[Originally published in Hebrew in August 2016 here.]
Even during the age of the Roman Empire, groups of soldiers used to break their stride while crossing a bridge, as even a strong bridge can collapse beneath a military march. The fact is, herd behavior can cause an immediate disaster or a long-term ecological catastrophe. One of the natural ways that lead to heterogeneous behavior is the use of partial, incomplete, or ambiguous information. This article, by Professor Ido Erev of the Technion and his colleagues, explained and elegantly demonstrated – 25 years ago – how imprecise communication can save a population from loss and destruction, while precise and accurate information can severely harm the population’s survivability likelihood.
We tend to believe that when we cater information to someone, everything should be precise and crystal clear. In some cases, even when we think that the best instructions and guidance were provided, good intelligent people make mistakes. The more so, in the case of imprecise or vague communication, some individuals will understand things differently – in the best-case scenario, part of the population will perform well, and some will err; in the worst-case scenario, everyone shall be confused. So why go in the direction of ambiguity and vagueness anyway? When is it right to encourage heterogeneous behavior? Let us explain this with two examples.
- “The problem of the commons” – Every member of a community of shepherds owns 10 sheep grazing in a common pasture. A member that adds an 11th sheep to his own herd increases personal wealth, while the whole community shares the loss resulting from the reduced food supply per animal. If other members do the same, soon enough, the entire community will suffer from the shortage of grazing.
In this case, a clear and unambiguous rule is required – one that the community can also monitor and regulate. Here, there is no room for ambiguity; a homogeneous behavior works in favor of the common good.
- Now, let’s imagine a different situation that the same community of shepherds may face – a drought that forces it to move to a new location with water. Only a few routes are available, and it is known that there are wolves, that can kill all the sheep passing their way, around. Scouts were sent to observe the wolves in order to predict their location during the group’s migration. The scouts can only form an uncertain prediction of which route will be open – the information they provide about the safest route is of a probabilistic nature. If precise information is presented, all the shepherds will select the same route. In that case, there is still a chance that the wolves will lurk there and wipe out the whole sheep population. Clearly, such consistency is not in the group’s best interests.
But since the information, which the scouts provide, is of a probabilistic nature, different individuals will grasp it differently and select different migration routes; hence, eliminating the chances of wiping out the whole sheep population. Such a heterogeneous behavior will create the variability, which can prevent a disaster, and is so necessary to the survival of the community.
The researchers conducted an experiment, which demonstrated that people tend to use vague communication when the common good requires heterogeneous behavior. Thus, participants in the experiment managed to generate and communicate ambiguous information in order to produce the variability that was in the group’s best interests.
What is true to communities in danger of extinction and to bridges can also be true to our daily lives. Stock exchanges and markets tend to collapse when traders behave homogeneously. When, based on different interpretations of the same information, a heterogeneous behavior is created, goods have both buyers and sellers, and a healthy equilibrium is created. Even in the drafting of contracts and agreements, there is sometimes an advantage to vague wording so that each side understands the agreement a bit differently and can explain it to its colleagues and senders a little differently – which is much harder to do when leaving no room for vagueness.
How does this affect the design of user interfaces? Whenever a herd behavior may be a risk to the people themselves or to the stability of the system that serves them, it is best to try to use the media a little more vaguely and thereby neutralize some of the risks.
Some of the examples are:
- The websites of ticket sellers tend to crash when they start selling a popular show or concert. The less accurate distribution of the opening time of ticket sale for such events can neutralize the herd behavior, spread the demand over a longer period of time, and effectively balance the peak demand (which cause servers to crash). Instead of writing “Sunday at two,” write “Sunday in the afternoon”…
- Not too long ago, I designed a system that encourages households to save on various savings channels with government support and participation. If everyone chooses the same savings channel, all the funds will be in the hands of one entity, and in the event of a collapse, embezzlement, or improper investment, the risk to the entire savers will be huge. If, based on the same information, different people will decide on various savings channels, the risk will spread and in any case will be smaller for the entire crowd of savers.
- In the planning of systems for responding and population management in times of emergency, there is a clear interest in preventing the flooding of communication networks and escape routes. Consider, for example, estimates of tsunamis in the coastal plain – if we communicate to the general population information about escape routes, clearly and unequivocally, they will immediately be blocked; While if a more general guidance is given (for example, to move eastward), the evacuees will use most roads, and the load will dissipate, increasing the chances of survival.
Shay Ben-Barak helps companies and startups to identify their users’ needs and to cater for these needs relevant product-features with the best User eXperience! He is running a small design boutique that is focusing on complex systems and on meaningful artifacts. Shay owns a master’s degree in Cognitive Engineering from the Technion – Israel Institute of Technology and he is deeply involved in product management and UX design of digital products for almost two decades now. He also teaches UX design in the academy and evangelizes for ethical design practices.