Speaker Points

Absolute Metrics and Ranking at Swiss Tournaments -- British Parliamentary Debating and Speaker Points: An Empirical Analysis

Abstract

Standard BP debating tournaments consist of Swiss preliminary rounds followed by elimination rounds. For preliminary round debates, teams receive team points, representing win/loss, and speaker points, an absolute/cardinal measure of performance. 

We analyse data from some 337 tournaments over 2020--2024. We find that ranking teams by (total) speaker points as the first metric is more accurate than ranking teams by team points (and various other metrics) with respect to empirical elimination round results. We suggest reasons for this. The result holds for tournaments of different length, size, year, and region. Our work is notable---to our knowledge, the first of its kind---for investigating an absolute/cardinal metric of performance at tournaments and evaluating metrics on empirical results (as opposed to simulation) post-ranking. 

We examine objections to ranking teams by speaker points, including allegations of bias, which we investigate and find to be unfounded in our companion paper. We discuss generalisations and explore several extensions, including hybrids of speaker points and team points, opponent-adjusted team points, retrospective tapers, and judge-adjusted speaker points. For the last, we propose a method for adjusting judges' speaker points for mean and spread that accounts for the teams they judged, and we find that it offers an improvement on (unadjusted) speaker points. 

Keywords: speaker points, absolute/cardinal metrics, ranking, Swiss tournaments, BP debating