Java performance anti-patterns like string concatenation in loops create O(n²) copying costs, while autoboxing in hot paths generates unnecessary GC pressure. Jonathan Vogel's analysis in Hacker News highlights how developers inadvertently slow their applications through these subtle mistakes. The article details how streaming inside loops causes repeated full collection scans per element, while String.format() proves the slowest string-building method due to parsing overhead.

Vogel also warns against using exceptions for control flow, which forces full stack traces, and too-broad synchronization that becomes concurrency bottlenecks. Repeated object creation of reusable tools like ObjectMapper incurs constructor costs on every call, and virtual thread pinning from synchronized blocks in JDK 21-23 creates subtle performance drains. Vogel's piece serves as a practical guide for developers auditing their Java codebases. Developers should audit their Java codebases for these patterns to avoid hidden performance costs.