The joint standard problem

Two technology companies, one shared opportunity, and a structural tension that game theory has studied since Rousseau. Axis is an established platform infrastructure player with a risk-averse board, an installed base, and a strong preference for predictable outcomes. Lumen is a growth-stage challenger with a distribution advantage, faster iteration cycles, and a higher risk tolerance baked into its investor expectations. Their asymmetry matters because it shapes how each player reads uncertainty about the other's choice.

Each firm faces exactly two strategies: commit to a joint interoperability standard (Hunt Stag) or ship a proprietary, incompatible system (Hunt Hare). The metaphors come from Jean-Jacques Rousseau's parable: a group of hunters can cooperate to catch a stag, which feeds everyone, or each can defect to catch a hare independently, which feeds only that hunter. The stag requires everyone to commit; the hare does not.

The payoff structure sets this apart from the prisoner's dilemma. In a prisoner's dilemma, defection is a dominant strategy: whatever the other player does, you are better off defecting. Here, neither strategy dominates. If both firms commit to the joint standard, each earns a payoff of 8. If one defects while the other commits, the defector earns 3 and the cooperator earns 0. If both go proprietary, each earns 3. Cooperation is not dominated; it is simply risky. The risk is that you bear the full coordination cost alone if the other player does not show up.

The stag, in this case, is a shared interoperability standard that opens a large joint market. The hare is a proprietary system that captures a smaller but certain niche. Both are rational choices. The game does not resolve through dominance analysis. It resolves through belief.

Each player's optimal move depends entirely on what they believe the other player will do. This is the defining feature of the coordination game, and the source of its difficulty. There is no calculation that resolves the question without a prior belief about the other's action.

Consider Axis. If Lumen commits to the joint standard, Axis earns 8 by joining and only 3 by going proprietary. The correct response is to join. If Lumen goes proprietary, Axis earns 0 by joining and 3 by going solo. The correct response is to go proprietary. The payoff from each of Axis's strategies is entirely conditional on Lumen's choice. The same logic applies symmetrically to Lumen. Neither player has a dominant strategy. The game is genuinely contingent on mutual belief.

The asymmetry between the two firms enters here in a subtle way. Axis, with its risk-averse board, weights the (0,3) outcome heavily. The prospect of bearing the full coordination cost while Lumen captures its niche is the nightmare scenario. This risk-aversion pushes Axis toward the safer equilibrium. Lumen, with higher risk tolerance and a growth mandate from its investors, can absorb a bad outcome more easily. Lumen is therefore more willing to commit first, which is strategically significant for Phase 5.

BATNA analysis reinforces this. Axis's outside option is its existing proprietary infrastructure, which already generates revenue. Going proprietary is not a failure for Axis; it is a defensible fallback. Lumen's outside option is weaker: without a standard, it faces a fragmented market and higher customer acquisition costs. Lumen therefore has more to gain from the joint standard and more to lose from the stalemate of mutual defection.

Pareto optimal equilibrium Risk-dominant equilibrium Unstable outcome

Figure 1. Both Nash equilibria plotted. The upper-right quadrant (stag) is Pareto optimal; the lower-left (hare) is risk-dominant. Axis alone cannot pull the game to the better outcome. The two unstable off-diagonal outcomes, where one firm is exposed, are the outcomes each player fears most.

Four strategy combinations exist, and none can be eliminated by dominance. This is the structural difference from the prisoner's dilemma. In that game, the dominant strategy analysis produces a unique surviving strategy for each player. Here, no row is dominated and no column is dominated. Both strategies survive for each player.

The two off-diagonal outcomes are not equilibria: in each case, the player who committed to the joint standard would gain by unilaterally switching to proprietary (earning 3 instead of 0). These outcomes are therefore unstable. The two diagonal outcomes are the only ones that survive as Nash equilibria.

The critical distinction is between payoff dominance and risk dominance. (Joint Standard, Joint Standard) is payoff-dominant: it produces the highest total payoff and the highest individual payoffs for both players. Neither player can do better by switching, and the outcome is Pareto optimal. (Proprietary, Proprietary) is risk-dominant: it is the safer choice under uncertainty about the other player's action. If Axis assigns even probability to Lumen's two strategies, the expected value of going proprietary (3) exceeds the expected value of going joint (4, but conditional on Lumen also committing). Risk-dominance is not about who earns more; it is about which equilibrium is less exposed to the fear that the other player might not show up.

