In a mathematical game, two or more players, each with her own information, make moves. The outcome of moves is a payoff (often probabilistic) for each player and, commonly, a state from which to play the next round of the game. Players optimize rationally – or almost-rationally – to maximize their expected payoffs.
What game theory can’t do
If you want to numerically calculate something about the real world, that kind of game theory rarely helps. Major reasons include:
Those “payoffs” (i.e. quantified preferences) and “information” (which includes possibly incorrect beliefs) are hard to estimate.
Further, people can have incentives to conceal their true states of mind.
If, despite these difficulties, you magically knew every number within a small range of error, game theory might give you fragile and unstable results nonetheless. (Even the famed Prisoner’s Dilemma lacks a useful calculated solution.)
The rationality assumption is … questionable.
How game theory does help
But less formal game-theoretic thinking can be very useful. For example, if you want to analyze a business, you should wonder:
Who are they selling to? (Identify some of the players)
Who are they competing with? How might the set of competitors change? (Ditto)
Who else are they engaging with, such as suppliers, regulators, or press? (Ditto)
What motivates those buyers, competitors, or partners? What are their decision processes and criteria? (Payoffs, information, perhaps moves – basically the whole game shebang)
What alternate strategies could they employ? (Moves and more)
If you look through my best-regarded post on business strategy, a strategic worksheet for tech companies, almost every item has that kind of game-theoretic flavor.
Know the players
Theoretically, every individual Ukrainian, Russian, American or Pole might be a separate player in the Russia-Ukraine mess. But of course, nobody would analyze things that way. Rather, they’d consider who the individuals, groups and organizations are that could act independently of each other, then assess each of the substantive players’ motivations, capabilities, strategies, and likely moves.
Mistakes in such analyses are legion. Major Russia-Ukraine examples include:
Putin thought Ukraine contained multiple factions that would turn on each other in important ways. He was wrong.
Putin evidently thought his military could execute his orders competently. If so, he was largely wrong.
Putin evidently thought the Ukrainian military would be similarly (in)competent to how it was in 2014. He was wrong.
Many people thought Putin wouldn’t launch a high-cost military operation with little chance of positive practical payoff. We were wrong.
Russians probably thought the West wouldn’t unite about economic sanctions. So far, that view has been correct about oil and gas imports, but wrong about almost everything else.
That list could be a lot longer.
Business examples are even more common. For most of the past 40 years, I’ve been an analyst of or a consultant to technology companies, both of which roles involve noticing errors that smart people make. It turns out that many of those errors violate the “Know the players” dictum. Common templates include:
Underrating or overrating actual competitors.
Underrating or overrating potential new competitive entrants.
Misjudging what potential buyers want.
Forgetting how many different people, roles and interests are engaged in a single corporation’s buying decision.
Incorrectly viewing the influencer community as a single homogenous whole.
That list could be much longer too.
Basic discipline for avoiding know-the-players errors goes something like this:
If there’s a complex decision process (e.g. a corporate buying decision, a law-making negotiation, etc.), research how it actually works.
Consider who (else) has something at stake in a situation.
Consider who (else) has agency for affecting the outcome.
Analyze their goals, opinions and capabilities as best you can.
Some mistakes are inevitable, since you’re assessing other people’s states of mind. But analytic discipline may help you avoid a lot of preventable blunders.
Change the rules
Now we get to the creative part: If you want to improve your prospects in a real-world game, change its rules. In one ancient example, Sun-Tzu suggested that an army should burn its boats and bridges, so that it would be united in the realization that there was no turning back. Alexander the Great actually used that strategy. So, two millennia later, did Hernan Cortes.
In game-theoretic terms, boat-burning eliminates what otherwise might have been playable moves, such as:
Commander: Order a retreat.
Officers: Mutiny and retreat.
Soldiers: Desert and retreat.
The remaining situation allows for few outcomes other than victory or death.
Donald Trump owes much of his political success to the violation of norms. Except for any parts that were provably illegal, that’s a change-the-rules kind of strategy. Vladimir Putin is a norm-breaker too, internationally, wreaking vast amounts of death and destruction. It will be hard to stop him unless the West counters with some taboo-breaking of its own.
Other areas where rule-changing can be important include military tactics, electoral law, administrative law, contract law, market positioning, and negotiation. I plan to discuss most of those areas in future posts.
Change the whole game
“Change the rules” isn’t always the best formulation; sometimes you change the whole game.
The Founding Fathers did that when they drafted the United States Constitution. Federal Express, Microsoft, eBay and many others did that when inventing new models of business. I plan to return to this theme as well.