Understanding decision errors
Source: https://academic.oup.com/book/62545/chapter/559822431 ↗
Palminteri and Wyart have established themselves as a productive collaboration in the library's decision-making research, with their previous work scoring 8 for bringing computational insights to human judgment under uncertainty.
This follow-up work likely extends their framework for understanding systematic biases in decision-making — the kind of foundational cognitive science that helps product leaders understand why teams consistently make predictable errors in estimation, prioritisation, and strategic choice.
The decision-making theme remains underdeveloped in the library relative to its importance for product direction, and these authors have proven they can bridge the gap between laboratory research and practical organisational insight.
Their approach to decision errors as computational rather than merely psychological offers product people a more precise framework for designing processes that account for human limitations.
Central argument
Palminteri and Wyart argue that decision errors are not random noise or simple cognitive failures but structured, predictable outputs of computational processes in the brain — meaning that systematic biases in judgment arise from the architecture of how the mind models uncertainty and reward, not merely from irrationality or lack of information. Their framework treats errors as signals with diagnostic value: understanding the generative mechanism behind a mistake reveals which part of the decision process broke down and under what conditions it will break down again. This positions decision error not as something to be corrected through willpower or better incentives, but as something to be managed through process design that accounts for known computational constraints.
Critique
The computational framing, while precise, risks a kind of explanatory closure: mapping errors onto neural or algorithmic mechanisms can make the account feel complete while leaving the social and contextual dimensions of organisational decision-making largely unaddressed. A team's estimation errors in a roadmap review are not simply the sum of individual computational biases — they are shaped by hierarchy, psychological safety, and shared mental models that the laboratory paradigms underlying this research typically neutralise by design. The translation from controlled cognitive science to messy group decision environments therefore requires assumptions the work may not adequately foreground.
Why it matters for product
For a CPO, the key implication is architectural: if estimation and prioritisation errors are structurally predictable rather than accidental, then the right response is to redesign the decision process itself — the sequencing of information, the structure of options presented, the timing of commitment — rather than investing in training people to think more carefully. This is directly applicable to recurring failure modes in product organisations: teams that consistently underestimate complexity, overweight recent evidence in roadmap decisions, or anchor strategy to the first framing they encounter are exhibiting computational regularities that structured process design can partially neutralise.