### Elo vs Bradley-Terry: Which is Better for Comparing the Performance of LLMs?

March 17, 2024 | 4 min readChatbot Arena updated its LLM ranking method from Elo to Bradley-Terry. What changed? Let's dig into the differences.

Under the sea, in the hippocampus's garden...

10 posts tagged with "math"

Chatbot Arena updated its LLM ranking method from Elo to Bradley-Terry. What changed? Let's dig into the differences.

This post attempts to take a deeper look at F1 score. Do you know that, for calibrated classifiers, the optimal threshold is half the max F1? How come? Here it's explained.

This post steps forward to multiple linear regression. The method of least squares is revisited --with linear algebra.

This post summarizes the basics of simple linear regression --method of least squares and coefficient of determination.

Is the sample correlation coefficient an unbiased estimator? No! This post visualizes how large its bias is and shows how to fix it.

The correlation coefficient is a familiar statistic, but there are several variations whose differences should be noted. This post recaps the definitions of these common measures.

How come ROC-AUC is equal to the probability of a positive sample ranked higher than negative ones? This post provides an answer with a fun example.

Causal inference is becoming a hot topic in ML community. This post formulates one of its important concepts called doubly robust estimator with simple notations.

How does Google's PageRank work? Its theory and algorithm are explained, followed by numerical experiments.