Friam

Advantages of groups

Classic

List

Threaded

5 messages Options

Russ Abbott

Advantages of groups

A recent paper in PNAS shows that groups make faster and more accurate decisions the larger they are (presumably up to a limit). The primary reason is that larger groups have "more eyes" than smaller groups. As long as information transmission through groups is not significantly delayed, larger is better.

Abstract. Although it has been suggested that large animal groups should make better decisions than smaller groups, there are few empirical demonstrations of this phenomenon and still fewer explanations of the how these improvements may be made. Here we show that both speed and accuracy of decision making increase with group size in fish shoals under predation threat. We examined two plausible mechanisms for this improvement: first, that groups are guided by a small proportion of high-quality decision makers and, second, that group members use self-organized division of vigilance. Repeated testing of individuals showed no evidence of different decision-making abilities between individual fish. Instead, we suggest that shoals achieve greater decision-making efficiencies through division of labor combined with social information transfer. Our results should prompt reconsideration of how we view cooperation in animal groups with fluid membership.

... The “many eyes” theory predicts that predators will be detected sooner by larger groups, increasing the probability of escape for the individual that first detects them, as well as for other group members as the information is transmitted across the group (9–13). Despite the fact that numerous studies have provided evidence in support of this functional benefit of larger group sizes, little is known about the decision-making mechanisms that underlie this key benefit of sociality. If we assume that members independently gather information and scan different parts of the environment, then the first individual to detect a given cue could initiate a response that proliferates rapidly throughout the group (14–17). Under this scenario, improvements in both speed and accuracy of collective decision making with group size can result because a) the probability that the group contains individuals that are better at scanning increases with group size and/or b) individuals do not need to scan as much of their environment when in groups, effectively focusing their information gathering to a smaller area of space.

... Both speed and accuracy of collective decision making increase with group size. In comparison with single fish, groups of 8 and of 16 fish were significantly faster and more accurate. The improvements in accuracy were remarkably similar to a perfect many eyes prediction, where detection of the predator by a single individual equates to detection by the entire group.

... Clustering of decisions is indicative of positive feedback through social cues. This feedback may well involve quorum rules, where the probability of moving in a particular direction increases nonlinearly with group size. These quorums allow errors by individual fish to be filtered out. However, what is most remarkable in the current experiment is that improvements in accuracy start even in groups of two and then accelerate rapidly.

... Our results are not consistent with the game-theoretic reasoning often applied when thinking about predator avoidance situations. For example, there is no evidence of individuals in larger groups slowing down to force another group member to take the risk of inspecting the two options. On the contrary, the efficiency of the fish shoal's many eyes suggests a high degree of cooperation in detection and avoidance of predators. The evolution of this cooperation can be understood only by identifying the cues used by a fish that allow it to determine when another fish has spotted a predator.

I have always thought that "many eyes" is also the reason that market-based system work better than centrally controlled systems. A market-based system has many eyes all looking for efficiencies to bring to market. A centrally controlled system simply can't match that.

-- Russ Abbott
_____________________________________________

Professor, Computer Science
California State University, Los Angeles

Google voice: 747-999-5105
blog: http://russabbott.blogspot.com/
vita: http://sites.google.com/site/russabbott/
_____________________________________________

============================================================
FRIAM Applied Complexity Group listserv
Meets Fridays 9a-11:30 at cafe at St. John's College
lectures, archives, unsubscribe, maps at http://www.friam.org

Nick Thompson

Re: Advantages of groups

Interesting, Russ. Thanks for bringing this to our attention.

The legislature is in town these days¸ and given the appalling inefficiency of its decision making processes, it would seem that the fish are better at collective decision making than we are.

Nick

From: [hidden email] [mailto:[hidden email]] On Behalf Of Russ Abbott
Sent: Saturday, March 05, 2011 11:15 PM
To: The Friday Morning Applied Complexity Coffee Group
Subject: [FRIAM] Advantages of groups

Abstract. Although it has been suggested that large animal groups should make better decisions than smaller groups, there are few empirical demonstrations of this phenomenon and still fewer explanations of the how these improvements may be made. Here we show that both speed and accuracy of decision making increase with group size in fish shoals under predation threat. We examined two plausible mechanisms for this improvement: first, that groups are guided by a small proportion of high-quality decision makers and, second, that group members use self-organized division of vigilance. Repeated testing of individuals showed no evidence of different decision-making abilities between individual fish. Instead, we suggest that shoals achieve greater decision-making efficiencies through division of labor combined with social information transfer. Our results should prompt reconsideration of how we view cooperation in animal groups with fluid membership.

