**Size:** the dataset has 5 tables and 718,470 rows in total. | Table | Rows | |---|---| | raw.raw_kaggle | 237,205 | | raw.return_kaggle | 237,196 | | raw.serve_kaggle | 237,185 | | raw.players_tournament_man_ | 6,422 | | raw.players_man_ | 462 | **Columns:** the two sources I have disagree, so I can't give a single figure. - The dataset card lists 21, 21, 21, 8 and 7 columns for the five tables, which is 78 in total. - The column counts from `analytics.table_overview` (step 0) are 17, 17, 17, 4 and 3, which is 58 in total. - I did not reconcile the two. I'd treat the card's 78 as the likely true physical count and the overview figure as possibly undercounting. **Time period:** the data covers about 23 years, from August 2001 to October 2024. - **Match tables:** `Date` is stored as text in all three match tables (raw, return and serve). Every row parsed successfully (237,205, 237,196 and 237,185 rows respectively). The earliest date is 2001-08-13 and the latest is 2024-10-28 in each table (steps 3, 5 and 6). - **Tournament table:** `players_tournament_man_.year` runs from 2002 to 2024, with no nulls. - **Player table:** `players_man_.playing_years` is a text list of years with no date type. The card shows years from 2002 to 2024. Because `Date` is text, a plain MIN/MAX on it sorts alphabetically. That is why the card shows misleading ranges such as "1-Apr-2013..9-Sep-2024" or a serve start of "10‑Apr‑2006". The parsed dates above are the reliable ones.