The 12 mental models and biases for engineers

The article argues that thinking quality matters in software development and management because mental models and cognitive biases shape problem-solving, decisions, and collaboration. It presents several models and biases that help engineers make better choices and avoid common reasoning errors.

Reading notes

#

• Second-order thinking asks what follows after the immediate result and is useful for spotting longer-term consequences before acting.
• First-principles thinking breaks a problem into its basic parts instead of relying on assumptions or analogies.
• The circle of competence marks what you know, what you know you do not know, and what you do not know you do not know.
• Inversion means asking what could make a system fail instead of only asking how to make it work.
• The Pareto principle says roughly 80% of effects come from 20% of causes, so focus should go to the highest-impact factors.
• The sunk cost fallacy is the tendency to keep investing in a failing project because resources have already been spent.
• Confirmation bias shows up when people favor evidence that supports their existing beliefs and overlook conflicting evidence.
• Type 1 decisions are irreversible and consequential, while Type 2 decisions are reversible and less impactful.
• Occam’s Razor favors simpler explanations, designs, and fixes when they can address the problem.
• The Dunning-Kruger effect describes how people with little knowledge may overestimate their competence, while experts may underestimate theirs.
• The map is not the territory means models and diagrams are only abstractions and do not fully capture reality.
• The article also mentions Parkinson’s law, where work expands to fill available time, and links it to time-boxing and prioritization.