# Input Matters: Evaluating Input Structure's Impact on LLM Summaries of Sports Play-by-Play

arXiv: https://arxiv.org/abs/2510.21034 · HTML: https://mlsanity.com/paper/2510.21034/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 7 of its 19 references that are in the map.
- First posted on arXiv 23 Oct 2025.

## Builds on (most cited first, 7 of 7)

- [ToTTo: A Controlled Table-To-Text Generation Dataset](https://mlsanity.com/paper/2004.14373/) (2020)
- [Leak, Cheat, Repeat: Data Contamination and Evaluation Malpractices in Closed-Source LLMs](https://mlsanity.com/paper/2402.03927/) (2024)
- [Challenges in Data-to-Document Generation](https://mlsanity.com/paper/1707.08052/) (2017)
- [Data-to-Text Generation with Content Selection and Planning](https://mlsanity.com/paper/1809.00582/) (2018)
- [When Reasoning Meets Information Aggregation: A Case Study with Sports Narratives](https://mlsanity.com/paper/2406.12084/) (2024)
- [Extending Context Window of Large Language Models from a Distributional Perspective](https://mlsanity.com/paper/2410.01490/) (2024)
- [Improving Factual Accuracy of Neural Table-to-Text Output by Addressing Input Problems in ToTTo](https://mlsanity.com/paper/2404.04103/) (2024)

## Abstract (arXiv)

A major concern when deploying LLMs in accuracy-critical domains such as sports reporting is that the generated text may not faithfully reflect the input data. We quantify how input structure affects hallucinations and other factual errors in LLM-generated summaries of NBA play-by-play data, across three formats: row-structured, JSON and unstructured. We manually annotated 3,312 factual errors across 180 game summaries produced by two models, Llama-3.1-70B and Qwen2.5-72B. Input structure has a strong effect: JSON input reduces error rates by 69% for Llama and 65% for Qwen compared to unstructured input, while row-structured input reduces errors by 54% for Llama and 51% for Qwen. A two-way repeated measures ANOVA shows that input structure accounts for over 80% of the variance in error rates, with Tukey HSD post hoc tests confirming statistically significant differences between all input formats.