Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1].

Metric-Distortion Bounds under Limited Information

Authors: Ioannis Anagnostides, Dimitris Fotakis, Panagiotis Patsilinakos

JAIR 2022 | Venue PDF | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental From an experimental standpoint, we present several empirical findings on real-life voting applications, comparing the scoring systems employed in practice with a mechanism explicitly minimizing (metric) distortion. Interestingly, for our case studies, we find that the winner in the actual competition is typically the candidate who minimizes the distortion.
Researcher Affiliation Academia Ioannis Anagnostides EMAIL Carnegie Mellon University, Pittsburgh, USA Dimitris Fotakis EMAIL National Technical University of Athens, Greece Panagiotis Patsilinakos EMAIL National Technical University of Athens, Greece
Pseudocode Yes Algorithm 1: Domination Root Input: Set of candidates C, Pairwise comparison oracle O; Output: Winner w C; 1. Initialize S := C; 2. Construct arbitrarily a set Π of S/2 pairings from S; 3. For every {a, b} Π remove O(a, b) from S; 4. If |S| = 1 return w S; otherwise, continue from step 2;
Open Source Code Yes Our code is publicly available at https://github.com/ioannis Anagno/Voting-Metric Distortion.
Open Datasets Yes For our experiments we used a dataset from Kaggle.
Dataset Splits No The paper analyzes "real-life voting applications" (Eurovision, Formula One), which typically involve complete historical data for analysis rather than train/test/validation splits for model training. The text does not specify any dataset splits.
Hardware Specification No The induced LPs in Minimax-LP will be solved via the Gurobi software (Gurobi Optimization, 2021). No other specific hardware details are provided for the experiments.
Software Dependencies Yes The induced LPs in Minimax-LP will be solved via the Gurobi software (Gurobi Optimization, 2021).
Experiment Setup Yes Eurovision employs a specific positional scoring system which works as follows. Every country assigns 12 points to its highest preference, 10 points to its second-highest preference, and from 8 to 1 points to each of its next 8 preferences, respectively... The scoring rule employed in F1 assigns to the first 10 drivers the points 25, 18, 15, 12, 10, 8, 6, 4, 2, 1, respectively...