---
title: "From Metrics to Meaning — paper by Piłka et al. (2026)"
description: "Wearable devices give youth footballers, their parents and coaches many training numbers, but these figures rarely explain… 23 references."
authors: ["Tomasz Piłka", "Tomasz Kuczyński", "Mateusz Czajka"]
year: 2026
venue: "Proceedings of the 21st Workshop on Innovative Use of NLP for Building Educational Applications (BEA 2026)"
source: "https://aclanthology.org/2026.bea-1.34/"
doi: "10.18653/v1/2026.bea-1.34"
licence: "CC BY 4.0 (ACL Anthology)"
licence_url: "https://creativecommons.org/licenses/by/4.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/pilka-2026-metrics-meaning-rule-grounded-llm/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/pilka-2026-metrics-meaning-rule-grounded-llm/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2026](https://md.research.elamur.ai/free/en/papers/2026/page/30/) / From Metrics to Meaning: Rule-Grounded LLM Explanations for Data Literacy in the Case of Youth Football

# From Metrics to Meaning: Rule-Grounded LLM Explanations for Data Literacy in the Case of Youth Football

**Authors:** Tomasz Piłka, Tomasz Kuczyński and Mateusz Czajka · **Year:** 2026 · **Venue:** [Proceedings of the 21st Workshop on Innovative Use of NLP for Building Educational Applications (BEA 2026)](https://md.research.elamur.ai/free/en/venues/bea/) ([BEA 2026](https://md.research.elamur.ai/free/en/venues/bea/2026/)) · **Publisher:** [Association for Computational Linguistics](https://md.research.elamur.ai/free/en/publishers/association-for-computational-linguistics/) · **Licence:** [CC BY 4.0 (ACL Anthology)](https://creativecommons.org/licenses/by/4.0/)

**Links:** [Publisher's page](https://aclanthology.org/2026.bea-1.34/) · [PDF](https://aclanthology.org/2026.bea-1.34.pdf) · [DOI 10.18653/v1/2026.bea-1.34](https://doi.org/10.18653/v1/2026.bea-1.34) · [OpenAlex](https://openalex.org/W7166882317)

**Citations:** 0 citations (OpenAlex, 2026-10-14)

## Summary
Wearable devices give youth footballers, their parents and coaches many training numbers, but these figures rarely explain themselves. The authors treat session summarisation as a data-literacy support task rather than generic data-to-text. In their pipeline, metrics are normalised by session duration and sorted into ordinal bands. Next, a hand-crafted set of rules converts these bands into structured, prioritised facts covering four areas: the session's overall load, how intensely the body worked internally, how much the player was exposed to high speeds, and the dynamics of their movement. An LLM converts these facts into short texts for a coach, a parent or a player.

Three input conditions are compared:
- **RAW**: metrics only.
- **RULES**: rule-derived facts only.
- **ENRICHED**: facts plus raw metrics and threshold definitions, using LoRA-adapted open-weight models.

The data are 122 anonymised U15 player-session records. Ten sessions were held out and paired with audience-specific reference summaries. Rule grounding made outputs more reliable and better matched to each audience, mainly by reducing unsupported or exaggerated interpretations. Physical education teachers rated ENRICHED player-facing texts as accurate, clear and useful. The work matters because explanations for young people must be faithful and restrained, not merely fluent.

## Key points
- Dataset: 122 anonymised player-session records from one U15 football environment (partner: JUSTWIN sp. z o.o.). Ten held-out sessions were used for evaluation, and the remaining 112 for LoRA adaptation. No cross-validation was performed.
- The rule layer maps duration-normalised indicators to bands (low, medium, high, elite). Activated rules carry a priority value for content selection; for example, HighDist fires on high distance per minute.
- The RAW vs RULES comparison for coach-facing texts used GPT-4o mini as an LLM-as-a-judge with a fixed rubric. Given the small sample, results are reported descriptively.
- Expert study: nine physical education teachers assessed only the player-facing ENRICHED summaries.
- Limitations noted by the authors: ENRICHED mixes several factors (richer input, thresholds, LoRA), so their individual effects cannot be separated. LoRA was also not compared against few-shot prompting.

*Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.*

📚 Shelves (6): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 16,667 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 16,194 works · [🩺 Medicine](https://md.research.elamur.ai/free/en/shelves/medicine/) · 241 works · [📈 Economics, Econometrics and Finance](https://md.research.elamur.ai/free/en/shelves/economics-econometrics-and-finance/) · 17 works · [📈 Economics and Econometrics](https://md.research.elamur.ai/free/en/shelves/economics-econometrics-and-finance/economics-and-econometrics/) · 12 works · [🩺 Orthopedics and Sports Medicine](https://md.research.elamur.ai/free/en/shelves/medicine/orthopedics-and-sports-medicine/) · 1 work

🧭 Topics (2): [🧠 Language models & pretraining](https://md.research.elamur.ai/free/en/topics/language-models/) · 3,320 works · [📏 Evaluation & benchmarks](https://md.research.elamur.ai/free/en/topics/evaluation-benchmarks/) · 1,220 works

## Cite

Cite as: Tomasz Piłka, Tomasz Kuczyński and Mateusz Czajka. 2026. From Metrics to Meaning: Rule-Grounded LLM Explanations for Data Literacy in the Case of Youth Football. Proceedings of the 21st Workshop on Innovative Use of NLP for Building Educational Applications (BEA 2026).

The paper itself, not this page. BibTeX from the library's record of the work.

```bibtex
@inproceedings{pilka2026metrics,
  title = {{From Metrics to Meaning: Rule-Grounded LLM Explanations for Data Literacy in the Case of Youth Football}},
  author = {Tomasz Piłka, Tomasz Kuczyński and Mateusz Czajka},
  booktitle = {Proceedings of the 21st Workshop on Innovative Use of NLP for Building Educational Applications (BEA 2026)},
  year = {2026},
  doi = {10.18653/v1/2026.bea-1.34},
  url = {https://doi.org/10.18653/v1/2026.bea-1.34}
}
```

- [← previous From Logical to Computational Sparsity: Structure-Aware Block-Sparse Attention for Long-Code Completion](https://md.research.elamur.ai/free/en/papers/wang-2026-logical-computational-sparsity-structure-aware/)
- [next → From Mimesis to Metamorphosis: Evolving VLM Judges via In-Context Comparing and Knowledge Internalization](https://md.research.elamur.ai/free/en/papers/wang-2026-mimesis-metamorphosis-evolving-vlm-judges/)
- [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/page/30/)

