Let’s play a (hypothetical) game. You must meet a stranger in New York City tomorrow. You can’t call, text, or signal them in any way. You pick a place and a time, show up, and hope they picked the same. No instructions of any kind may be used. Just you, more than eight million people, and a city that agrees on almost nothing, except maybe the Knicks after their championship run.

Most people land on the same answer: Grand Central Terminal, at noon.

Thomas Schelling, who would later win the Nobel in economics, posed this puzzle in The Strategy of Conflict (1960). His point wasn’t that Grand Central is special. It’s that coordination doesn’t require communication. Rather, it requires a focal point: an option so obvious that everyone expects everyone else to expect it. You don’t pick Grand Central because it’s your favorite. You pick it because you think they’ll pick it because they think you’ll pick it. The whole thing runs on a hall of mirrors.

The Part Everyone Skips

That hall of mirrors has a name: common knowledge. And it’s sneakier than it sounds.

Common knowledge isn’t everyone knowing a thing. It’s everyone knowing that everyone knows – all the way down. A fact can sit in every head in the building and still fail to be common knowledge, because each head privately suspects it’s the only one holding it.

Shared information is what’s in the room. Common knowledge is what the room knows it shares.

The Deadline Nobody Mentioned

Watch it fail in the wild. The deadline slipped. Maria knows. Devon knows. Priya knows. All three found out separately, and all three assumed they were the only one who noticed – so all three said nothing. The information was everywhere. The coordination was nowhere.

Three people, fully informed, doing nothing in perfect synchrony.

This isn’t a knowledge problem; every relevant fact was present and accounted for. It’s a common knowledge problem. The fact never became something the group could act on, because nobody knew the others knew.

Earlier in this series, the piece on the MUM Effect explained why bad news struggles to travel up. This is the floor below it. MUM keeps the news from arriving. Focal points explain why news that did arrive – that landed in every inbox – still produces nothing. The signal got through. The “everyone knows that everyone knows” never formed.

“I assumed everyone knew” is the failure point of most coordination.

The Slack Channel Is Not Common Knowledge

Modern work makes this worse, because we’ve confused broadcasting with common knowledge, and they are not the same animal.

A message posted to a channel of forty people and acknowledged by none of them is not common knowledge. An email sent to fifteen people with no replies is not common knowledge. It’s a group hallucination of alignment. You can see that the messages were sent. You cannot see whether anyone read it, registered it, or believes the others did. Everyone scrolls past assuming someone else has it – which means everyone assumes everyone else is on it, which means nobody is.

Political scientist Michael Chwe wrote a whole book on this, called Rational Ritual (2001). His unlikeliest exhibit: the Super Bowl ad. The assumption is that companies pay millions because it’s a rare opportunity to put content in front of several million people, only that’s not the reason. Chwe’s insight was that companies pay millions because everyone watching knows everyone else is watching. That’s why it’s better to spend a giant sum of the advertising budget on a Super Bowl ad instead of a series of smaller spots to reach more customers. The common knowledge – the thing only a public, shared, unmissable event can manufacture.

Your team rarely has Super Bowl events. It has a Slack thread, a periodic all-hands, and the amusing assumption that attendance equals agreement.

Why Acknowledgment Is the Whole Game

There’s a classic problem in distributed computing, posed by Jim Gray in 1978: the coordinated attack problem, also known as the Two Generals’ Problem. Two generals on opposite hills must strike at the same moment, but can only send messengers. One sends “attack at dawn.” But did it arrive? The other sends confirmation. Now did that arrive? Now the first general must confirm the confirmation – and so on and so on. Halpern and Moses (1990) later proved why: the two sides can never reach common knowledge, only increasingly hopeful odds.

Teams aren’t generals, and a quick “got it” or “I’m on it” usually closes the loop the generals never can. But the lesson stands: alignment isn’t the moment a fact is sent. It’s the moment everyone can see that everyone else received it.

Coordination doesn’t break where the information is missing. It breaks where the “everyone knows” is missing.

So the question for any stalled team isn’t did they get the message? It’s does each of them know the others got it too? The first is a delivery report. The second is a focal point – and only one of them gets everyone to Grand Central, at noon, no conversation required.

References

Chwe, M. S.-Y. (2001). Rational ritual: Culture, coordination, and common knowledge. Princeton University Press.

Gray, J. (1978). Notes on data base operating systems. In R. Bayer, R. M. Graham, & G. Seegmüller (Eds.), Operating systems: An advanced course (Lecture Notes in Computer Science, Vol. 60, pp. 393–481). Springer.

Halpern, J. Y., & Moses, Y. (1990). Knowledge and common knowledge in a distributed environment. Journal of the ACM, 37(3), 549–587.

Schelling, T. C. (1960). The strategy of conflict. Harvard University Press.

Michael Morand

Michael Morand is the founder of Evo Associates, a consulting firm that helps organizations build stronger, more effective teams. Drawing on nearly 20 years of leadership experience across the U.S. Military, Johnson & Johnson, and academia, he brings a background grounded in complexity science to his work on team dynamics. He holds a Master's in Economics from Oklahoma, an MBA in Strategy from NYU Stern, a Fellowship at MIT, and a Doctorate from Temple University.