If you've watched enough detective shows, you've seen a scene like this: two criminal co-conspirators are arrested and placed in separate interrogation rooms. A couple of no-nonsense cops shake down conspirator #1, telling him his buddy in the other room is about to rat him out. They tell him his only choice is to confess, testify against the other guy, and get a deal. Meanwhile, conspirator #2 is told it's #1 who's on the verge of confessing. If they both confess, they're both going to jail. If neither confess, the case against them is weak and they can beat the rap.
But here's the deal: They're in separate rooms, and unsure what the other guy is going to do. So what happens? They each confess as quickly as they can. They throw their accomplice under the bus, hoping to do so first, because they'd rather go to jail with a deal than get a harsher sentence by saying nothing. They're caught in what's called a "prisoner's dilemma."
Many situations in real life are essentially prisoner's dilemmas. Athletes take performance-enhancing drugs, because there's no guarantee that their competitors won't. Students in highly competitive academic environments are reluctant to help others study for fear that the favor won't be returned, and they will be out-performed.
Some prisoner's dilemmas are iterated. This is when you find yourself with the same person (or people) for an extended period of time, asking over and over again, "should I cooperate or should I screw this person before s/he screws me?" Think Cold War. Or political advertising.
Or divorce. Reaching a divorce agreement involves a series of negotiations during which each party has to decide: Do I cooperate and risk getting taken advantage of? Or should I protect myself by not being too nice? One "round" of an iterated prisoner's dilemma game might involve conversations about who gets the house. Another might occur as the terms of child custody are negotiated. Throughout the process, when it comes to the car, the savings account, the family heirlooms, etc., the divorcing couple has to repeatedly decide whether to cooperate or whether to, as game theorists call it, "defect."
What's interesting about an iterated game is that if you provoke your opponent too much during one round of negotiations (over the house, for instance), they'll remember it as they enter the second round (over the kids, for example). Thus, decisions made in each iteration could have long-lasting repercussions. When both parties constantly defect, over time they can end up with a nasty divorce. On the other hand, cooperating too much carries its own risk: if you always give in, maybe in the spirit of being the better or bigger person, and they always defect, you could end up in pretty bad shape.
Ideally, two parties would be able to cooperate and file for an uncontested divorce, which is relatively easy to get these days and spares you from the pit of rage and exhaustion that can accompany an all-out war. But not all spouses are amenable to cooperation, especially if they're heart-broken, angry, or even narcissistic.
So if you're facing divorce, what should you do? What's your best strategy? As it happens, experts on the subject use computer simulations to answer this question. A simulation is a make-believe and simplified reproduction of the real world. Prisoner's dilemma simulations use algorithms to create "players" or "actors" with unique strategies, then pit these actors against each other in an iterated prisoner's dilemma game. Mock scenarios are played out to see which strategies are superior against others. For instance, a game can put a "Jesus" (always cooperates) up against a "Lucifer" (always defects) to see which fares better. (You'll be disheartened to learn that in this scenario Lucifer wins).
A particularly interesting player is one who employs a "tit-for-tat" strategy: they start out cooperating (because why not be nice first?) but then do whatever the other player did on their last move. If the other player defected in the last round of the simulation, the tit-for-tat actor is programmed to defect in the current round. If the other player starts cooperating, the tit-for-tat player starts cooperating too. When tested against a variety of different players, the tit-for-tat strategy is a strong winner. But not always. It can be beaten by a Lucifer, because matching Lucifer's moves leads the game down an eye-for-an-eye-until-the-whole-world-is-blind situation. However, if the tit-for-tat algorithm is tweaked to become a "generous tit-for-tat strategy" - so that for every ten moves the tit-for-tat player cooperates even as their opponent continues to defect - the downward spiral isn't as severe, and the damages not as great.
What does this mean for divorcing couples? If computer simulations tell us a modified tit-for-tat strategy puts you in a better spot against all other opponents, then perhaps acting like a modified tit-for-tat player during a divorce is the way to go. If your spouse is trying to bilk you out of everything, then do the same, but be nice once in a while. If your spouse is throwing out an olive branch, then set aside your rage and reciprocate, for the sake of an optimal outcome.
A "successful" divorce doesn't mean doing better than your spouse. It means minimizing your emotional scars and arriving at an outcome that puts you in a better situation than all possible alternatives, which in many cases is the most you can hope for. So start out nice, then do what your soon-to-be ex does. And if he or she is never willing to cooperate, then suck it up and cooperate once in a while. It might not feel great. But it's the scientific way to "win" at divorce.