When do you use pairwise versus pointwise evaluation for LLM outputs, and what does each get wrong?
Pairwise asks which one wins, pointwise asks how good a single output is. The signal is knowing which is more reliable, how Elo/Bradley-Terry aggregates comparisons, and where each method quietly misleads.
Updated Sep 2026 · Grounded in real GenAI, LLM, and AI/ML engineering interview loops and written to a senior-engineer editorial bar.
Pairwise asks which one wins, pointwise asks how good a single output is. The signal is knowing which is more reliable, how Elo/Bradley-Terry aggregates comparisons, and where each method quietly misleads.
Lead with where the obvious approach breaks, because that is the judgment they are screening for — most candidates jump straight to the happy path and lose the room.
Then walk the failure back through the pipeline in order, naming the one metric the customer's exec sponsor actually cares about before you propose the fix.