... The “many eyes” theory predicts that predators will be detected sooner by larger groups, increasing the probability of escape for the individual that first detects them, as well as for other group members as the information is transmitted across the group (9–13). Despite the fact that numerous studies have provided evidence in support of this functional benefit of larger group sizes, little is known about the decision-making mechanisms that underlie this key benefit of sociality. If we assume that members independently gather information and scan different parts of the environment, then the first individual to detect a given cue could initiate a response that proliferates rapidly throughout the group (14–17). Under this scenario, improvements in both speed and accuracy of collective decision making with group size can result because a) the probability that the group contains individuals that are better at scanning increases with group size and/or b) individuals do not need to scan as much of their environment when in groups, effectively focusing their information gathering to a smaller area of space.

... Both speed and accuracy of collective decision making increase with group size. In comparison with single fish, groups of 8 and of 16 fish were significantly faster and more accurate. The improvements in accuracy were remarkably similar to a perfect many eyes prediction, where detection of the predator by a single individual equates to detection by the entire group.

... Clustering of decisions is indicative of positive feedback through social cues. This feedback may well involve quorum rules, where the probability of moving in a particular direction increases nonlinearly with group size. These quorums allow errors by individual fish to be filtered out. However, what is most remarkable in the current experiment is that improvements in accuracy start even in groups of two and then accelerate rapidly.

... Our results are not consistent with the game-theoretic reasoning often applied when thinking about predator avoidance situations. For example, there is no evidence of individuals in larger groups slowing down to force another group member to take the risk of inspecting the two options. On the contrary, the efficiency of the fish shoal's many eyes suggests a high degree of cooperation in detection and avoidance of predators. The evolution of this cooperation can be understood only by identifying the cues used by a fish that allow it to determine when another fish has spotted a predator.

-- Russ Abbott
_____________________________________________

Professor, Computer Science
California State University, Los Angeles

Google voice: 747-999-5105
blog: http://russabbott.blogspot.com/
vita: http://sites.google.com/site/russabbott/
_____________________________________________

Jochen Fromm-5

Re: Advantages of groups

In reply to this post by Russ Abbott

James Surowiecki named for requirements for
"wise groups" in his book "The Wisdom of Crowds":
diversity of opinion, independence, decentralization
and aggregation. Large fish shoals should fulfill
them: "many eyes" lead to a diversity of opinion
about potential threats, the opinion of the "selfish"
fishes about a threat is not determined by the opinions
of those around them, and they are decentralized
by default. Somehow a fishy discussion thread..

There is also a new book by Patrick R. Laughlin
named "Group Problem Solving" which goes in
the same direction.
http://press.princeton.edu/titles/9339.html

-J.

============================================================
FRIAM Applied Complexity Group listserv
Meets Fridays 9a-11:30 at cafe at St. John's College
lectures, archives, unsubscribe, maps at http://www.friam.org

Jochen Fromm-5

Re: Advantages of groups

Sorry I meant "four" requirements of course. In principle,
the article says nothing new, it says that a large group
has less cognitive limitations than an individual.
Iain Couzin his written this before. But swarms are
always fascinating! I updated the wiki pages for
the corresponding topics:

http://wiki.cas-group.net/index.php?title=Collective_Intelligence
http://wiki.cas-group.net/index.php?title=Crowd
http://wiki.cas-group.net/index.php?title=Swarm
http://wiki.cas-group.net/index.php?title=Swarm_Intelligence

-J.

============================================================
FRIAM Applied Complexity Group listserv
Meets Fridays 9a-11:30 at cafe at St. John's College
lectures, archives, unsubscribe, maps at http://www.friam.org

Nick Thompson

Re: Advantages of groups

Jochen,

Thanks for this. Your message led me to a whole cascade of wikis on
subjects I am supposed to know something about.

Note the use of modal language.

Nick

-----Original Message-----
From: [hidden email] [mailto:[hidden email]] On Behalf
Of Jochen Fromm
Sent: Sunday, March 06, 2011 7:40 AM
To: The Friday Morning Applied Complexity Coffee Group
Subject: Re: [FRIAM] Advantages of groups

Sorry I meant "four" requirements of course. In principle, the article says
nothing new, it says that a large group has less cognitive limitations than
an individual.
Iain Couzin his written this before. But swarms are always fascinating! I
updated the wiki pages for the corresponding topics:

http://wiki.cas-group.net/index.php?title=Collective_Intelligence
http://wiki.cas-group.net/index.php?title=Crowd
http://wiki.cas-group.net/index.php?title=Swarm
http://wiki.cas-group.net/index.php?title=Swarm_Intelligence

-J.

============================================================
FRIAM Applied Complexity Group listserv
Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives,
unsubscribe, maps at http://www.friam.org

============================================================
FRIAM Applied Complexity Group listserv
Meets Fridays 9a-11:30 at cafe at St. John's College
lectures, archives, unsubscribe, maps at http://www.friam.org