Game Theory in Real-Life Partnerships

Game Theory in Real-Life Partnerships

By Ken Phelan
Posted in Uncategorized
On February 12, 2020

Game theorists like to look at game play as a way of modeling human behavior. One of their favorite games is called the prisoner’s dilemma. You’ve probably heard of it. In summary, it goes like this:

You and a partner commit a crime. You’re not caught in the act, but the police bring you both downtown for questioning. The police put you in separate rooms and question you. At this point, the following outcomes are possible:

  • Option 1 - You both keep quiet and each spend one year in jail for a lesser crime. 
  • Option 2 - One of you betrays the other. The traitor goes free, the other does four years in jail. 
  • Option 3 - Both of you betray the other and you both do three years in jail. 

From a team perspective, it’s best to keep the secret. Only two years served total. 

Individually, it’s a different story. Your partner has either betrayed you or not. If he has, you’re better off betraying him too, getting three years instead of four. If he hasn’t, you’re still better off betraying him, getting off scot free vs. doing one year in jail. So if you’re good at math, or have read any books on game theory, you can see it’s in your best interest to betray your partner. 

As artificial as this seems, it bears a strong similarity to many real-life partnership scenarios. There are almost always strategies that are good for one party and toxic for the team as a whole. Here are some examples: 

  • You can sell a client something they won’t get value from. You make money, but in the long run, the client pays the price. 
  • A client can drive a deal into an unprofitable place for providers. The client wins, but in the long run, the providers are either driven out of business or are forced to cut quality or services. 
  • Manufacturers can work with one reseller throughout the sales cycle and then take a deal direct or move to a low-margin reseller in the eleventh hour. It’s good for the manufacturer, but bad for the reseller. 

Like our criminals, this bad behavior is almost always in one party’s short term best interests, often at the cost of the other partner.

Interestingly though, it doesn’t always happen, even though betrayal is obviously the better strategy. Do people not betray their partners out of morality or are they just bad at math? 

This is where game theory comes back into play. What if you’re playing the game more than once? On the surface, a team that learns to trust each other should beat teams that betray each other on a regular basis. 

Computer simulation and some intensive research provides answers. It turns out that the optimal strategy is a form of “tit for tat.” Start out by trusting each partner. If they betray you, then the next time you’re with them, you should do the same. If they cooperate, then the next time you’re with them, you cooperate. 

This is the basis of cooperation as an optimal strategy. We need to be aware that tomorrow we’re all going to wake up and have to play again. Cooperative players will be trusted, uncooperative players will not. 

Warren Buffet once said, “You can’t make a good deal with a bad person.” But how to tell a good person from a bad one? Well, based on this, it’s relatively simple. The person who’s playing again tomorrow is likely to be cooperative. 

Paul Roberts of “The Impulse Society” talks about the environment inside many of the financial firms leading up to the crash of 2008. As they were assembling the financial vehicles leading to the bubble, they knew it was a bubble. They knew it couldn’t last. But as they discussed it, they said when it goes bad, IBGYBG - I’ll be gone, you’ll be gone.

A couple of years ago, a manufacturer came to me to propose a deal for a mutual customer that we had been partnered in for a number of years. It was described as a scenario in which everyone wins. The manufacturer was going to move a bunch of new product into the customer and the customer was going to save money. The problem was, the customer didn’t actually need or want the extra product. The savings were also short term, with lower payments in year one, and larger payments as time went on. Gotham walked away from the deal and the deal went through another partner. We don’t operate in IBGYBG mode. 

The next time you’re working a deal, look for the IBGYBG people at the table. Who will still be dealing with the outcomes of this deal next year and the year after that? Who is interested in playing again tomorrow? And when you find the IBGYBG’s, don’t be surprised when they betray you. It’s just good math.

Ken Phelan

Ken Phelan

Ken is one of Gotham’s founders and its Chief Technology Officer, responsible for all internal and external technology and consulting operations for the firm. A recognized authority on technology and operations, Ken has been widely quoted in the technical press, and is a frequent presenter at various technology conferences. Ken is the Chairman of the Wall Street Thin Client Advisory Council.