This post first appeared at harxsan.github.io .

In a large study of RAG pipelines, the three retrieval techniques with the biggest separately measured gains predicted a +46.6% NDCG@10 lift over the default_standard baseline, which scores 0.412. The balanced pipeline that combines them realizes only +17.2%, a 63% shortfall against that prediction.

The three techniques are cross-encoder reranking, disabling MMR, and document-awareness prefixes. Each one looks like a clear win in its own ablation. Together they deliver much less than their separate results add up to.

Why it matters

Most teams decide what stays in a RAG pipeline one ablation at a time. If a stage wins its ablation, it goes in, and the expected gain of the full stack is the sum of the wins. Every stage adds latency and cost to every query, so when the sum does not hold, the team pays for gains it never sees.

HyDE is a small example from the same paper.…