The reason why humans make seemingly irrational decisions is not a random malfunction of the brain, but is the product of specific learning mechanisms which, while evolutionarily advantageous, can fail in modern contexts. At the basis of our decisions is the “reinforcement“, a fundamental cognitive process shared with many animal species, which drives us to repeat actions followed by rewards and avoid those followed by punishments. However, recent research has shown that this system is not so much about accuracy, but about practical effectiveness: the brain is interested in surviving and getting immediate resultsnot necessarily having a perfect solution, even at the cost of an incorrect representation of reality. This gap between objective reality and subjective perception generates systematic biases that influence our choices in a predictable way.
What is repetition bias
The first major culprit of our poor choices is a form of cognitive inertia known as “repetition bias“. Contrary to classical economic theory, which imagines man as a rational calculator who evaluates each option based on its expected value, we tend to repeating a choice simply because we have already made it in the past, regardless of the result obtained. This phenomenon, which recalls the “Law of Exercise” formulated by Thorndike over a century ago, suggests that thevery act of choosing an option increases the probability to select it again in the future, creating a vicious circle of perseveration.
Recent studies have shown that this tendency to repetition alters our perception of value: we end up overestimating the options we often choose and develop a unjustified certainty on their goodness, even when objectively better alternatives exist. It has been observed that the repetition of a choice reduces the feeling of perceived uncertaintya mechanism that probably serves to save mental energy in complex environments. In practice, the brain prefers the consistency and ease of habitual action over the effort of constantly reevaluating alternatives, leading us to persevere in error even in the face of contrary evidence.
Slaves of the context
A second determining factor is the brain’s inability to evaluate options in isolation: we are, instead, slaves to context and optimism. Our internal evaluations are subject to distortions in the way we store and process information. If a mediocre option is presented in a context of terrible alternatives, it will seem excellent to us, and we will continue to choose it even when the context changes and it would become disadvantageous.
Added to this is the “positivity bias”: we tend to update our beliefs much faster when we receive “good news” than when we receive bad news. This inherent optimism leads us to overestimate our probability of success and to ignore warning signs, a trait that can foster resourcefulness but often leads to unrealistic assessments. Furthermore, we are naturally more confident when we try to make a profit than when we try to avoid a loss, even if the objective difficulty of the task is identical. The ventromedial prefrontal cortex plays a crucial role in this process, encoding a kind of “unconditional trust“subjective that integrates these affective biases, sometimes detaching itself from the real precision of our performances.
Same brain, different decisions
Finally, our poor choices are often the result of an ancestral conflict between different decision-making systems: The Pavlovian system (by the great Nobel Prize winning physiologist Ivan Pavlov) and that instrumental.
The Pavlovian system is a “pre-programmed” mechanism that pushes us to approach stimuli that promise rewards and to inhibit or flee when faced with threats of punishment. While useful in nature, this instinct can come into conflict with more complex modern goals requiring, for example, acting actively to avoid harm or holding back to gain a future advantage. When we find ourselves in contexts of potential loss or punishment, a slow reaction times and a decline in confidence, as if the brain “freezes” cognitive and motor resources.
Neurostimulation studies have shown that the Dorsolateral prefrontal cortex it is critical to suppressing these automatic impulses and instead fostering goal-oriented behavior. When this area of the brain fails to exercise effective control, the Pavlovian bias prevails and automatic reactions take over: thus we end up avoid necessary actions oh perform impulsive actions. Interestingly, stimulation of this brain area can specifically reduce bias, suggesting that our brain manages reward and punishment circuits through partially distinct pathways. As a result, the error is not just a logical short circuit, but often the result of one lack of neural mediation between immediate survival instincts and rational planning.
Sources
Degni et al., 2026, Reduced Pavlovian Value Updating Alters Decision-Making in Sign-Trackers. Wagner et al., 2024, Explaining decision biases through context-dependent repetition. Badioli et al., 2024, Unraveling the influence of Pavlovian cues on decision-making: A pre-registered meta-analysis on Pavlovian-to-instrumental transfer. Kim et al., 2023, Causal role of the dorsolateral prefrontal cortex in modulating the balance between Pavlovian and instrumental systems in the punishment domain. Palmintieri, 2025, Human reinforcement learning processes and biases: computational characterization and possible applications to behavioral public policy. Ting et al., 2020, Robust Pavlovian-to-Instrumental and Pavlovian-to-Metacognitive Transfers in human reinforcement learning. Ting et al., 2023, Neural and computational underpinnings of biased confidence in human reinforcement learning. Sosa et al., 2021, The Role of the Lateral Habenula in Inhibitory Learning from Reward Omission.
