A Benchmark for Multi-Party Negotiation Games from Real Negotiation Data

TL;DR

This paper introduces a new benchmark and evaluation framework for multi-party negotiation games based on real data, emphasizing sequences of commitments rather than final outcomes.

Contribution

It presents a configurable negotiation game generator, document-grounded instances from climate negotiations, and baseline solvers, highlighting the need for novel methods that handle partial commitments.

Findings

No single solver dominates across different game regimes.

Performance varies with the structural properties of the negotiation game.

Existing methods are limited in handling diverse strategic regimes.

Abstract

Many real-world multi-party negotiations unfold as sequences of binding, action-level commitments rather than a single final outcome, yet this regime remains under-studied in existing benchmarks. We introduce a benchmark and evaluation framework for this setting, combining a configurable negotiation game generator with document-grounded instances derived from a climate negotiation exercise. We also provide several baseline solvers. Exact evaluation on small games and comparative evaluation on larger instances show that no solver dominates across regimes; performance depends on the structural properties of the game. These results motivate the creation of novel negotiation methods that value partial commitments robustly across diverse strategic regimes. Code and data for the benchmark are available at: https://anonymous.4open.science/r/negotiation_MARL-46B8