Coordinate on what stands out

When everyone gains by choosing alike, pick what stands out to everyone, not your favourite. To get a group to aim high together, show first that you will.

Thomas Schelling, in The Strategy of Conflict (1960), asked people to solve problems where they won only by matching a partner they could not talk to. Told to name heads or tails, 36 of his 42 respondents said heads. Asked where and when to meet someone in New York with no plan agreed, most of his students chose the information booth at Grand Central Station, at noon. Nothing made those answers better in themselves; they simply stood out, and each person expected the other to expect it. Schelling called such an answer a focal point.

Judith Mehta, Chris Starmer and Robert Sugden (1994) tested this. People told only to pick an answer chose differently from people told to try to match a partner: those trying to match crowded onto the obvious answer more, and coordinated better (in a later summary, 87% of those trying to match said heads). The lesson is to ask not “which do I like?” but “which will we all see?”.

Jean-Jacques Rousseau, in his Discourse on Inequality (1755), told of hunters who could catch a stag only if every one of them held his post, while any one of them could catch a hare alone. Brian Skyrms, in The Stag Hunt and the Evolution of Social Structure (2004), took it as a model of social cooperation: going for the stag is best if everyone does, but going for the hare is safe whatever the others do. Both are stable; which one a group lands on depends on trust.

In the lab, trust often loses. John Van Huyck, Raymond Battalio and Richard Beil (1990) had groups of people each choose an effort level from 1 to 7. Pay rose with the lowest effort anyone in the group chose, and fell with your own, so effort above the lowest was wasted. Large groups slid quickly to the lowest, worst outcome; smaller groups did better. One cautious person sets the result for all. Talk can help: Russell Cooper and colleagues (1992) found that in games where one choice was safer but the other paid more, letting both players send each other a message before playing always led them to the better outcome.

Words are cheap, though, and someone who was let down before may need more than a promise: a commitment they can see, or a track record of the plan working. These are lab and classroom findings, a guide to how groups behave rather than a rule for every team.

Why it matters

In a coordination problem nobody gains by being different, so the hard part is guessing what the others will choose. An answer that stands out to everyone lets people meet without talking; a bigger shared prize that needs everyone is reached only when each trusts the rest to go for it.

Away from the game

Meeting places, deadlines, standards and joint projects are coordination problems. Name the obvious meeting point; when a plan needs everyone, make your own commitment visible early, and keep your word so the next plan is easier.