# Generalized Bradley-Terry Models for Score Estimation from Paired Comparisons

arXiv: https://arxiv.org/abs/2308.08644 · HTML: https://mlsanity.com/paper/2308.08644/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 16 Aug 2023.

## Built on by (most cited first, 9 of 9)

- [Arena Learning: Build Data Flywheel for LLMs Post-training via Simulated Chatbot Arena](https://mlsanity.com/paper/2407.10627/) (2024)
- [AI Alignment at Your Discretion](https://mlsanity.com/paper/2502.10441/) (2025)
- [From Efficiency to Equity: Measuring Fairness in Preference Learning](https://mlsanity.com/paper/2410.18841/) (2024)
- [Generalizing while preserving monotonicity in comparison-based preference learning models](https://mlsanity.com/paper/2506.08616/) (2025)
- [PieArena: Ranking and Profiling Language Agents in Realistic Negotiation Scenarios](https://mlsanity.com/paper/2602.05302/) (2026)
- [On Monotonicity in AI Alignment](https://mlsanity.com/paper/2506.08998/) (2025)
- [Designing User-Centric Metrics for Evaluation of Counterfactual Explanations](https://mlsanity.com/paper/2507.15162/) (2025)
- [Bayesian Preference Learning for Test-Time Steerable Reward Models](https://mlsanity.com/paper/2602.08819/) (2026)
- [Pairwise Reference Alignment as a Model-Level Ordinal Observable](https://mlsanity.com/paper/2605.30758/) (2026)

## Abstract (arXiv)

Many applications, e.g. in content recommendation, sports, or recruitment, leverage the comparisons of alternatives to score those alternatives. The classical Bradley-Terry model and its variants have been widely used to do so. The historical model considers binary comparisons (victory or defeat) between alternatives, while more recent developments allow finer comparisons to be taken into account. In this article, we introduce a probabilistic model encompassing a broad variety of paired comparisons that can take discrete or continuous values. We do so by considering a well-behaved subset of the exponential family, which we call the family of generalized Bradley-Terry (GBT) models, as it includes the classical Bradley-Terry model and many of its variants. Remarkably, we prove that all GBT models are guaranteed to yield a strictly convex negative log-likelihood. Moreover, assuming a Gaussian prior on alternatives' scores, we prove that the maximum a posteriori (MAP) of GBT models, whose existence, uniqueness and fast computation are thus guaranteed, varies monotonically with respect to comparisons (the more A beats B, the better the score of A) and is Lipschitz-resilient with respect to each new comparison (a single new comparison can only have a bounded effect on all the estimated scores). These desirable properties make GBT models appealing for practical use. We illustrate some features of GBT models on simulations.