WANT to learn how to influence people? Rather than take dubious advice from a self-help guru, you could try listening to the mathematicians who have cracked the problem.
David Kempe and his colleagues at Cornell University in New York have found a way to identify who you should talk to if you want your ideas spread to as many people as possible. And contrary to what common sense might predict, it is not necessarily the person with the most contacts.
Kemp鈥檚 team studied a community of 10,750 particle physicists. The researchers envisaged this community as a network, in which each person is represented as a node and two nodes are linked if the physicists have co-authored a research paper.
Advertisement
Kemp then devised an algorithm to identify the most influential physicists in the network. The mathematical program did this not by counting the number of links each physicist had, but by looking farther and analysing how information spreads through the network.
What the researchers found is that only certain people are able to influence large proportions of the network. While someone may have co-authored papers with 20 colleagues, their influence might not spread further than those 20 individuals. Another physicist, with direct links with just two colleagues, might be able to influence far more of the network if his two contacts are themselves linked to hundreds more scientists.
For example, one of the physicists the algorithm identified was Christopher Pope from Texas A&M University. When it went on to search for a physicist capable of influencing parts of the network to which Pope was not connected, Hong Lu from the University of Michigan at first seemed to be the obvious candidate as the best-connected person. But both Lu and Pope influenced the same areas of the network, so Kempe鈥檚 algorithm skipped on to someone else.
鈥淭he extent of a word-of-mouth cascade depends not just on who an initially targeted person influences, but who those people influence, and who those people influence, and so on,鈥� says Duncan Watts, a sociologist at Columbia University, New York, and author of the book Six Degrees: The science of a connected age. 鈥淚t鈥檚 an extremely impressive piece of work.鈥�
Identifying the few most influential people is a huge challenge, even for a computer algorithm. Kemp鈥檚 team had to run their algorithm thousands of times for it to identify the few people capable of triggering the cascade of influence that affected the largest proportion of the network.