program can evaluate a very large
  number of positions).
• But, in human chess very many fewer positions can
  be considered, and they are selected using
  domain specific knowledge. i.e. EXPERTISE
  (strong methods again)

9
THE NATURE OF (HUMAN) EXPERTISE IN CHESS

• Chess masters and grandmasters typically think
  through a relatively small number of developments
  of the game.
• They assume a rational opponent, and try to force
  play to the development that seems most
  favourable.
• Choices of move are assessed not in terms of the
  position they create immediately, but in terms of
  the position that they will eventually lead to
  (at a so-called quiet position).
• Most moves are not followed up. They are assumed
  to be less valued that those that are.

10
DE GROOT - THOUGHT AND CHOICE IN CHESS

• (Dutch thesis - 1946 English book - 1965)
• Main Findings
• Human chess players consider only a few
  developments of the game at each move (Not
  surprising,given our knowledge about limitations
  on short-term memory).
• Excellent players (at the grandmaster level) do
  not follow up any more moves than good tournament
  players.
• They follow up better moves (as rated by other
  players), and they assess moves more quickly.
• Better players reconstruct (from memory) briefly
  presented board positions more accurately than
  less good players.

11
CHASE AND SIMON (1973)

• Chase and Simon (1973) showed that this finding
  only held for real chess positions (not random
  board positions with the same number of pieces).
• Chase and Simon suggested that a grandmaster
  might have 50,000 chunks of chess-related
  information in long-term memory.
• Explains why chess masters study previous games
  and why it takes 10 years to become a grandmaster
  (THE 10-YEAR RULE FOR THE ACQUISITION OF
  EXPERTISE IN A COMPLEX SKILL - Ericsson)

12
OTHER WORK ON EXPERTS vs. NOVICES

• Mainly on tasks such as
• Solving physics (applied maths) problems
• Computer programming
• Looks at differences in
• Knowledge
• Problem-Solving Methods
• planning

13
DIFFERENCES BETWEEN EXPERTS AND NOVICES

• (A) Knowledge
• Larkin (1983) novices representations of
  physics problems are naive. They use everyday
  concepts instead of the specialised concepts of
  physics that experts use. Naive representations
  often fail to suggest a solution.
• Chi, Feltovich, and Glaser (1981) showed that,
  when asked to sort problems into groups, novices
  relied on such superficial features as whether
  the problem was about an inclined plane, whereas
  experts classified the problems according to the
  physical principles that were relevant to their
  solution.

14
DIFFERENCES BETWEEN EXPERTS AND NOVICES

• (B) Problem Solving Methods
• Larkin, McDermott, Simon, and Simon (1980)
  suggested that experts try to work forwards, from
  the information given to the answer, whereas
  novices try to work backwards from the answer.
• Priest and Lindsay (1992), in a more extensive
  study, cast doubt on this idea, showing that both
  experts and novices prefer working forwards. The
  main difference was in the ability to produce a
  high-level plan before attempting a solution.
• Soloway and Ehrlich (1984) also suggest
  differences in levels of planning between expert
  and novice computer programmers.

15
EXPERT SYSTEMS

• Computer programs that attempt to embody human
  expertise.
• Often domain specific production (if....then)
  rules, with a simple inference engine.
• Same type of rules used to capture problem
  solving heuristics in the Newell and Simon
  approach

16
EXPERT SYSTEMS WELL-KNOWN EXAMPLES

• DENDRAL - works out structure of certain kinds of
  complex organic molecules (using mass
  spectrograph data)
• MYCIN - diagnoses potentially fatal bacterial
  infections and suggests treatments
• R2 - configures VAX computer systems for DEC
• PROSPECTOR - processes data from bore hole
  drilling to predict presence of minerals and/or
  oil.

17
EXPERTS BORN OR MADE?

• The ten-year rule suggests that experts are
  largely made.
• Autobiographic accounts of child prodigies are
  often misleading (e.g. H.G. Wells, Bernard Shaw
  exaggerating their childhood poverty and lack of
  resources).

18
EXPERTS BORN OR MADE? (cont.)

• Howe - stresses importance of parental pressure
  and child's temperament (devotion to duty will
  he or she put up with an hour's piano practice a
  day?).
• Sosniak (1990) - highly successful concert
  pianists were not recognised as outstanding until
  late in their childhood, after they started
  devoting unusually large amounts of time to
  practising.

19
EXPERTS BORN OR MADE? (cont.)

• The specificity of practice explains why the
  skill is isolated (practising piano does not
  help in mastering the mechanics of flute playing,
  for example, even if the ability to read music
  could, in principle, transfer).
• Practice is also crucial in the development of
  the skills of savants (e.g. the artist Stephen
  Wiltshire). However, it remains unclear why
  savants have a preserved island of talent.

20
EXPERTS BORN OR MADE? (cont.)

• Howe and others do not address the question of
  whether people have particular aptitudes (e.g.
  for music, or chess, or tennis).
• Such aptitudes could determine
• which domain a future expert decides to take
  devote themselves to
• whether someone who chooses to practice, e.g.
  piano, will succeed in becoming very talented.