The human desire for revenge, even when self-destructive, has important implications for politics and game theory. A study found that 90% of subjects retaliated against a simulated Soviet nuclear first strike despite knowing they were dooming themselves.

This is irrational from a self-preservation standpoint but reflects that revenge is a powerful innate drive, perhaps evolutionarily adaptive for deterring aggression against one's tribe. Mutually assured destruction works because of this glitch in human reasoning.

The instinct for retaliation deters conflict but makes de-escalation difficult. Game theory's assumption of pure self-interest misses these messy human motives. Understanding them is key to anticipating others' actions, both at the poker table and in geopolitics.

Section: 2, Chapter: 8

Author: Nate Silver

The advent of computer solvers has revolutionized the game of poker. These AI-aided programs, developed by pioneers like Piotr Lopusiewicz, find the game theory optimal (GTO) solution for poker by running through billions of iterations. They reveal that optimal play often involves frequent randomization and mixed strategies to remain unexploitable. As a result, modern poker players devote hundreds of hours to studying solver outputs, bringing game theory into practice at an unprecedented level. While humans still maintain an edge through psychological reads and exploitative play, the technical gap between top players and computers is rapidly shrinking.

Section: 1, Chapter: 1

Author: Nate Silver

One of the most important skills for top poker players is strategic empathy - the ability to put yourself in your opponent's shoes and understand how they perceive the game state. By paying attention to subtle physical tells, betting patterns, and emotional states, elite players like Vanessa Selbst and Maria Ho gain an edge by narrowing their opponent's range and finding exploitative plays. However, this is a skill that needs to be carefully honed and applied, as making hero calls or huge bluffs based on a hunch can be disastrous without hard evidence to support a read.

Section: 1, Chapter: 2

Author: Nate Silver

The ludic fallacy refers to the misuse of games, gambling, and probability theory to model real-world phenomena. It occurs when we confuse the clean, tidy, clearly defined randomness found in games of chance with the far more complex, open-ended randomness found in real life. An example is equating the odds of winning a casino roulette spin (where probability can be cleanly defined) with the odds of a real-world event like a war breaking out (where probability is far more intractable). Those who spend too much time in artificial, ludic environments can become "nerds" - people who think explicitly about probability but fail to understand how randomness operates in the real world.

Section: 1, Chapter: 9

Author: Nassim Nicholas Taleb

Bitcoin is the dominant cryptocurrency largely because it serves as what game theorists call a "focal point" - an option people tend to choose in the absence of communication because it seems natural or special. Bitcoin benefits from being the first and most well-known cryptocurrency.

If users want to transact seamlessly, the community needs to converge on one or a few cryptocurrencies. Any one cryptocurrency would work, but Bitcoin has the advantage of incumbency and brand recognition. This makes it the default choice and very difficult to displace, even by technically superior alternatives.

Section: 2, Chapter: 6

Author: Nate Silver

Sam Bankman-Fried (SBF) had an extremely unhealthy view of risk, believing it was correct to take gambles that could cause serious damage to one's position as long as there was a small chance of a huge payoff.

SBF explicitly extended this philosophy to existential risk, telling colleagues he'd happily "flip a coin, if it came up tails and the world was destroyed, as long as if it came up heads the world would be like more than twice as good."

Combined with his utilitarianism and earning-to-give mentality, SBF's approach to risk was a recipe for unhinged, destructive decision making with no regard for downside.

Section: 2, Chapter: 8

Author: Nate Silver

The Prisoner's Dilemma is the most famous scenario in game theory: two prisoners are being interrogated separately. If both stay silent, they each get a short sentence. If one rats the other out, the snitch goes free while their partner gets a long sentence. If they both snitch, both get medium sentences. The rational play is to always snitch.

The question is, how can cooperation ever emerge in a dog-eat-dog world? The surprising answer: forgiveness. Computer simulations of evolutionary processes show that the most successful long-term strategies for iterated Prisoner's Dilemma games are "nice" ones - they start out cooperating and only defect if their partner does first. But critically, they also quickly return to cooperating if the partner starts behaving nicely again.

Tit-for-tat, but with reconciliation. Strict punishment of defectors, but no grudges. Why does this work? Essentially, being forgiving avoids an endless cycle of retaliation. One party's selfish action doesn't lead to an arms race of increasing betrayal.

This sheds light on the evolutionary origins of seemingly irrational human behaviors like empathy and mercy. Natural selection can counterintuitively favor "nice" strategies.

Section: 1, Chapter: 11

Author: Brian Christian