## References

23 references, 3 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Erling Algroy, Halvard Grendstad, Amund Riiser, Tone Nybakken, Atle Hole Saeterbakken, Vidar Andersen, and Hilde Stokvold Gundersen. 2021. Motion analysis of match play in U14 male soccer players and the influence of position, competitive level, and contextual variables. International Journal of Env… [find](https://scholar.google.com/scholar?q=Erling+Algroy%2C+Halvard+Grendstad%2C+Amund+Riiser%2C+Tone+Nybakken%2C+Atle+Hole+Saeterbakken%2C+Vidar+Andersen%2C+and+Hilde+Stokvold+Gundersen%2E+2021%2E+Motion+analysis+of+match+play+in+U14+male+soccer+players+and+)
2. Siti A. Atan, Andrew Foskett, and Ajmol Ali. 2016. Motion analysis of match play in new zealand U13 to U15 age-group soccer players. Journal of Strength and Conditioning Research, 30(9):2416–2423. [find](https://scholar.google.com/scholar?q=Siti+A%2E+Atan%2C+Andrew+Foskett%2C+and+Ajmol+Ali%2E+2016%2E+Motion+analysis+of+match+play+in+new+zealand+U13+to+U15+age%2Dgroup+soccer+players%2E+Journal+of+Strength+and+Conditioning+Research%2C+30%289%29%3A2416%E2%80%932423%2E)
3. Wenhu Chen, Jianshu Chen, Yu Su, Zhiyu Chen, and William Yang Wang. 2020. Logicnlg: [Logical natural language generation from open-domain tables](https://md.research.elamur.ai/free/en/papers/chen-2020-logical-natural-language-generation-open/). In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pages 7929– 7942, Online. Association for Computational Linguis…
4. Terje Dalen and Havard Loras. 2019. Monitoring training and match physical load in junior soccer players: Starters versus substitutes. Sports, 7(3):70. [find](https://scholar.google.com/scholar?q=Terje+Dalen+and+Havard+Loras%2E+2019%2E+Monitoring+training+and+match+physical+load+in+junior+soccer+players%3A+Starters+versus+substitutes%2E+Sports%2C+7%283%29%3A70%2E)
5. Martina Galletti and Valeria Cesaroni. 2025. From end-users to co-designers: Lessons from teachers. In Proceedings of the 20th Workshop on Innovative Use of NLP for Building Educational Applications (BEA 2025), pages 505–516, Vienna, Austria. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Martina+Galletti+and+Valeria+Cesaroni%2E+2025%2E+From+end%2Dusers+to+co%2Ddesigners%3A+Lessons+from+teachers%2E+In+Proceedings+of+the+20th+Workshop+on+Innovative+Use+of+NLP+for+Building+Educational+Applications+%28)
6. Albert Gatt and Emiel Krahmer. 2018. Survey of the state of the art in natural language generation. Journal of Artificial Intelligence Research, 61:65–170. [find](https://scholar.google.com/scholar?q=Albert+Gatt+and+Emiel+Krahmer%2E+2018%2E+Survey+of+the+state+of+the+art+in+natural+language+generation%2E+Journal+of+Artificial+Intelligence+Research%2C+61%3A65%E2%80%93170%2E)
7. Marcus P. Hannon, Nicholas M. Coleman, Lloyd J. F. Parker, John McKeown, Viswanath B. Unnithan, Graeme L. Close, Barry Drust, and James P. Morton. 2021. Seasonal training and match load and micro-cycle periodization in male premier league academy soccer players. Journal of Sports Sciences, 39(16):18… [find](https://scholar.google.com/scholar?q=Marcus+P%2E+Hannon%2C+Nicholas+M%2E+Coleman%2C+Lloyd+J%2E+F%2E+Parker%2C+John+McKeown%2C+Viswanath+B%2E+Unnithan%2C+Graeme+L%2E+Close%2C+Barry+Drust%2C+and+James+P%2E+Morton%2E+2021%2E+Seasonal+training+and+match+load+and+micro%2Dcycl)
8. J. A. Harley, C. A. Barnes, M. Portas, R. Lovell, S. Barrett, D. Paul, and M. Weston. 2010. Motion analysis of match-play in elite U12 to U16 age-group soccer players. Journal of Sports Sciences, 28(13):1391– 1397. [find](https://scholar.google.com/scholar?q=J%2E+A%2E+Harley%2C+C%2E+A%2E+Barnes%2C+M%2E+Portas%2C+R%2E+Lovell%2C+S%2E+Barrett%2C+D%2E+Paul%2C+and+M%2E+Weston%2E+2010%2E+Motion+analysis+of+match%2Dplay+in+elite+U12+to+U16+age%2Dgroup+soccer+players%2E+Journal+of+Sports+Sciences%2C+28%281)
