Against Tit-for-Tat as a Heuristic
This is probably obvious to some of you, but flawed humans like me need reminding
The dominance of tit-for-tat as a strategy in game theory is well-established. It maximizes cooperative gains while protecting against exploitation. Robert Axelrod's seminal work, "Evolution of Cooperation," could be distilled to a simple maxim: "Be nice by default, unless given reason not to."
While tit-for-tat proves remarkably robust in games and game-like scenarios, internalizing it as an ethical heuristic can be surprisingly destructive. In life, choices and payoffs are infinitely fuzzier, and our situational awareness is systematically biased.
In our relationships, we aren't confined to binary choices of "cooperate" or "defect." What appears as a partner's "defection" might be unintentional: perhaps they spoke dismissively or failed to provide support when needed. Even if an objective observer might label it as defection, your partner likely has justified it to themselves differently.
More importantly, your psychology is always on your side. We make decisions and perceive reality through our perceptive and emotional filters. Sometimes we find ourselves retaliating not from strategic choice, but from anger or hurt. Our emotional apparatus, evolved for self-preservation, colors our perception toward self-interest. A good relationship heuristic should account for and counterbalance this bias.
Responding with your own "defection"—by withholding affection or support—risks triggering a cascade of mutual retaliation, with both parties feeling equally justified. Relationships offer a crucial alternative to action-reaction cycles: communication and to some extent, norms.
During a charged exchange with your partner the seductive game theoretic thought appeals to you: 'If I let this slide, what precedent am I setting?' This reveals we're already rationalizing retaliation rather than following a robust heuristic. Instead of emotionally charged tit-for-tat, relationships need clear boundaries established in advance and protocols for communication when those boundaries are crossed. What feels intuitively like tit-for-tat in the moment won't accomplish what the bot version accomplishes in a game.
The tit-for-tat mindset often appeals to driven, ambitious individuals. It's adaptive in low-trust contexts where exploitation is common—business deals with unfamiliar parties, political negotiations, or competitive markets. In these domains, having a careful eye for reciprocity protects one's interests. But even in lower trust domains, "tit for tat" as a vibe can be misleading.
Suppose someone breaks established norms by pulling out of a business deal last minute without cause. The move that feels most intuitively like tit-for-tat—engineering another deal just to burn them back—is obviously unwise, even in purely self-interested terms. True strategic reciprocity might look more like carefully imposing reputational costs. Even in physical conflict, our most primal domain, martial artists train for years precisely to override emotional tit-for-tat impulses. They develop automatic responses that work across scenarios without requiring heated calculations in the moment.
Think through why vigilante justice wreaks such havoc. Civilization itself emerges from recognizing that we're all agents optimizing for self-interest most of the time. Across different domains—relationships, business, or even physical conflict—we need robust norms and heuristics to protect our interests. But these rarely involve calculating, in the heat of the moment, whether it's worth scoring a retaliatory point. The mechanisms that actually prevent exploitation look more like clear boundaries, established protocols, and measured responses—not the immediate emotional satisfaction of getting even.
To friends, family or strangers that have been at the receiving end of my reflective tit-for-tat, let me take this opportunity to apologize deeply and sincerely.
Tit for tat is a robust evolutionary strategy in finite, adversarial games where zero-sum is implicit. If the ecosystem is saturated with suckers, then other models will outperform it. But life is not a finite game, and game theory isn't an appropriate way to model it.
Recommending Alfie Kohn's, No Contest, the Case Against Competition for interesting related insights.
Wise & profound - very well written.