When the problem isn't winning — it's agreeing

01 · THE SHIFTWhen rationality alone isn't enough

Lesson 04 gave you a powerful tool: Nash equilibrium — no player wants to deviate unilaterally. But it handed you something uncomfortable at the same time. Many real games have more than one Nash equilibrium.

Remember the stag-hunt from last lesson — you and a referring partner each choose whether to Adopt a new shared clinical platform or Keep your old tools. Two cells were Nash equilibria: both-Adopt and both-Keep. Neither of you wants to deviate from either. Rationality alone cannot tell you which one you'll end up in.

This lesson is entirely about that convergence problem. The answer involves a concept that is simultaneously obvious and underrated: the focal point.

The problem of equilibrium selection is a central research thread — see SEP — Game Theory §3 on refinements; the focal-point concept originates with Thomas Schelling, The Strategy of Conflict (1960).

02 · FOCAL POINTSThe salience that solves the puzzle

In 1960, economist Thomas Schelling ran a simple experiment. He asked people: "You've agreed to meet a stranger in New York City tomorrow. No time, no location agreed. You can't communicate. Where and when do you go?"

Most people said noon, at the information booth in Grand Central Terminal. Not because it was "optimal." Because it was obvious — a landmark both parties would expect the other to expect. Schelling called this a focal point (or Schelling point).

Now anchor this to a real situation you've been in. You and a colleague are running the skull-base course at HSJ. During the coffee break on day two, you get separated — phones dead, no prior meeting-point plan. You need to find each other before the afternoon session starts.

The focal point isn't the best spot. It's the spot that each person expects the other to expect. The main auditorium entrance, the registration desk, the coffee stand — whatever is most salient given your shared knowledge of the venue and each other. Shared salience does what communication would otherwise do.

Thomas Schelling, The Strategy of Conflict (Harvard, 1960) — the original meeting-in-New-York experiment. Popularized for practitioners in Dixit & Nalebuff, The Art of Strategy (Norton, 2008).

03 · BATTLE OF THE SEXESCoordination with a distributional edge

Pure coordination games are the easy case: both players are indifferent about which equilibrium, they just want to match. But many real coordination problems have a twist — both parties still need to match, yet they each prefer a different match. Game theorists call this battle of the sexes.

Here's your version. You and a high-volume referring partner are both losing efficiency because your referral workflows don't talk to each other. You both agree: coordinating on one platform beats not coordinating at all. But you've already invested in Platform A, and they've built around Platform B.

Read the cells: if you mismatch (A,B or B,A), both of you get 0 — chaos, no integration, wasted effort. If you both adopt A, you get 2 and your partner gets 1 — coordination achieved, you come out ahead. If you both adopt B, you get 1 and your partner gets 2 — coordination achieved, they come out ahead.

Notice: both (A,A) and (B,B) are Nash equilibria. At (A,A), you'd gain nothing switching to B alone — you'd drop to 0. Same logic at (B,B). Both are self-enforcing. The game's strategic tension is entirely about which equilibrium you end up in — coordination with a distributional fight built in.

04 · INTERACTIVERead the battle-of-the-sexes matrix

The distributional edge matters: whoever wins the coordination battle — getting the other side to adopt their preferred platform — earns the higher payoff. That's why first-mover commitment, public announcements, and setting the default matter so much in battle-of-the-sexes games. The side that moves first or makes a credible commitment often captures the favorable equilibrium.

Open Yale ECON 159 (Polak), Lectures 6–7: coordination games and multiple equilibria.

05 · SELECTION TOOLKITHow real players choose among equilibria

When you face a coordination problem — multiple equilibria, no dominant strategy pointing at one — four forces do the selecting:

• Focal points (salience). The option that stands out culturally, visually, historically, or by simplicity. In your SaaS tiers: if every competitor calls the mid-tier "Pro," that word has focal gravity — fighting it costs more than it's worth. The market's default is the focal point.
• First-mover commitment. Publicly announcing "we're going with Platform A from January" changes the game — your partner's best response to a committed player is now to match you. (This is a teaser for Lesson 06: sequential games and backward induction.)
• Communication and cheap talk. When interests align — both of you just want to pick the same thing — a simple pre-game message ("I'll be at the auditorium entrance") is credible and solves it. But when interests conflict, like in battle of the sexes, both sides have an incentive to claim their preferred equilibrium — "Platform A is better for us both" — and the other side knows this. Cheap talk loses its coordination power when preferences diverge.
• Conventions and standards. Industries converge on one format not because it's the best one — often it isn't — but because the coordination benefit of everyone being on the same thing outweighs the benefits of any individual superior alternative. This is how inferior standards can persist and why being the default carries real strategic value in healthcare IT.

For your skull-base course and SaaS work: when partners adopt your tool, you're often setting a focal point for future cohorts. The first platform a cohort's program director uses has outsized staying power — not because it's optimal, but because it's salient.

06 · KNOW THE LIMITSWhen a focal point won't save you

Focal points are powerful, but they only solve one specific problem — pure convergence. They don't fix a fight. Watch for these three failure modes before you lean on "just find the obvious option":

Naming when the tool doesn't apply is what keeps "find the focal point" from becoming a party trick. If interests conflict, salience isn't shared, or someone has a reason to lie — you need bargaining (Lesson 10) or commitment (Lesson 07), not just a landmark everyone recognizes.

07 · LOCK IT INWhat you just learned

• A game with multiple Nash equilibria doesn't get resolved by "be rational" — convergence needs something extra.
• A focal point (Schelling point) is an equilibrium selected by salience — the option each expects the other to expect. It doesn't have to be the best option, just the most obvious shared one.
• Pure coordination: players want to match and are (roughly) indifferent which equilibrium — focal points, conventions, or any signal work.
• Battle of the sexes: players both want to match but prefer different equilibria — distributional tension overlays the coordination problem. First-mover commitment and establishing the default tend to win.
• Cheap talk coordinates when interests align; it loses credibility when both sides want different equilibria and say so strategically.
• Conventions and standards are equilibria selected by history and network effects — disrupting them requires a clear coordination mechanism, not just a better product.

Primary sources: SEP — Game Theory (equilibrium selection) · Open Yale ECON 159 (Polak). Focal points: Thomas Schelling, The Strategy of Conflict (1960); Dixit & Nalebuff, The Art of Strategy (Norton, 2008).