WEKO3
Item
Information-Theoretical Analysis of Team Dynamics in Football Matches
https://doi.org/10.15102/0002000971
https://doi.org/10.15102/0002000971852ec9a7-fbce-43df-af52-17529db71916
| Name / File | License | Actions |
|---|---|---|
ChengYi-ShanFulltext.pdf (9.2 MB)
|
|
|
|
ChengYi-ShanExamAbstract.pdf (48 KB)
|
|
| Item type | 学位論文 / Thesis or Dissertation(1) | |||||||
|---|---|---|---|---|---|---|---|---|
| PubDate | 2025-09-05 | |||||||
| Title | ||||||||
| Title | サッカー試合におけるチームダイナミクスの情報理論的分析 | |||||||
| Language | ja | |||||||
| Title | ||||||||
| Title | Information-Theoretical Analysis of Team Dynamics in Football Matches | |||||||
| Language | en | |||||||
| Language | ||||||||
| Language | eng | |||||||
| Keyword | ||||||||
| Subject Scheme | Other | |||||||
| Subject | Information theory | multi-agent systems | network science | team sports science | |||||||
| Resource Type | ||||||||
| Resource Type Identifier | http://purl.org/coar/resource_type/c_db06 | |||||||
| Resource Type | doctoral thesis | |||||||
| Identifier Registration | ||||||||
| Identifier Registration | 10.15102/0002000971 | |||||||
| Identifier Registration Type | JaLC | |||||||
| Access Right | ||||||||
| Access Rights | open access | |||||||
| Access Rights URI | http://purl.org/coar/access_right/c_abf2 | |||||||
| Author |
Cheng, Yi-Shan
× Cheng, Yi-Shan
|
|||||||
| Abstract | ||||||||
| Description Type | Abstract | |||||||
| Description | This thesis investigates how coordinated behavior emerges within football teams by applying information-theoretic frameworks to spatiotemporal player tracking data. The first part of the thesis applies the causal emergence framework to 34 professional football matches, demonstrating that spatially defined macro-level features provide stronger explanatory power than their micro-level origins based on published work. This analysis reveals that emergent coordination is closely linked to match-critical situations including possession rate and shooting. The second part applies integrated information to the same dataset, revealing higher integration during defensive phases, which is opposite to the causal emergence results. This divergence reflects the frameworks’ differing emphases: integrated information captures internal irreducibility, while causal emergence highlights differentiated roles. Simulation based validation supports this interpretation. The third part analyzes transfer entropy matrices to uncover latent patterns of players’ interaction. Clustering in TE space reveals consistent separation between static and active phases, while non-negative matrix factorization (NMF) identifies dominant interaction components linked to key match events. Together, these analyses demonstrate the complementary strengths of causal emergence, integrated information, and transfer entropy in characterizing multiscale patterns of coordination in team sports. The thesis offers practical insights relevant to fields ranging from sports analytics to collective systems. |
|||||||
| Language | en | |||||||
| Exam Date | ||||||||
| 2025-07-03 | ||||||||
| Degree Conferral Date | ||||||||
| Date Granted | 2025-08-31 | |||||||
| Degree | ||||||||
| Degree Name | Doctor of Philosophy | |||||||
| Degree Referral Number | ||||||||
| Dissertation Number | 甲第204号 | |||||||
| Degree Conferrral Institution | ||||||||
| Degree Grantor Name Identifier Scheme | kakenhi | |||||||
| Degree Grantor Name Identifier | 38005 | |||||||
| Degree Grantor Name | Okinawa Institute of Science and Technology Graduate University | |||||||
| Version Format | ||||||||
| Version Type | VoR | |||||||
| Version Type Resource | http://purl.org/coar/version/c_970fb48d4fbd8a85 | |||||||
| Copyright Information | ||||||||
| Rights | © 2025 The Author. | |||||||