Both equilibria are self-enforcing once reached, which is why the game cannot be resolved by reasoning alone. Logic confirms both are stable but offers no basis for choosing between them.

The concrete move is not a strategic action inside the game but a coordination investment before the game resolves. Because both equilibria are Nash equilibria, the way to reach the better one is to shift the shared expectation of both players toward it. This requires three types of intervention: a credible commitment, a visible first mover, and an off-ramp that removes the fear of the worst outcome.

First, schedule a joint declaration of intent with a real cost to each party. A low-cost announcement that either firm could retract cheaply does not move beliefs. A joint roadmap published to potential customers, a shared developer portal, or a co-signed letter to a standards body all carry a reputational cost if either firm defects afterward. The cost of the signal is what makes it credible. Cheap talk does not raise the probability that Axis will commit, because Lumen knows it can be ignored. Costly signalling does.

Second, Lumen should commit first and make the commitment visible. Lumen has the higher risk tolerance and the stronger incentive to reach the (8,8) outcome. A unilateral visible commitment from Lumen raises the probability that Axis will follow: if Lumen is already committed, Axis's expected payoff from joining is no longer 4 (half of 0 plus half of 8) but closer to 8, because the uncertainty about Lumen's action has been resolved. Lumen absorbs the risk of first-mover exposure; in return, it shifts the game's focal point decisively toward (Joint Standard, Joint Standard).

Third, build an off-ramp. The fear of the (0,3) outcome is what drives risk-averse players like Axis toward the hare. An explicit agreement that if only one party commits by a defined date, both revert to proprietary development without penalty removes this fear. The off-ramp does not change the payoffs of the Nash equilibria; it changes the perceived risk of the transition period. Axis is more willing to commit if it knows it will not be left exposed alone. With the off-ramp in place, the coordination problem shrinks: both firms need only believe that the other will commit before the deadline, not that the other will commit unconditionally.

The game changes as the environment around it changes, and four shifts in particular can flip which equilibrium prevails. Coordination games are sensitive to changes in the shared expectation landscape. An event that shifts what either player believes the other will do can move the game from one equilibrium to another without altering the payoffs at all.

The most powerful shift is a public anchor: a regulator or industry body that endorses the joint standard as the preferred direction for the market. This creates a focal point in the sense of Thomas Schelling: a point that stands out as the natural, obvious choice, even without communication. When a regulator names a standard, the equilibrium selection problem simplifies. Each player now expects the other to choose Joint Standard because defecting from a publicly endorsed standard carries a reputational and regulatory cost that is not in the original payoff matrix. Focal points do not change payoffs; they change beliefs, and beliefs are what the coordination game runs on.

A second shift is the entry of a third competitor. If a new entrant arrives with its own proprietary system, the (Proprietary, Proprietary) equilibrium between Axis and Lumen becomes less attractive: two proprietary systems in a market with a third entrant produces fragmentation that damages both. The entry changes the game's topology, making (Joint Standard, Joint Standard) between Axis and Lumen the only option with enough combined market power to resist the entrant. Coalition logic replaces coordination logic.

A third shift is a change in Axis's risk profile. A new board, a new CEO with a growth mandate, or a failed proprietary product launch can raise Axis's risk appetite. As Axis becomes less risk-averse, the risk-dominant equilibrium loses its pull. The threshold at which Axis prefers the joint standard even under uncertainty shifts downward. What was a coordination problem becomes a near-trivial commitment problem.

A fourth shift is an expansion of the market for the joint standard. If adjacent use cases emerge that only a shared standard can serve, the payoff at (8,8) rises above 8 while the payoff at (3,3) remains unchanged. As the gap between the two equilibria widens, the joint standard becomes increasingly attractive even to risk-averse players. The coordination problem does not disappear, but its difficulty decreases as the asymmetry between the two equilibria grows. Monitor adjacent market developments as a leading indicator: a market expansion that only a standard can capture is an argument to invest in coordination mechanisms now, before a competitor acts as focal point instead.