ForzaJuve1 commited on
Commit
04b970c
1 Parent(s): 813ba16

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +46 -1
README.md CHANGED
@@ -164,8 +164,8 @@ For the flat variables without nested structures:
164
  "WindSpeed": 13
165
  }
166
  ```
167
- ```
168
  For the nested variables, since the nested structures are complex and long, it will be difficult to show a complete instance here. I will instead provide the structure.
 
169
  {
170
  "MatchEvent": Features({
171
  "1-First Half": Sequence(Features({
@@ -657,5 +657,50 @@ For the nested variables, since the nested structures are complex and long, it w
657
  }
658
  ```
659
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
660
  ## Citation
661
  Mikhail Zhilkin @cervus (2021). "UEFA Euro 2020." Hosted by [data.world]. Available at: “https://data.world/cervus/uefa-euro-2020".
 
164
  "WindSpeed": 13
165
  }
166
  ```
 
167
  For the nested variables, since the nested structures are complex and long, it will be difficult to show a complete instance here. I will instead provide the structure.
168
+ ```
169
  {
170
  "MatchEvent": Features({
171
  "1-First Half": Sequence(Features({
 
657
  }
658
  ```
659
 
660
+ ### Data Fields
661
+ - `HomeTeamName`: Name of the home team.
662
+ - `AwayTeamName`: Name of the away team.
663
+ - `DateandTimeCET`: Date and time of the match in Central European Time (CET).
664
+ - `MatchID`: A unique identifier for the match.
665
+ - `RoundName`: The round of the competition (e.g. semi-final, final).
666
+ - `Stage`: The competition stage (e.g. Group stage, Knockout Stage).
667
+ - `MatchDay`: The matchday number within the competition.
668
+ - `Session`: The session number of the match.
669
+ - `MatchMinute`: Total minutes played in the match.
670
+ - `InjuryTime`: Total injury time added.
671
+ - `NumberOfPhases`: Number of phases in the match.
672
+ - `Phase`: Current phase of the match.
673
+ - `ScoreHome`: Final score for the home team.
674
+ - `ScoreAway`: Final score for the away team.
675
+ - `MatchStatus`: Status of the match (e.g., played).
676
+ - `StadiumID`: Unique identifier for the stadium.
677
+ - `RefereeWebName`: Name of the referee.
678
+ - `TotalNumberofMatchesRefereed`: Total number of matches refereed by the official.
679
+ - `NumberofMatchesRefereedinGroupStage`: Number of matches refereed in the group stage.
680
+ - `NumberofMatchesRefereedinKnockoutStage`: Number of matches refereed in the knockout stage.
681
+ - `AssistantRefereeWebName`: Name of the assistant referee.
682
+ - `Humidity`: Humidity percentage at the match venue.
683
+ - `Temperature`: Temperature at the match venue in Celsius.
684
+ - `WindSpeed`: Wind speed at the match venue in km/h.
685
+ - `MatchEvent`: The event information in different phases of the game.
686
+ - `TeamLineUps`: The starting 11, benched players, and staff information for home and away team.
687
+ - `TeamStats`: The detailed team-level statistics for home and away team categorized into different game areas (eg. attacking, defending, etc.).
688
+ - `PlayerStats`: The detailed player-level statistics for players from home and away team categorized into different game areas (eg. attacking, defending, etc.).
689
+ - `PlayerPreMatchInfo`: Player-based pre-match information for home and away team.
690
+
691
+ ### Data Splits
692
+ Since the final dataset contains only 51 rows, and many supported tasks using this dataset might not need train-test-validation splits, I only keep one 1 split of train dataset.
693
+
694
+ ## Dataset Creation
695
+
696
+ ### Curation Rationale
697
+ The primary purpose of the creation of this dataset is to combine useful information on different levels of matches (individual player, team, match, event, etc.) to facilitate user access and insight-extraction from this tournament's data.
698
+
699
+ ### Source Data
700
+ "UEFA Euro 2020" from “https://data.world/cervus/uefa-euro-2020".
701
+
702
+ ## Limitations
703
+ If users hope to conduct advanced machine learning modeling using the dataset, the sample size might be too small to attain generalizable conclusions, and the preprocessing step could involve extensive parsing and flattening of variables. The dataset is more suitable for summrization, story telling, insight and feature extraction, and analysis rather than modeling. There are also a small number of missing data that could be hard to impute in the nested variables.
704
+
705
  ## Citation
706
  Mikhail Zhilkin @cervus (2021). "UEFA Euro 2020." Hosted by [data.world]. Available at: “https://data.world/cervus/uefa-euro-2020".