9. Ekaterina Kochmar, Kaushal Maurya, Kseniia Petukhova, K. V. Aditya Srivatsa, Anaïs Tack, and Justin Vasselli. 2025. Findings of the BEA 2025 shared task on pedagogical ability assessment of AI-powered tutors. In Proceedings of the 20th Workshop on Innovative Use of NLP for Building Educational Appli… [find](https://scholar.google.com/scholar?q=Ekaterina+Kochmar%2C+Kaushal+Maurya%2C+Kseniia+Petukhova%2C+K%2E+V%2E+Aditya+Srivatsa%2C+Ana%C3%AFs+Tack%2C+and+Justin+Vasselli%2E+2025%2E+Findings+of+the+BEA+2025+shared+task+on+pedagogical+ability+assessment+of+AI%2Dpowered)
10. Clifford Konold, Trevor L. Higgins, Stephen J. Russell, and Khalil Khalil. 2015. Data seen through different lenses. Educational Studies in Mathematics, 88(3):305–325. [find](https://scholar.google.com/scholar?q=Clifford+Konold%2C+Trevor+L%2E+Higgins%2C+Stephen+J%2E+Russell%2C+and+Khalil+Khalil%2E+2015%2E+Data+seen+through+different+lenses%2E+Educational+Studies+in+Mathematics%2C+88%283%29%3A305%E2%80%93325%2E)
11. Ellen B. Mandinach and Edith S. Gummer. 2016. What does it mean for teachers to be data literate: Laying out the skills, knowledge, and dispositions. Teaching and Teacher Education, 60:366–376. [find](https://scholar.google.com/scholar?q=Ellen+B%2E+Mandinach+and+Edith+S%2E+Gummer%2E+2016%2E+What+does+it+mean+for+teachers+to+be+data+literate%3A+Laying+out+the+skills%2C+knowledge%2C+and+dispositions%2E+Teaching+and+Teacher+Education%2C+60%3A366%E2%80%93376%2E)
12. Joshua Maynez, Shashi Narayan, Bernd Bohnet, and Ryan McDonald. 2020. [On faithfulness and factuality in abstractive summarization](https://md.research.elamur.ai/free/en/papers/maynez-2020-faithfulness-factuality-abstractive-summarization/). In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pages 1906–1919, Online. Association for Computational Linguistics.
13. Konstantinos Michos, Maria-Luisa Schmitz, and Dominik Petko. 2023. Teachers’ data literacy for learning analytics: A central predictor for digital data use in upper secondary schools. Education and Information Technologies, 28:14453–14471. [find](https://scholar.google.com/scholar?q=Konstantinos+Michos%2C+Maria%2DLuisa+Schmitz%2C+and+Dominik+Petko%2E+2023%2E+Teachers%E2%80%99+data+literacy+for+learning+analytics%3A+A+central+predictor+for+digital+data+use+in+upper+secondary+schools%2E+Education+and+In)
14. Samraj Moorjani, Adit Krishnan, Hari Sundaram, Ewa Maslowska, and Aravind Sankar. 2022. Audience-centric natural language generation via style infusion. In Findings of the Association for Computational Linguistics: EMNLP 2022, pages 1919–1932, Abu Dhabi, United Arab Emirates. Association for Computa… [find](https://scholar.google.com/scholar?q=Samraj+Moorjani%2C+Adit+Krishnan%2C+Hari+Sundaram%2C+Ewa+Maslowska%2C+and+Aravind+Sankar%2E+2022%2E+Audience%2Dcentric+natural+language+generation+via+style+infusion%2E+In+Findings+of+the+Association+for+Computationa)
15. Sankalan Pal Chowdhury, Nico Daheim, Ekaterina Kochmar, Jakub Macina, Donya Rooein, Mrinmaya Sachan, and Shashank Sonkar. 2025a. Large language models for education: Understanding the needs of stakeholders, current capabilities, and the path forward. In Proceedings of the 20th Workshop on Innovative… [find](https://scholar.google.com/scholar?q=Sankalan+Pal+Chowdhury%2C+Nico+Daheim%2C+Ekaterina+Kochmar%2C+Jakub+Macina%2C+Donya+Rooein%2C+Mrinmaya+Sachan%2C+and+Shashank+Sonkar%2E+2025a%2E+Large+language+models+for+education%3A+Understanding+the+needs+of+stakeho)
