# The Leaderboard Illusion

arXiv: https://arxiv.org/abs/2504.20879 · HTML: https://mlsanity.com/paper/2504.20879/

## Key facts

- Cited by 66 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 29 Apr 2025.

## Built on by (most cited first, 30 of 66)

- [Neither Valid nor Reliable? Investigating the Use of LLMs as Judges](https://mlsanity.com/paper/2508.18076/) (2025)
- [HeuriGym: An Agentic Benchmark for LLM-Crafted Heuristics in Combinatorial Optimization](https://mlsanity.com/paper/2506.07972/) (2025)
- [Evaluating LLM Metrics Through Real-World Capabilities](https://mlsanity.com/paper/2505.08253/) (2025)
- [Is Your LLM Overcharging You? Tokenization, Transparency, and Incentives](https://mlsanity.com/paper/2505.21627/) (2025)
- [Hidden Measurement Error in LLM Pipelines Distorts Annotation, Evaluation, and Benchmarking](https://mlsanity.com/paper/2604.11581/) (2026)
- [Autonomous Research Agents: A Survey of AI Scientists and the Verification Gap](https://mlsanity.com/paper/2608.05179/) (2026)
- [Dropping Just a Handful of Preferences Can Change Top Large Language Model Rankings](https://mlsanity.com/paper/2508.11847/) (2025)
- [Position: LLM Watermarking Should Align Stakeholders' Incentives for Practical Adoption](https://mlsanity.com/paper/2510.18333/) (2025)
- [Visually Prompted Benchmarks Are Surprisingly Fragile](https://mlsanity.com/paper/2512.17875/) (2025)
- [Exploiting Leaderboards for Large-Scale Distribution of Malicious Models](https://mlsanity.com/paper/2507.08983/) (2025)
- [Reasoning Core: A Scalable RL Environment for LLM Symbolic Reasoning](https://mlsanity.com/paper/2509.18083/) (2025)
- [A Single Character can Make or Break Your LLM Evals](https://mlsanity.com/paper/2510.05152/) (2025)
- [Emergence WebVoyager: Toward Consistent and Transparent Evaluation of (Web) Agents in The Wild](https://mlsanity.com/paper/2603.29020/) (2026)
- [VERA-MH Concept Paper](https://mlsanity.com/paper/2510.15297/) (2025)
- [K-Sort Eval: Efficient Preference Evaluation for Visual Generation via Corrected VLM-as-a-Judge](https://mlsanity.com/paper/2602.09411/) (2026)
- [On Evaluating LLM Alignment by Evaluating LLMs as Judges](https://mlsanity.com/paper/2511.20604/) (2025)
- [Leaderboard Incentives: Model Rankings under Strategic Post-Training](https://mlsanity.com/paper/2603.08371/) (2026)
- [LLM Evaluation as Tensor Completion: Low Rank Structure and Semiparametric Efficiency](https://mlsanity.com/paper/2604.05460/) (2026)
- [Evaluating the Unseen Capabilities: How Many Theorems Do LLMs Know?](https://mlsanity.com/paper/2506.02058/) (2025)
- [A Protocol for Trustless Verification Under Uncertainty](https://mlsanity.com/paper/2507.00631/) (2025)
- [EigenBench: A Comparative Behavioral Measure of Value Alignment](https://mlsanity.com/paper/2509.01938/) (2025)
- [Reward Models are Metrics in a Trench Coat](https://mlsanity.com/paper/2510.03231/) (2025)
- [compar:IA: The French Government's LLM arena to collect French-language human prompts and preference data](https://mlsanity.com/paper/2602.06669/) (2026)
- [MedArena: Comparing LLMs for Medicine-in-the-Wild Clinician Preferences](https://mlsanity.com/paper/2603.15677/) (2026)
- [Who Defines "Best"? Towards Interactive, User-Defined Evaluation of LLM Leaderboards](https://mlsanity.com/paper/2604.21769/) (2026)
- [Reflections and New Directions for Human-Centered Large Language Models](https://mlsanity.com/paper/2605.06901/) (2026)
- [From Uncertain Judgments to Calibrated Rankings: Conformal Elo Estimation for LLM Evaluation](https://mlsanity.com/paper/2606.13221/) (2026)
- [Right Answer, Wrong Method: Shortcut Hacking Misleads the Evaluation of LLM Reasoning on Frontier Science Benchmarks](https://mlsanity.com/paper/2608.02442/) (2026)
- [K-Bench: measuring model performance on real scientific agent requests](https://mlsanity.com/paper/2608.21601/) (2026)
- [Rank Confidence Sequences:Anytime-valid Leaderboards](https://mlsanity.com/paper/2609.32211/) (2026)

## Abstract (arXiv)

Measuring progress is fundamental to the advancement of any scientific field. As benchmarks play an increasingly central role, they also grow more susceptible to distortion. Chatbot Arena has emerged as the go-to leaderboard for ranking the most capable AI systems. Yet, in this work we identify systematic issues that have resulted in a distorted playing field. We find that undisclosed private testing practices benefit a handful of providers who are able to test multiple variants before public release and retract scores if desired. We establish that the ability of these providers to choose the best score leads to biased Arena scores due to selective disclosure of performance results. At an extreme, we identify 27 private LLM variants tested by Meta in the lead-up to the Llama-4 release. We also establish that proprietary closed models are sampled at higher rates (number of battles) and have fewer models removed from the arena than open-weight and open-source alternatives. Both these policies lead to large data access asymmetries over time. Providers like Google and OpenAI have received an estimated 19.2% and 20.4% of all data on the arena, respectively. In contrast, a combined 83 open-weight models have only received an estimated 29.7% of the total data. We show that access to Chatbot Arena data yields substantial benefits; even limited additional data can result in relative performance gains of up to 112% on the arena distribution, based on our conservative estimates. Together, these dynamics result in overfitting to Arena-specific dynamics rather than general model quality. The Arena builds on the substantial efforts of both the organizers and an open community that maintains this valuable evaluation platform. We offer actionable recommendations to reform the Chatbot Arena's evaluation framework and promote fairer, more transparent benchmarking for the field