Where to look and why? :
Utility and heuristics as explanations
of cooperation in one-shot PD games
Shigehito Tanida & Toshio Yamagishi
Hokkaido University
Overview
• Our Question:
Why do people cooperate in one-shot
PD games with an unrelated partner?
• We measured subject’s eye movements
via an eye camera while Ss were looking at
the matrix and making a decision in one-
shot PD game.
• We examine the validity of explanation of cooperation in one-shot PD,
analyzing relationship between the subject’s
attention to the matrix and their decision.
Two alternative explanations of
cooperation in one-shot PD game
1. Utility transformation explanation
– Social Value Orientation (SVO) (e.g., van Lange, 1999):
Utility of an outcome in a social exchange
(e.g., PD game) is a combination of
1. Utility of own outcome 2. Utility of partner’s outcome
3. Disutility of inequality of the two players’ outcomes
Cooperators (Pro-Socials) are expected to pay
more attention to the partner’s outcome than
defectors.
Defectors (Pro-Selfs) care only about their own
outcome.
Two alternative explanations of
cooperation in one-shot PD game
2. Social Exchange Heuristic explanation
• Whether one cooperates or defects depends on
how he/she define the game the two
The relative plausibility of situation.
In this approach, players’ decisions depend not for
• explanations will depend on how easy it is so
much on the utilities of outcomes but rather as
the player to intuitively grasp the game on an
activation of a particular heuristic
incidence of social exchange.
SEH:A default decision rule in a social exchange
that minimizes a condition that was
We conducted two particular type of error
When activated by cues of being in sense of
manipulated the kind of intuitive a social
exchange, it makes people seek mutual
exchange. while avoiding being a sucker.
cooperation
Simple Matrix Condition
• The gain of mutual cooperation is easy to understand
Partner
Cooperate Defect
Own Partner Own Partner
Cooperate
Subject
1200 1200 0 1800
Defect Own Partner Own Partner
1800 0 600 600
Complex Matrix Condition
• It is difficult to intuitively grasp that it represents a
social exchange
(simple payoff x .93) + 165 Partner
Cooperate Defect
Own Partner Own Partner
Cooperate
Subject
1281 1281 165 1839
Defect Own Partner Own Partner
1839 165 723 723
Eye Movement Tracking
• We observed where in the PD matrix players
looked at with an eye camera.
If the utility transformation explanation is valid,
In both condition, Cooperators should look at
their partner’s payoffs (compared to own
payoffs) more than Defectors.
If the Heuristic explanation is valid,
The cooperation rate and the proportion of
fixation of mutual cooperation cell that Ss
looked at in Simple Matrix condition should be
higher than in the Complex Matrix Condition.
Result1: Cooperation rate & Proportion of the fixation
• Cooperation rate
Simple Matrix Condition: 44.8%
Complex Matrix Condition: 36.7% (χ2(1)=0.30, n.s.）
• Proportion of the fixation
Cooperators Defectors
C D C D
C .38 .25 C .23 .21
D
.21 .16 D .32 .25
Which did Ss look at matrix? - Own or
Partner's payoff?
Complex Matrix Condition
Result : Simple Matrix Condition
1
1 Own
Own PayoffCooperate
p<.05 Partner
Partner's
Defect
Partner's Payoff
Own Partner Own Partner
Cooperate
1200 1200 0 1800
Subject0.5
0.5
Defect Own Partner Own Partner
1800 0 600 600
0
0
Cooperator
Cooperator Defector
Defector
• Proportion of fixation in Simple Matrix condition
C D C D
C .39 .27 C .27 .22
D D
.21 .13 .28 .24
Cooperators Defectors
• Proportion of fixation in Complex Matrix condition
C D C D
C .38 .21 C .17 .20
D
.21 .21
D .37 .26
Cooperators Defectors
Discussion
• The result of Defectors’ attention in
Complex Matrix Condition indicated that
players’ social motivation plays an important
role in the situation that is lacking the kind of
intuitive sense of exchange.
• Defectors in Simple Matrix Condition paid
attention to the mutual cooperation cell,
suggested that their decision is more likely
to be driven by heuristics rather than
transformed utilities.