16. Sankalan Pal Chowdhury, Terry Jingchen Zhang, Donya Rooein, Dirk Hovy, Tanja Käser, and Mrinmaya Sachan. 2025b. Educators’ perceptions of large language models as tutors: Comparing human and AI tutors in a blind text-only setting. In Proceedings of the 20th Workshop on Innovative Use of NLP for Buil… [find](https://scholar.google.com/scholar?q=Sankalan+Pal+Chowdhury%2C+Terry+Jingchen+Zhang%2C+Donya+Rooein%2C+Dirk+Hovy%2C+Tanja+K%C3%A4ser%2C+and+Mrinmaya+Sachan%2E+2025b%2E+Educators%E2%80%99+perceptions+of+large+language+models+as+tutors%3A+Comparing+human+and+AI+tutors)
17. Luiz Henrique Palucci Vieira, Christopher Carling, Fabio Augusto Barbieri, Rodrigo Aquino, and Paulo Roberto Pereira Santiago. 2019. Match running performance in young soccer players: A systematic review. Sports Medicine, 49(6):917–931. [find](https://scholar.google.com/scholar?q=Luiz+Henrique+Palucci+Vieira%2C+Christopher+Carling%2C+Fabio+Augusto+Barbieri%2C+Rodrigo+Aquino%2C+and+Paulo+Roberto+Pereira+Santiago%2E+2019%2E+Match+running+performance+in+young+soccer+players%3A+A+systematic+rev)
18. Kseniia Petukhova and Ekaterina Kochmar. 2025. Intent matters: Enhancing AI tutoring with fine-grained pedagogical intent annotation. In Proceedings of the 20th Workshop on Innovative Use of NLP for Building Educational Applications (BEA 2025), pages 860– 872, Vienna, Austria. Association for Comput… [find](https://scholar.google.com/scholar?q=Kseniia+Petukhova+and+Ekaterina+Kochmar%2E+2025%2E+Intent+matters%3A+Enhancing+AI+tutoring+with+fine%2Dgrained+pedagogical+intent+annotation%2E+In+Proceedings+of+the+20th+Workshop+on+Innovative+Use+of+NLP+for+B)
19. Ratish Puduppully, Li Dong, and Mirella Lapata. 2019. Data-to-text generation with content selection and planning. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 33, pages 6908–6915. [find](https://scholar.google.com/scholar?q=Ratish+Puduppully%2C+Li+Dong%2C+and+Mirella+Lapata%2E+2019%2E+Data%2Dto%2Dtext+generation+with+content+selection+and+planning%2E+In+Proceedings+of+the+AAAI+Conference+on+Artificial+Intelligence%2C+volume+33%2C+pages+69)
20. Ehud Reiter and Robert Dale. 2000. Building Natural Language Generation Systems. Cambridge University Press. [find](https://scholar.google.com/scholar?q=Ehud+Reiter+and+Robert+Dale%2E+2000%2E+Building+Natural+Language+Generation+Systems%2E+Cambridge+University+Press%2E)
21. Hieu Tran, Zonghai Yao, Lingxi Li, and Hong Yu. 2025. Readctrl: Personalizing text generation with readability-controlled instruction learning. In Proceedings of the Fourth Workshop on Intelligent and Interactive Writing Assistants (In2Writing 2025), pages 19–36, Albuquerque, New Mexico, US. Associa… [find](https://scholar.google.com/scholar?q=Hieu+Tran%2C+Zonghai+Yao%2C+Lingxi+Li%2C+and+Hong+Yu%2E+2025%2E+Readctrl%3A+Personalizing+text+generation+with+readability%2Dcontrolled+instruction+learning%2E+In+Proceedings+of+the+Fourth+Workshop+on+Intelligent+and)
22. Natercia Valle, Pavlo Antonenko, Kara Dawson, and Anne Corinne Huggins-Manley. 2021. Staying on target: A systematic literature review on learner-facing learning analytics dashboards. British Journal of Educational Technology, 52(4):1724–1748. [find](https://scholar.google.com/scholar?q=Natercia+Valle%2C+Pavlo+Antonenko%2C+Kara+Dawson%2C+and+Anne+Corinne+Huggins%2DManley%2E+2021%2E+Staying+on+target%3A+A+systematic+literature+review+on+learner%2Dfacing+learning+analytics+dashboards%2E+British+Journal+)
23. Sam Wiseman, Stuart Shieber, and Alexander M. Rush. 2017. [Challenges in data-to-document generation](https://md.research.elamur.ai/free/en/papers/wiseman-2017-challenges-data-document-generation/). In Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing, pages 2253–2263, Copenhagen, Denmark. Association for Computational Linguistics.

*Updated 1 November 2026*