Skip to content
TheChatArchiveThe Chat Archive
Claudeclaude-3-7-sonnet-20250219· AnthropicHugging Face LMArena Human Preference 140K

how is this ranking created

From a Chatbot Arena battleThe same person sent this prompt to two anonymous models at once. Compare it with grok-3-mini-high’s answer.
  1. User

    how is this ranking created

  2. Claude Assistant

    This question is a bit ambiguous without more context. Are you asking about a specific ranking system or methodology that was mentioned earlier? If you could provide more details about which ranking you're referring to, I'd be happy to explain how it's created or calculated.

  3. User

    the lmaarena rankings, is it just here? choosing one or the other?

  4. Claude Assistant

    The LM Arena rankings are typically created through a series of head-to-head comparisons between different language models. Users are presented with two AI responses to the same prompt and asked to choose which one they prefer. These binary choices accumulate to create an overall ranking.

    The system uses these pairwise comparisons rather than absolute scoring, similar to an Elo rating system used in chess. Models gain or lose points based on whether they "win" or "lose" these comparisons, and their relative position in the rankings adjusts accordingly.

    This approach helps create a more objective measure of model performance based on actual user preferences rather than predetermined metrics alone.

Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.