Scientists Find the Cheapest Way to Connect Everyone in a Network

This is a math paper, not a product announcement. Two researchers, Juan M.C. Larrosa and Fernando Tohmé, study "game theory" — the math of how people make decisions when everyone else is deciding too. Their question: if each person pays only when they start a connection, but once linked, information flows both directions, who should connect to whom? One person's payment ends up helping the other person for free, which changes everything about the ideal setup.

Their answer is surprisingly tidy. The best structure is a straight line of people where every other person is a relay point — like a relay race, or a phone tree where each person calls two others and stops. Each relay point connects backward to the one before and forward to the one after, then hands off. That breaks the chain of accumulating costs while still getting messages where they need to go. Other arrangements still work, but this one is the most stable: nobody has a reason to quietly rearrange things for their own benefit.

So what does that mean outside the chalkboard? Think of mesh Wi-Fi at home, where devices pass signal to each other instead of all hitting one router. Or delivery routes, or a team coordinating a project through a few well-placed coordinators instead of everyone emailing everyone. The pattern says: don't connect everything to everything, and don't chain everyone in a long line where costs snowball. Space out the hubs.

The honest catch: this is pure theory. There's no app, no company, no launch date — just a proof. It also assumes people behave rationally and selfishly, and that the group is small enough for everyone to see what everyone else is doing. Real networks — office politics, family group chats, global supply chains — are far messier. But the underlying lesson is useful: cheap networks come from smart relay points, not more connections.