**Size** - **Rows:** 718,470 across 5 tables, which matches the dataset card and the row counts I queried. - `raw_kaggle`: 237,205 - `return_kaggle`: 237,196 - `serve_kaggle`: 237,185 - `players_tournament_man_`: 6,422 - `players_man_`: 462 - **Columns:** the dataset card lists 78 columns in total: 21 each in the three match tables, 8 in `players_tournament_man_` and 7 in `players_man_`. - The metadata overview I queried reports fewer: 17, 17, 17, 4 and 3, which sums to 58. - I did not resolve this mismatch. I'd trust the card's 78, since it gives the full per-table breakdown, but the metadata count may reflect a different way of counting columns. **Time period** - **Match tables:** `Date` is stored as text. Once parsed, all 237,205 `raw_kaggle` rows, 237,196 `return_kaggle` rows and 237,185 `serve_kaggle` rows converted cleanly. All three tables span **13 Aug 2001 to 28 Oct 2024**, so the data covers roughly 23 years. - **Earlier ranges were misleading:** ranges like "1-Apr-2013 to 9-Sep-2024" on the card come from sorting the date text alphabetically, not chronologically. Don't rely on them. - **Players-by-tournament table:** `players_tournament_man_.year` runs from 2002 to 2024 with no nulls. - **Player summary table:** `players_man_.playing_years` is a text list of years (for example 2002 through 2024), not a proper date column. The dataset therefore covers about 718 thousand rows of men's tennis data from 2001 to 2024. The match-level tables hold one row per player per match, and the two player tables are smaller summaries.