10% of the code in the world is poorly written
Analysis
The claim that "10% of the code in the world is poorly written" lacks credible and verifiable evidence from trusted sources. The provided references are all non-trusted, anecdotal, or opinion-based discussions from forums, blogs, and social media platforms, which do not offer systematic studies or industry-wide analyses to substantiate such a precise figure. While it is widely acknowledged in the software development community that a significant portion of code can be suboptimal due to various factors like legacy systems, rushed development, or lack of expertise, no reliable data quantifies this as exactly 10%. The sources mostly discuss poor coding practices anecdotally or in specific contexts (e.g., data science, legacy business code) without generalizing to the entire global codebase. Therefore, the claim remains unsubstantiated and should be regarded as an unsupported estimate rather than a factual statement.
Sources
Discusses code coverage and quality metrics anecdotally without providing data on global code quality percentages.
Focuses on data scientists’ coding practices without quantifying poorly written code globally.
Personal opinion on some data scientists writing bad code, no broad statistical evidence.
Talks about variable naming issues, no global code quality claims or data.
Discusses C++ code quality traits, no global or quantitative claims.
Comments on academic code quality attitudes, anecdotal and context-specific.
Reddit discussion about legacy code quality, no reliable statistics.
Stack Overflow discussion on code quality perception, no data on percentages.
Speculation about AI-generated code quality, no global code quality data.
Personal workflow and AI code generation, no evidence supporting the claim.
Verify any claim in seconds
Download AI Fact Checker and check headlines, quotes, and claims with AI.