Behavioral Software Engineering
summary
Compares behavioral-economics biases to everyday engineering decisions and gives concrete examples: removing reflection from non‑critical paths wastes effort, benchmarks fluctuate ±5% so a claimed 3% win is likely noise, and synthetic benchmarking produces overfitting. Recommends practices that worked: measure everything and be skeptical, file fast and prototype to avoid sunk cost in code reviews, reserve effort for refactoring and quantify technical debt; also warns about anchoring, priming, temporal discounting and the cobra‑effect of metric-driven incentives.
AI-written summary · about
chapters (9)
- 0:00Behavioral software engineering
- 4:58Patternicity in development
- 11:28Cobra effect and metrics
- 12:13Anchoring in decisions
- 16:45Sunk cost fallacy
- 19:37Temporal discounting and decay
- 22:10Status quo bias
- 26:08Framing effect and loss aversion
- 29:28Cognitive biases roots
related talks
01h 02mCognitive biases in software development
01h 00mCognitive biases in software development
39mHow Cognitive Biases Affect our Software Architectures
43mHow to deal with the social side of computing life
53mHow Not to Be Wrong in Software Engineering
21mDerailing irrationality
44mThe Impact of Behavioral Biases to Real Data Science Projects
38mCognitive Bias: A reflection and path forward